ebook
THE GUILFORD PRESS
Biofeedback
Also Available
Biopsychosocial Assessment in Clinical Health Psychology
Edited by Frank Andrasik, Jeffrey L. Goodie,
and Alan L. Peterson
Passive Muscle Relaxation:
A Program for Client Use (CD-ROM)
Mark S. Schwartz and Stephen N. Haynes
Biofeedback
A Practitioner’s Guide
F o u r t h
E d i t i o n
Edited by
Mark S. Schwartz
Frank Andrasik
THE GUILFORD PRESS
New York London
© 2016 The Guilford Press
A Division of Guilford Publications, Inc.
370 Seventh Avenue, Suite 1200, New York, NY 10001
www.guilford.com
Paperback edition 2017
All rights reserved
No part of this book may be reproduced, translated, stored in a retrieval
system, or transmitted, in any form or by any means, electronic, mechanical,
photocopying, microfilming, recording, or otherwise, without written permission
from the publisher.
Printed in the United States of America
This book is printed on acid-free paper.
Last digit is print number: 9
8
7
6
5
4
3 2
The authors have checked with sources believed to be reliable in their efforts to
provide information that is complete and generally in accord with the standards
of practice that are accepted at the time of publication. However, in view of the
possibility of human error or changes in behavioral, mental health, or medical
sciences, neither the authors, nor the editors and publisher, nor any other party
who has been involved in the preparation or publication of this work warrants
that the information contained herein is in every respect accurate or complete,
and they are not responsible for any errors or omissions or the results obtained
from the use of such information. Readers are encouraged to confirm the
information contained in this book with other sources.
Library of Congress Cataloging-in-Publication Data
Names: Schwartz, Mark S. (Mark Stephen), editor. | Andrasik, Frank, 1949–
editor.
Title: Biofeedback : a practitioner’s guide / edited by Mark S. Schwartz,
Frank Andrasik.
Description: Fourth edition. | New York : The Guilford Press, [2016] |
Includes bibliographical references and index.
Identifiers: LCCN 2015049500 | ISBN 9781462522545 (hardcover : alk. paper)
| ISBN 9781462531943 (paperback : alk. paper)
Subjects: LCSH: Biofeedback training.
Classification: LCC RC489.B53 S39 2016 | DDC 615.8/514—dc23
LC record available at http://lccn.loc.gov/2015049500
In Memoriam
David E. Krebs, DPT, PhD
We include this dedication in profound honor and in memory of David E. Krebs because he
was an integral part of the first three editions of this book; because of his illustrious career
and accomplishments; and because we wanted him to know that we remain very appreciative of him and in awe of his many accomplishments. We prepared this dedication and
shared it with him, via his family, in 2013, several months before he died, on February 7,
2014. He was 57.
In late June 2009, David had a fall at his home that left him unable to continue
his illustrious professional career. However, he is still being honored for his 30-plus years
of work in the physical therapy field, and his students and colleagues have continued his
important work and contributions. His loving family continued to remain close with him
and were dedicated to his welfare and happiness.
David’s curriculum vitae is long and very impressive. We put it, along with other information about him, at www.marksschwartzphd.com. His obituary is at http://hosting-24864.
tributes.com/obituary/show/david-e-krebs-99398993.
For those readers who did not know about David and his career, we include the following summary:
David E. Krebs received his BS in Physical Therapy in 1977 and MA in Applied
Physiology in 1979 from Columbia University, his PhD in Pathokinesiology and Physical
Therapy in 1986 from New York University, and his DPT in 2002 from the MGH Institute
of Health Professions.
In 1983, he agreed to write a major chapter for the first edition of Biofeedback: A Practitioner’s Guide (published in 1987). Although this was very early in his career, it was apparent that his accomplishments would be legendary. We have been fortunate and grateful
that he agreed to join the first edition and remained a contributor to this book throughout
his career.
v
He became Professor of Physical Therapy and Clinical Investigation at the MGH
Institute of Health Professions and Director of Massachusetts General Hospital’s Biomotion Laboratory. He also held academic appointments in Orthopaedics at Harvard Medical School and in Mechanical Engineering at the Massachusetts Institute of Technology.
David had more than 300 publications and was awarded more than $5 million as principal
investigator on federal (National Institutes of Health, National Institute on Disability and
Rehabilitation Research) and foundation research grants, primarily in the area of neural
and biomechanical constraints of human locomotor control.
He was the featured speaker at the 12th annual Eugene Michels Research Forum of
the American Physical Therapy Association and received the Association’s 2003 Helen J.
Hislop Award for Outstanding Contributions to Professional Literature, its 1998 Marian
Williams Award for Research in Physical Therapy, and its 1994 Golden Pen Award for
Scientific Writing.
We are very proud to continue to have David’s name and chapter, for which he was the
original author and major contributor, included in this book. Thank you, Dave, from the
many professionals and patients who forever are very respectful of and very grateful to you.
For the information about David E. Krebs’s career and other documents about his life,
we are indebted to his family and to Timothy Fagerson, DPT, his former student and coauthor. They all helped in the review and editing of this dedication.
vi
About the Editors
Mark S. Schwartz, PhD, is past chair of the Biofeedback Certification Institute of America (now Biofeedback Certification International Alliance; BCIA); serves on the Mayo
Clinic Emeritus Staff; and has a private practice in Jacksonville, ­Florida. He recently served
as Visiting Professor in the Department of Psychology at the University of North Florida.
Past president of the Association for Applied Psychophysiology and Biofeedback (AAPB),
Dr. Schwartz is board certified by the BCIA, a Diplomate in Clinical Sexology of the
American Board of Sexology, and a Diplomate of the American Board of Assessment Psychology. His website is www.marksschwartzphd.com.
Frank Andrasik, PhD, is Distinguished Professor and Chair of Psychology at the University of Memphis. He is Editor-in-Chief of Applied Psychophysiology and Biofeedback, Associate Editor of Cephalalgia, and past Editor-in-Chief of Behavior Therapy. His extensive
­publications include, most recently, the coedited volume Biopsychosocial Assessment in
Clinical Health Psychology. Dr. Andrasik is past president of the Association for Behavioral
and Cognitive Therapies and the AAPB. He is board certified as a Senior Fellow by the
BCIA and is a Fellow of the Society of Clinical Psychology and Society for Health Psychology (Divisions 12 and 38 of the American Psychological Association), the Association for
Psychological Science, the Society of Behavioral Medicine, the Association for Behavioral
and Cognitive Therapies, and the American Headache Society.
vii
Contributors
Frank Andrasik, PhD, Department of Psychology, University of Memphis, Memphis, Tennessee
John G. Arena, PhD, Mental Health Service Line, Department of Veterans Affairs Medical
Center, and Department of Psychiatry and Health Behavior, Medical College of Georgia,
Augusta, Georgia
Gerard A. Banez, PhD, Pediatric Pain Rehabilitation Program, Cleveland Clinic Children’s
Hospital for Rehabilitation, Cleveland, Ohio
Anat Barnea, DSc, private practice, Givat Chaim Ichud, Israel
Dana Bassett, MA, Hornsby Psychology Clinic, Sydney, New South Wales, Australia
Niels Birbaumer, PhD, Institute for Medical Psychology and Behavioural Neurobiology,
University of Tübingen, Tübingen, Germany
Keith I. Block, MD, Block Center for Integrative Cancer Treatment, Skokie, Illinois
Eugenia Bodenhamer-Davis, PhD, Department of Rehabilitation, Social Work and Addictions,
University of North Texas, Denton, Texas
Jeffrey E. Bolek, PhD, Emeritus Staff, Cleveland Clinic, and Motor Control Restoration, LLC,
Cleveland, Ohio
Rex L. Cannon, PhD, Neural Potential, West Palm Beach, Florida
Thomas F. Collura, PhD, BrainMaster Technologies, Inc., Bedford, Ohio
Timothy Culbert, MD, FAAP, Integrative Medicine, PrairieCare Medical Group,
Chaska, Minnesota
Richard E. Davis, MS, Neurotherapy Associates of Texas, Denton, Texas
Peter T. Dorsher, MSc, MD, Department of Physical Medicine and Rehabilitation, Mayo Clinic,
Jacksonville, Florida
Eugene Eisman, PhD, Emeritus Staff, Department of Psychology, University of Milwaukee,
Milwaukee, Wisconsin
Timothy L. Fagerson, DPT, Spine–Orthopaedic–Sports Physical Therapy,
Wellesley, Massachusetts
ix
x
Contributors
Herta Flor, PhD, Department of Cognitive and Clinical Neuroscience, Central Institute
of Mental Health, Medical Faculty Mannheim, University of Heidelberg, Mannheim, Germany
Udi Gal, MA, private practice, Manof, Israel
Richard N. Gevirtz, PhD, Department of Clinical Psychology, California School
of Professional Psychology, Alliant International University, San Diego, California
Alan G. Glaros, PhD, School of Dentistry, University of Missouri–Kansas City,
Kansas City, Missouri
Charlotte Gyllenhaal, PhD, Block Center for Integrative Cancer Treatment, Skokie, Illinois
Daniel Hamiel, PhD, Baruch Ivcher School of Psychology, IDC Herzliya, Herzliya, Israel;
Cohen–Harris Resilience Center, Haifa, Israel; and Tel-Aviv Brull Community Mental Health
Services, Tel-Aviv, Israel
Timothy Harkness, MA, Chelsea Football Club, London, United Kingdom
Joe Kamiya, PhD, Emeritus Staff, Department of Medical Psychology, University of California,
San Francisco, San Francisco, California
Maria Katsamanis, PsyD, Department of Psychiatry, Robert Wood Johnson Medical School,
Rutgers, The State University of New Jersey, Piscataway, New Jersey
Tobias Kaufmann, PhD, Institute of Clinical Medicine, University of Oslo, Oslo, Norway
Boris Kotchoubey, PhD, Institute for Medical Psychology and Behavioral Neurobiology,
University of Tübingen, Tübingen, Germany
David E. Krebs, DPT, PhD (deceased), Institute of Health Professions
and Biomotion Laboratory, Massachusetts General Hospital, Boston, Massachusetts
Andrea Kübler, PhD, Institute for Psychology, University of Würzburg, Würzburg, Germany
Deloris M. Lakia, DNP, CNP, CDE, College of Nursing, University of Toledo, Toledo, Ohio
Leonard L. Lausten, DDS, Department of Otorhinolaryngology, University of Kansas Medical
Center, Kansas City, Kansas
Paul M. Lehrer, PhD, Department of Psychiatry, Robert Wood Johnson Medical School, Rutgers,
The State University of New Jersey, Piscataway, New Jersey
Wolfgang Linden, PhD, Department of Psychology, University of British Columbia, Vancouver,
British Columbia, Canada
Joel F. Lubar, PhD, Emeritus Staff, Department of Psychology, University of Tennessee,
Knoxville, Tennessee and Southeastern Neurofeedback Institute, Pompano Beach, Florida
Angele V. McGrady PhD, LPCC, Department of Psychiatry, University of Toledo, Toledo, Ohio
Susan Middaugh, PhD, Emeritus Staff, Department of Anesthesiology and Perioperative
Medicine, Medical University of South Carolina, Charleston, South Carolina
Vincent J. Monastra, PhD, FPI Attention Disorders Clinic, Endicott, New York
Nicola Neumann, PhD, Institute of Diagnostic Radiology and Neuroradiology,
University of Greifswald, Greifswald, Germany
C. J. Peek, PhD, Department of Family Medicine and Community Health,
University of Minnesota Medical School, Minneapolis, Minnesota
Jeanetta C. Rains, PhD, Center for Sleep Evaluation, Elliot Hospital,
Manchester, New Hampshire
Contributors
Andrea Reid-Chung, MA, ADD Centre, Biofeedback Institute of Toronto, Mississauga,
Ontario, Canada
Arnon Rolnick, PhD, Rolnick’s Clinic, Ramat Gan, Israel
Ronald L. Rosenthal, PhD, private practice, Miami, Florida
Mark S. Schwartz, PhD, Emeritus Staff, Department of Psychiatry and Psychology, Mayo Clinic,
and private practice, Jacksonville, Florida
Nancy M. Schwartz, MA, Department of Psychology, University of North Florida,
Jacksonville, Florida
Keith Sedlacek, MD, private practice, Stress Regulation Institute, New York, New York
Fredric Shaffer, PhD, BCB, Center for Applied Psychophysiology, Truman State University,
Kirksville, Missouri
Richard A. Sherman, PhD, Department of Psychology, Saybrook University, Oakland, California
Wesley E. Sime, PhD, MPH, Emeritus Staff, Department of Nutrition and Health Sciences,
University of Nebraska–Lincoln, and First Step Wellness, Lincoln, Nebraska
Jonathan C. Smith, PhD, Department of Psychology, Roosevelt University, Chicago, Illinois
Estate M. “Tato” Sokhadze, PhD, Department of Biomedical Sciences, University of South
Carolina School of Medicine, Greenville, South Carolina
Ute Strehl, PhD, Institute for Medical Psychology and Behavioral Neurobiology,
University of Tübingen, Tübingen, Germany
Sebastian Striefel, PhD, Emeritus Staff, Department of Psychology, Utah State University,
Logan, Utah
Robert W. Thatcher, PhD, Applied Neuroscience Research Institute, Largo, Florida
James W. G. Thompson, PhD, Evoke Neuroscience, New York, New York
Lynda Thompson, PhD, ADD Centre, Biofeedback Institute of Toronto, Mississauga,
Ontario, Canada
Michael Thompson, MD, ADD Centre, Biofeedback Institute of Toronto, Mississauga,
Ontario, Canada
Kirtley E. Thornton, PhD, Neuroscience Center, Charlotte, North Carolina
Jeannette Tries, PhD, OTR, Aurora West Allis Memorial Center, West Allis, Wisconsin,
and Clinic for Neurophysiological Learning, Milwaukee, Wisconsin
David L. Trudeau, MD, Foundation Neurofeedback and Neuromodulation Research (FNNR,
formerly ISNR-RF), Murfreesboro, Tennessee
Frederick Wamboldt, MD, Department of Medicine, National Jewish Health, Denver, Colorado
Robert Whitehouse, EdD, private practice, Boulder, Colorado
Vietta Sue Wilson, PhD, School of Kinesiology and Health Science, York University,
Toronto, Ontario, Canada
Marcie Zinn, PhD, Department of Community Research, DePaul University, Chicago, Illinois
Mark Zinn, MM, Department of Community Research, DePaul University, Chicago, Illinois
xi
Preface
The responses to the first three editions of Biofeedback have been very rewarding in terms of acceptance and reviews. The second edition essentially doubled the size, and despite our sincere efforts
and the publisher’s urging, the third edition remained the same size. The publisher reemphasized
the need for a substantial reduction in the fourth edition, and we again earnestly sought to comply.
Because new research areas and interests in biofeedback have emerged in recent years, it was clear
that more topics warranted inclusion. We therefore deemed it necessary to expand the chapters
involving electroencephalographic (EEG) biofeedback/neurotherapy and other topics.
Readers will find 11 new chapters among the total of 42. Several of the chapters from the third
edition have been combined and condensed into single chapters. Seven chapters focus entirely on
EEG biofeedback/neurofeedback, and several other chapters include substantial discussions of this
modality. The remaining chapters have been revised, updated, and streamlined to varying degrees.
One chapter was eliminated.
Three instrumentation chapters are entirely new, focusing on surface electromyography (Chapter
4), quantitative EEG (Chapter 7), and consumer products (Chapter 9). Other totally new chapters focus on relaxation (Chapter 12), cognitive-behavioral therapy and ways to integrate it with
biofeedback (Chapter 11), and workplace applications (Chapter 36). Five wholly new chapters are
devoted to disorders receiving increased attention by biofeedback therapists and researchers: anxiety disorders (Chapter 26), asthma (Chapter 29), traumatic brain injury (Chapter 38), autism spectrum disorders (Chapter 39), and substance use disorders (Chapter 41). These new chapters provide
unique content that is not found within any other single volume.
We have retained chapters on the same conditions and disorders that were present in the prior
editions, because these remain the focus of continued research and practice—recurrent headaches,
temporomandibular disorders, Raynaud’s disease, essential hypertension, diabetes mellitus, tinnitus, fibromyalgia, irritable bowel syndrome, attention-deficit/hyperactivity disorder, neuromuscular reeducation, and bowel/bladder and pelvic floor disorders—as well as chapters on applications
for performing artists, sports, and pediatrics. All of these contributions have been significantly
updated. As in prior editions, we include a chapter on frontier applications that focuses on several
other disorders and conditions. The volume again begins with a historical perspective and now
includes the new definition of biofeedback and a preliminary definition of applied psychophysiology.
The chapter on entering the field and ensuring competence has been thoroughly updated.
Biofeedback now includes a total of 65 authors, nearly half of whom (32) are new contributors.
This compares with 39 authors in the third edition, 21 authors in the second edition, and 7 in
xiii
xiv
Preface
the first edition. To the extent that is possible, we and the other authors have striven to integrate
academic/research content along with considerable applied content of a clinical and/or educational
nature.
Interested readers may find it useful to refer to the breadth of materials relevant to biofeedback,
and the specific topics discussed in this book, that are accessible at www.marksschwartzphd.com.
This website allows for posting of supplemental photos, graphics, text, protocols, patient/client/
subject educational material, and examination items for qualified educators, as well as for reader
comments and questions, and updating of content by some authors where appropriate.
This book continues to provide the broadest scope of topics in the field of biofeedback and
applied psychophysiology, prepared by a diverse and highly acclaimed set of authors.
On the occasion of the publication of this fourth edition, we thank all readers of the previous
editions for their support, ideas, dedication, and feedback, and look forward to receiving continued
feedback from readers of this new edition.
Mark S. Schwartz
Frank Andrasik
Acknowledgments
Mark S. Schwartz: In the prior three editions, I expressed my detailed acknowledgments to many
people, and I still am appreciative of them for all the reasons expressed previously. (I am including
acknowledgments from all three earlier editions at www.marksschwartzphd.com.) For this edition, I
will add a few that were not mentioned in the prior editions.
I am incredibly grateful that Frank Andrasik again agreed to coedit the text. He diligently coped
with the myriad factors that resulted in delays from many sources. His incredible and varied skills,
his patience, his perseverance, his devotion, and his humor all were clearly crucial in the completion of the book.
Frank Andrasik: In the third edition (my first as coeditor), I expressed my heartfelt thanks to Mark
Schwartz, for affording me the privilege and opportunity to assist him in revising what had already
become a classic text in the field. Now, as then, I treasured every step along the way, especially
those enabling me to get to know him more as a person and a friend. It is my continued honor to
be able to work with and learn from Mark, whose deep humility will leave him to disagree with me
when I acknowledge him as one of the clinical pioneers and most influential leaders in our field.
He has contributed in innumerable ways to the growth of biofeedback and applied psychophysiology. For this and more, all of us are indebted. I also expressed my deep appreciation to the many
mentors and colleagues who helped educate me, guide me, and shape my career, and to family and
loved ones who helped sustain me over my entire career. (My prior acknowledgments also appear
at www.marksschwartzphd.com, as I continue to appreciate all those mentioned, even more, if that
is possible.)
Mark S. Schwartz and Frank Andrasik: Many thanks to The Guilford Press staff for continued
acceptance, support, patience, and flexibility. Special thanks to Jane Keislar, Senior Assistant Editor, and Jim Nageotte, Senior Editor, for their incredible patience, flexibility, and attention to detail
during the lengthy process of compiling a book of this magnitude. Thanks also to Carolyn Graham,
who managed many of the preproduction administrative matters for the book; to the wonderful
copy editor, Jacquelyn Coggin, for her skills and tact; and to Senior Production Editor Anna Nelson,
for her patience and flexibility in complet­ing the final editorial and production stages for the book.
We also express our considerable appreciation to our authors, who remained devoted, patient,
and persistent with the unexpected and unplanned delays along the way.
Finally, we remain somewhat at a loss for words to adequately thank our wives, Nancy and Candy,
for coping with our late nights and long weekends, while we sat at our keyboards when we should
have been sitting with them.
xv
Contents
Part I. Orientation to Biofeedback
1.The History and Definitions of Biofeedback and Applied Psychophysiology
3
Mark S. Schwartz, Thomas F. Collura, Joe Kamiya, and Nancy M. Schwartz
2.Entering the Field and Assuring Competence
24
Fredric Shaffer and Mark S. Schwartz
Part II. Instrumentation
3. A Primer of Traditional Biofeedback Instrumentation
35
C. J. Peek
4. Advanced Topics in Surface Electromyography: Instrumentation and Applications
68
Jeffrey E. Bolek, Ronald L. Rosenthal, and Richard A. Sherman
5.Cardiorespiratory Measurement and Assessment in Applied Psychophysiology
85
Richard N. Gevirtz, Mark S. Schwartz, and Paul M. Lehrer
6.Electroencephalographic Measures and Biofeedback: A Primer
98
Nicola Neumann, Ute Strehl, Niels Birbaumer, and Boris Kotchoubey
7. Quantitative Encephalography and Electroencephalographic
Biofeedback/Neurofeedback
Robert W. Thatcher
xvii
113
xviii
Contents
8. Introduction to Psychophysiological Assessment and Biofeedback Baselines
128
John G. Arena and Mark S. Schwartz
9. Consumer‑ and Home‑Based Biofeedback
154
Mark S. Schwartz and Frank Andrasik
PART III. ADJUNCTIVE/COMPLEMENTARY INTERVENTIONS
10. Dietary Considerations
163
Keith I. Block, Charlotte Gyllenhaal, and Mark S. Schwartz
11. Biofeedback and Cognitive‑Behavioral Interventions: Reciprocal Contributions
176
Daniel Hamiel and Arnon Rolnick
PART IV. RELAXATION INTERVENTIONS
12. Relaxation Today: Self‑Stressing and Psychological Relaxation Theory
189
Jonathan C. Smith
13. Cardiorespiratory Biofeedback
196
Richard N. Gevirtz, Paul M. Lehrer, and Mark S. Schwartz
PART V. PRACTICE ISSUES
14. Intake and Preparation for Intervention
217
Mark S. Schwartz
15. Adherence
233
Jeanetta C. Rains and Mark S. Schwartz
16. Problems Associated with Relaxation Procedures and Biofeedback,
and Guidelines for Management
249
Mark S. Schwartz, Nancy M. Schwartz, and Vincent J. Monastra
17. Ethical Practice Issues and Concerns
260
Sebastian Striefel
18. Myths, Insurance, and Other Professional Topics and Issues
272
Sebastian Striefel, Ronald L. Rosenthal, Robert Whitehouse, and Mark S. Schwartz
19. Evaluating Research in Clinical Biofeedback
Frank Andrasik and Mark S. Schwartz
290
xix
Contents
Part VI. Clinical Applications:
Traditional General Biofeedback Practice
20. Headache
305
Frank Andrasik and Mark S. Schwartz
21. Temporomandibular Muscle and Joint Disorders
356
Alan G. Glaros and Leonard L. Lausten
22. Raynaud’s Disease and Raynaud’s Phenomenon
369
Maria Katsamanis, Mark S. Schwartz, and Keith Sedlacek
23. Essential Hypertension
383
Wolfgang Linden and Angele V. McGrady
24. Diabetes Mellitus
400
Angele V. McGrady and Deloris M. Lakia
25. Tinnitus: Nothing Is as Loud as a Sound You Are Trying Not to Hear
422
Herta Flor and Mark S. Schwartz
26. Anxiety Disorders
439
Arnon Rolnick, Dana Bassett, Udi Gal, and Anat Barnea
27. Fibromyalgia Syndrome
453
Peter T. Dorsher and Mark S. Schwartz
28. Irritable Bowel Syndrome
466
Mark S. Schwartz
29. Asthma
480
Paul M. Lehrer and Frederick Wamboldt
Part VII. Clinical Applications: Electroencephalographic,
Neuromuscular, and Pelvic Floor Biofeedback Specialties
30. Attention‑Deficit/Hyperactivity Disorder
493
Vincent J. Monastra and Joel F. Lubar
31. Neuromuscular Reeducation and Gait Training
525
Timothy L. Fagerson and David E. Krebs
32. Bowel, Bladder, and Pelvic Floor Disorders
Jeannette Tries and Eugene Eisman
545
xx
Contents
Part VIII. Optimization
33. Performing Artists
587
Marcie Zinn and Mark Zinn
34.Sports
607
Vietta Sue Wilson, Wesley E. Sime, and Timothy Harkness
Part IX. Other Special Populations and Applications
35. Pediatric Applications
629
Timothy Culbert and Gerard A. Banez
36. Work‑Related Musculoskeletal Disorders
651
Susan Middaugh
Part X. The Frontier and a Nod to the Future
37. Biofeedback of Slow Cortical Potentials in Epilepsy
671
Ute Strehl
38.Traumatic Brain Injury, Quantitative Electroencephalography,
and Electroencephalographic Biofeedback
677
Kirtley E. Thornton
39. Biofeedback Interventions for Autism Spectrum Disorders
686
Lynda Thompson, Michael Thompson, James W. G. Thompson, and Andrea Reid‑Chung
40. Brain–Computer Communication: An Alternative Communication Channel
for Paralyzed Patients
697
Tobias Kaufmann, Niels Birbaumer, and Andrea Kübler
41.Substance Use Disorders and Neurofeedback
707
Estate M. “Tato” Sokhadze, David L. Trudeau, Rex L. Cannon,
Eugenia Bodenhamer‑Davis, and Richard E. Davis
42. More Frontiers and Further Forward
717
Mark S. Schwartz and Frank Andrasik
Index
739
Part I
Orientation to Biofeedback
Chapter 1
The History and Definitions of Biofeedback
and Applied Psychophysiology
Mark S. Schwartz, Thomas F. Collura,
Joe Kamiya, and Nancy M. Schwartz
This chapter conveys the converging trends
that influenced the development and journey
of applied biofeedback, and the broader field of
applied psychophysiology.1 This historical perspective is designed to help the reader understand the
origins of the multifaceted and multimodality field
of biofeedback, including a history of specialty
modalities and applications (e.g., electroencephalographic [EEG] biofeedback, a.k.a. neurofeedback). It also seeks to help illuminate the broader
concept of applied psychophysiology, and to give
perspective to the name changes of the primary
professional membership organization and its journal.
Applied biofeedback began in the United States
with the convergence of many disciplines in the
late 1950s. The major antecedents and fields from
which it developed include the following. (Italics
on the first use of a term indicate that the term is
included in the Glossary.)
7. Consciousness, altered states of conscious-
ness, and electroencephalography (EEG biofeedback also known as neurofeedback)
8. Cybernetics
9. Cultural factors.
10. Professional developments2
11. Definitions
The order of the items in this list reflects neither historical sequence nor relative importance.
Other historical perspectives on biofeedback may
be found in Basmajian (1989), Shaffer (2010),
and Peper and Shaffer (2010). (See www.marksschwartzphd.com for other references and links to
selected historical perspectives.)
Instrumental Conditioning
of ANS Responses
Learning theory developed within experimental
psychology. Reinforcement is necessary for operant conditioning or instrumental conditioning to
occur. From this perspective, both overt behaviors
and covert behaviors, such as thoughts, feelings,
and physiological responses, are functions of the
antecedents and consequences of such behaviors.
This model describes the learning of responses
instrumental to obtaining positive or avoiding
negative consequences.
1. Instrumental conditioning of autonomic nervous
system (ANS) responses
2. Psychophysiology
3. Behavior therapy and behavioral medicine
4. Stress research and stress management strate-
gies
5. Biomedical engineering
6. Surface electromyography (EMG), diagnostic
EMG, and control of single motor units
3
4
The prevailing scientific viewpoint for several
decades has been that only the voluntary musculoskeletal system, mediated by the central nervous
system (CNS), is responsive to operant conditioning. The older view held that the ANS functioned
automatically beyond conscious awareness, and
hence beyond voluntary control. Most scientists
thought that the internal, homeostatic controls
for functions such as circulation and digestion
were innate and unaffected by self-regulatory
learning. Most scientists assumed that ANS functioning or visceral learning was modifiable only via
classical conditioning, if subject to learning at all. In
this view, responses are automatic after conditioning occurs. In classical conditioning, thoughts can
even become conditioned stimuli (CSs) and elicit
physiological responses.
The strong biases against instrumental conditioning of the ANS and the visceral responses it
controls limited the amount of experimental work
in this area until a few decades ago (Miller, 1978).
Studies with humans and animals showed that
instrumental training could produce increases and
decreases in bodily responses (see early reviews
by Harris & Brady, 1974; and Kimmel, 1979, and
Taub, 2010).
Research indicated that individuals could gain
volitional control over several different ANS functions without learning that could be attributed to
cognitive factors. Many scientists and professionals were very skeptical of these findings. There
was much disagreement concerning whether the
research really demonstrated cortical control over
ANS activity. As research advanced, it became
clear that to show operant learning effects in
the ANS, researchers needed more sophisticated
designs. They had to rule out skeletally mediated
mechanical artifacts and visceral reflexes.
The best organized and most articulate history
of the very challenging research on instrumental
conditioning of autonomic response systems, and
in particular the brilliance of Neal Miller, was
provided by Ed Taub (2010). Every student of the
history of biofeedback, the history of psychology,
and indeed research methods, should read this
presentation by Taub. (With permission, the entire
article is reproduced at www.marksschwartzphd.
com for readers who are interested.)
• Eliminating or ruling out somatic mediation of
the autonomic responses was the problem to be
resolved.
• Studies of heart rate changes with chemically
paralyzing and artificially respirated rats, thus
I. ORIENTATION TO BIOFEEDBACK
without somatic mediation, were reported
(Miller & DiCara, 1967; DiCara & Miller
1968a, 1968b, 1968c, 1968d).
• Instrumental condtioning of autonomic functioning was controversial among psychophysiological researchers.
• Several attempts by other researchers and by
Miller and his students to replicate these studies were unsuccessful. Most researchers who
were familiar with this research attributed the
DiCara and Miller (1967a) results to an anomaly despite other, similar and reliable, albeit
smaller, results by other investigators (Trowill,
1967; Hothersall & Brener, 1969; Slaughter et
al., 1970).
• Miller and his students made extensive and
meticulous efforts to reproduce the studies and
although unsuccessful, the process provided an
outstanding example of the Strong Inference
(Dworkin & Miller, 1986; Taub, 2010) research
model. They evaluated a large number of alternative hypotheses.
• Adverse publicity about the unsuccessful replications created a strong negative association.
• Taub (2010) pointed out the terrifying limitations of any attempted learning research with
paralyzed rats or any vertebrate.
• The second experimental question regarding
biofeedback instrumental conditioning of any
ANS response without somatic mediation was
reported by Miller and Brucker (1979) with
patients with quadriplegia, thus without sufficient somatic muscle activity mediating the
strong increase in blood pressure enough to
manage the low blood pressure (i.e., orthostatic) due to the patients typical reclining position. Miller and Brucker noted that the results
were “strongly indicating that these patients
can learn unusually large increases in blood
pressure and that this visceral response can be
performed independently of skeletal responses”
(Taub, 2010, p. 113).
• Taub’s thermal biofeedback studies (Slattery &
Taub, 1976; Taub & School, 1978) extended the
research regarding instrumental conditioning
of ANS without somatic mediation. Temperature biofeedback from varied and specific locations on a hand resulted in “very clear anatomical differentiation of the temperature response”
and “a large response around the feedback
locus, and much less or none at other locations”
(Taub, 2010, p. 113).
• Taub (2010) reported his research with 11 participants attempting to alter skin temperature
5
1. History and Definitions
up or down on one digit compared to another.
He reported that for eight subjects, there was
“significantly greater temperature response on
the designated digit than at the other one”
(Taub, 2010, p. 114). With other controls, they
concluded that “the anatomical specificity
results represented differential alterations in
blood flow and were not due to an artifact . . .
[and] in particular, not to somatic mediation
involving muscle activity changes from any of
the locations . . . recorded” (Taub, 2010, p. 114).
The research with instrumental conditioning of
visceral responses mediated by the ANS provided
a major impetus to the development of clinical
biofeedback. It appeared to resolve the controversy
over whether such conditioning was a legitimate
phenomenon. An assumption of clinical biofeedback is that it can help persons improve the accuracy of their perceptions of their visceral events.
These perceptions allow them to gain greater selfregulation of these processes.
This operant model of biofeedback has significant heuristic value. One can apply principles of
instrumental conditioning to physiological selfregulation.
Although it is helpful to view biofeedback primarily as instrumental conditioning of visceral
responses, this model is limiting in that some professionals believe that human learning includes
major cognitive dimensions, as well as environmental reinforcers, for example, thinking, expectation, visualization and imagery, foresight and
planning, and problem-solving strategies.
One can include cognitive factors within the
operant conditioning model. However, professionals adhering to more stringent interpretations
of the model consider cognitive factors inadmissible, because one cannot observe or objectively
measure them. Nevertheless, studies of motor skill
learning (Blumenthal, 1977) show that humans
develop mental models (“motor programs”) of
what a skilled movement should be like. Furthermore, research shows that one may acquire behavior without obvious practice or even reinforcement. This evidence comes from latent learning
experiments (Harlow & Harlow, 1962), studies
of discovery learning (Bruner, 1966), and studies
of observational learning involving imitation of a
model (Rosenthal & Zimmerman, 1978).
Increased acceptance for the role of mental
processes in learning led to cognitive-behavioral
therapies and studies of cognitively mediated strategies in the changes occurring during biofeedback
therapies. The emphasis on cognitive learning
also supported the applications of cybernetics to
biofeedback.
Psychophysiology
David Shapiro offered the first academic course
in psychophysiology at Harvard University in 1965.
The Handbook of Psychophysiology, a major publication, appeared 7 years later (Greenfield &
Sternback, 1972).
Psychophysiology involves the scientific study
of the interrelationships of physiological and cognitive processes. Some consider it a special branch
of physiology. Others also consider it an offspring
of psychobiology, which in turn is the child of the
marriage between the physical and social sciences
(Hassett, 1978). Physiological psychologists often
manipulate physiology and observe behavior. In
contrast, psychophysiologists often facilitate, manage, guide, hinder, or obstruct human psychological variables and observe the physiological effects.
As a form of “applied psychophysiology,”3 clinical biofeedback helps people alter their behaviors
with feedback from their physiology. Some providers of clinical biofeedback used to refer to themselves as “clinical psychophysiologists.”
Behavior Therapy and Behavioral Medicine
The fields of behavior therapy and behavioral
medicine are related outgrowths of both learning
theory and psychophysiology. “Behavior therapy”
developed in the 1950s as an alternative to insightoriented psychodynamic theories and therapies for
mental disorders. The roots of behavior therapy
include the notion that one learns maladaptive behaviors; therefore, in most cases, one can
unlearn them. The model is largely educational
rather than medical. It applies the principles of
operant and respondent conditioning, as well as
of cognitive learning, to change a wide range of
behaviors. Many professionals view some biofeedback applications as a form of operant learning.
Others view biofeedback more cognitively within
an information-processing model.
“Behavioral medicine” is another outgrowth of
learning theory, psychophysiology, and behavior
therapy. This specialty developed within behavior
therapy and psychosomatic medicine. It appeared
as a distinct entity in the late 1970s. Behavioral
medicine focuses on applications of learning theo-
6
I. ORIENTATION TO BIOFEEDBACK
ries to medical disorders and other health-related
topics. It does not focus on psychopathology or
mental disorders. G. E. Schwartz and Weiss (1978)
reported a definition of behavioral medicine proposed at the Yale Conference held in 1977:
Behavior medicine is the field concerned with the
development of behavior science knowledge and
techniques relevant to the understanding of physical
health and illness and the application of this knowledge and these techniques to diagnosis, prevention,
treatment, and rehabilitation. Psychosis, neurosis,
and substance abuse are included only insofar as
they contribute to physical disorders as an end point.
(p. 379)
Behavioral medicine also developed because
traditional medical approaches were insufficient
for managing and treating many chronic diseases,
conditions, and health-damaging or maladaptive
behaviors. This new specialty goes beyond the
traditional germ theory of the etiology and progression of diseases. It recognizes the important
roles of stress, lifestyle, habits, and environmental
variables in the development, maintenance, and
treatment of medical and dental diseases and conditions.4
Behavioral medicine places much emphasis on
the patient’s role in prevention of and recovery
from organic diseases and conditions. The same
emphases are clear in applied or clinical biofeedback. In fact, some professionals consider clinical
biofeedback to be a major specialty within the
broader field of behavioral medicine (Birk, 1973;
Olton & Noonberg, 1980).
The contributions of behavior therapy and
behavioral medicine to the development and
applications of applied biofeedback and applied
psychophysiology are clear. The interactions
among professionals from all of these fields will
continue to be enriching.
Stress Research, Relaxation Therapies,
and Other Stress Management Techniques
An important area of behavioral medicine is
research on the effects of stress on causing physical symptoms and altering the immune system.
However, research on stress began long before the
development of behavioral medicine or biofeedback; in fact, both fields have their roots partly in
stress research. One has only to remember Hans
Selye’s (1974) report of more than 130,000 entries
on stress that showed the extent of this already
immense body of research.
Pioneering research was conducted by the physicians Claude Bernard and Walter B. Cannon,
as well as by Selye. Pi Suñer (1955) observed that
Bernard developed the concepts of physiological
“homeostasis” as the major process by which the
body maintains itself. As Langley (1965) noted, the
concept became integral to the discipline of physiology. Physical and mental disease are thought to
occur because some homeostatic feedback mechanism is malfunctioning. One of the major effects
of such homeostatic imbalance is stress.
In his book The Wisdom of the Body, Cannon
(1932) indicated the natural causes and results of
the innate stress response. He named this response
fight or flight. Selye’s (1974, 1976, 1983) extensive
research led to a triphasic conceptualization of
the nature of the physiological stress response: the
stages of alarm, resistance, and exhaustion.
The brilliant and pioneering work of Cannon
and Selye contributed significantly to the development of the field of psychosomatic medicine.
Their work increased awareness of the role of
stress in physical and mental diseases. This awareness nurtured applied biofeedback, and many of
these applications focused on stress-related disorders. Furthermore, as noted by Miller (1978), the
emphasis of biofeedback on measuring and producing changes in bodily processes contributes
to other behavioral techniques for relieving stress
effects.
Many stress management systems evolved with
the awareness of the effects of stress on health and
disease. Included among these are many relaxation
therapies, and some observers perceive biofeedback
as a specific treatment modality within this group.
In practice, the effects of relaxation have a major
role in achieving the therapeutic effects with some
forms of biofeedback. A very early form of physical
relaxation is “hatha yoga,” adopted from the Far
East and popularized in Western countries in the
1960s. In the United States in the 1930s, Edmund
Jacobson (1938, 1978) developed “progressive
relaxation” (PR), sometimes also called “progressive muscle relaxation,” which consists of a series
of muscle activities designed to teach people ways
to distinguish degrees of tension and relaxation,
and to reduce specific and general muscle tension. It also helps reduce or stop many symptoms
and some causes and adverse effects of stress.
McGuigan and Lehrer (2007), as two of Jacobson’s
students and ardent authorities, discussed the history and techniques from their unique perspec-
7
1. History and Definitions
tive. Lichstein (1988) provided one of the most
thorough detailed texts on relaxation strategies
and research results. Other very useful resources
are the books by Smith (1989, 1990, 2001, 2005;
also see Chapter 12 in this volume). Modifications
of progressive relaxation have been developed by
Wolpe (1973), Bernstein and Borkovec (1973),
Bernstein, Carlson, and Schmidt (2007), and
Jacobson and McGuigan (1982). A related technique developed in England by Mitchell (1977,
1987) involves stretch–release procedures. In addition to the physiological relaxation procedures,
there has been a proliferation of primarily mental
techniques, most of which involve some form of
meditation. Islamic Sufis, Hindu yogis, Christian
contemplatives, and Hasidic Jews have practiced
religious meditation for centuries.
Meditation became popularized in the United
States in the 1960s as a result of the development
of Transcendental Meditation (TM), practiced and
promoted by a teacher from India named Maharishi Mahesh Yogi (Forem, 1974). More Westernized
variations of TM were subsequently developed as
“clinically standardized meditation” (Carrington,
1977, 1978, 1998, 2007) and the “relaxation
response” (Benson, 1975). Stroebel’s (1982) “quieting reflex” is a modification of a meditation technique combined with physiological relaxation.
Another meditation approach is “open focus,”
developed by Fehmi and Fritz (1980), which has
recently experiencined a contemporary updating
(Fehmi & Robbins, 2009). It is closer to Soto Zen
meditation in its goal of seeking a content-free and
quiet mind, by contrast with the focused concentration of yoga and TM. The emigration of Zen
Buddhist teachers to the United States beginning
in the 1940s was yet another factor contributing to
the meditation movement. See Carrington (2007)
and Kristeller (2007) for more history of modern
forms of mantra meditation and for mindfulness
meditation, respectively.
There are still other approaches involving relaxation/meditation: Ira Progoff’s (1980) “process
meditation,” José Silva’s (1977) “Silva mind control,” and C. Norman Shealy’s (1977) “biogenics.”
Practitioners often combine relaxation/meditation
techniques with biofeedback instrumentation to
enhance the learning of psychophysiological selfregulation.
Hypnosis is yet another approach developed to
help persons to control pain and stress. Hypnosis
developed slowly from the 1700s until the 20th
century. Over the past few decades it has become
more sophisticated and empirically grounded as
a set of therapeutic techniques. Liebeault, Charcot, and Freud were among the first to apply the
techniques to patients (Moss, 1965). Twentiethcentury researchers, such as Hull, Barber, Hilgard,
Weitzenhoffer, and Erickson, conducted rigorous
investigations into the parameters of hypnosis.
Some, like Wickramasekera (1976, 1988), reported
integrations of hypnosis and biofeedback.
In Germany, early in the 20th century, J.
H. Schultz developed a form of physiologically
directed, self-generated therapy called “autogenic
training” or “autogenic therapy.” Wolfgang Luthe
(1969) brought it to North America and reported
extensive research and therapeutic applications of
this technique, variations of which are now also in
common practice.
Biomedical Engineering
Without high-quality instrumentation for measuring physiological events accurately and reliably,
there would be no biofeedback. As Tarlar-Benlolo
(1978) reminds us, “prior to World War II, available equipment was not sufficiently sensitive for
measuring most of the body’s internally generated
electrical impulses” (p. 728). Progress occurred
after the war.
Biomedical engineers have developed technology that is both noninvasive and very sophisticated. Surface recordings used for biofeedback
measurement provide feedback for many different
physiological activities. Feedback can also be provided for angles of limbs and the force of muscles
and limbs. Instruments continuously monitor,
amplify, and transform electronic and electromechanical signals into audio and visual feedback—
understandable information.
Now multiple and simultaneous recordings of
several channels of physiological information are
available with instrumentation linked to computers. Computers allow great storage capabilities,
rapid signal processing and statistical analyses,
simultaneous recording and integration of multiple channels, and displays that only a few years
ago were impossible.
EMG, Diagnostic EMG,
and Single‑Motor‑Unit Control
The workhorse of the biofeedback field has long
been surface electromyography (abbreviated here as
EMG, though SEMG is also used). According to
8
Basmajian (1983), EMG instrumentation grew out
of the studies of neuromuscular and spinal cord
functions. He reminds us that “it began with the
classic paper in 1929 by Adrian and Bronk, who
showed that the electrical responses in individual
muscles provided an accurate reflection of the
actual functional activity of the muscles” (p. 2).
Physicians have used EMG for diagnosing neuromuscular disorders for many decades. As early as
1934, reports indicated that voluntary, conscious
control over the EMG potential of single motor
units was possible (Smith, 1934). Marinacci and
Horande (1960) added case reports of the potential value of displaying EMG signals to assist
patients in neuromuscular reeducation. Basmajian
(1963, 1978) also reported on the successful control of single motor units.
Several investigators reported EMG feedback in the rehabilitation of patients after stroke
(Andrews, 1964; Binder-MacLeod, 1983; Brudny,
1982; Basmajian, Kukulka, Narayan, & Takebe,
1975; Wolf & Binder-MacLeod, 1983). Such
research was important in the development of
applied biofeedback, especially for the field of neuromuscular rehabilitation. Thus, EMG biofeedback gained solid support among researchers and
clinicians.
Practitioners have also used EMG feedback for
treating symptoms and disorders such as tension
headaches and tension myalgias, temporomandibular disorders, pelvic floor disorders that include
incontinence, and many other conditions (see
Part VI, this volume).
Consciousness, Altered States
Of Consciousness, and EEG Feedback
Humanistic psychology reestablished the human
self as a legitimate source of inquiry, and scientists
in transpersonal psychology and neurophysiology renewed the study of human consciousness.
Theorists such as Tart (1969), Krippner (1972),
Ornstein (1972), Pelletier and Garfield (1976),
G. E. Schwartz and Beatty (1977), and Jacobson
(1982) are among those who have made significant
contributions to our understanding of human consciousness.
Many studies of altered states of consciousness induced by drugs, hypnosis, or meditation
have added to our knowledge of the relationships
between brain functioning and human behavior.
Such research helped stimulate the use of electro-
I. ORIENTATION TO BIOFEEDBACK
encephalography (EEG) in biofeedback, which also
focuses on the functional relationships between
brain and behavior.
In the early 1960s, studies began appearing on
the relationships between EEG alpha wave activity
(8–12 hertz [Hz]) on the one hand, and emotional
states and certain states of consciousness on the
other. Alpha biofeedback, commonly reported as
associated with a relaxed but alert state, received
its most attention in the late 1960s. Clinical applications were mostly for general relaxation.
Kamiya (1969) reported that one could voluntarily control alpha waves—previously believed
impossible Support came from Brown (1977), Nowlis and Kamiya (1970), and Hart (1968). “Though
these studies tended to lack systematic controls,
they nonetheless caught the imagination of many
serious scientists as well as the media” (Orne, 1979,
p. 493). Some investigators and practitioners continued to advocate the value of alpha biofeedback
through the early 1980s (e.g., see Gaarder & Montgomery, 1981, for a discussion), despite recognizing
that “there was no clear-cut and concrete rationale
to explain why it should help patients” (p. 155). In
contrast, Basmajian (1983) noted that “alpha feedback . . . has virtually dried up as a scientifically
defensible clinical tool. . . . It has . . . returned
to the research laboratory from which it probably
should not have emerged prematurely. Through
the next generation of scientific investigation, it
may return as a useful applied technique” (p. 3).
Other investigators studied specialized learning
processes and other EEG parameters, such as theta
waves, evoked cortical responses, and EEG phase
synchrony of multiple areas of the cortex (Beatty,
Greenberg, Deibler, & O’Hanlon, 1974). Selected
brain areas and EEG parameters (e.g., sensorimotor rhythm and slow-wave activity) became the
focus of well-controlled studies. These emerged
as effective therapeutic approaches for carefully
selected patients with CNS disorders such as
epilepsy (Lubar, 1982, 1983; Sterman, 1982; see
Strehl, Chapter 37, this volume), as well as for
some patients with attention-deficit/hyperactivity
disorder (Lubar, 1991; see also Monastra & Lubar,
Chapter 30, this volume).
More recently, EEG feedback procedures purport to be successful in treating patients with a
wide variety of other symptoms and disorders. The
growth and scope of EEG biofeedback is partially
reflected in the changes in this text now, with
eight chapters compared to two in the third edition and one in the first two editions.
1. History and Definitions
History and Development of EEG Biofeedback
Technology
EEG biofeedback, sometimes referred to as “neurofeedback,” began with the approach of enhancing
a particular frequency band, generally alpha, as a
means of achieving benefits associated with greater
presence of that band in the EEG. When initial
work began, some systems used conventional EEG
systems, and augmented them with additional
circuitry. Others were developed entirely “standalone,” with amplifiers, processing circuits, and
output devices (lights, speakers, etc.) as an integral part of the design. Early research used such
custom-engineered systems to produce important
initial results (e.g., Nowlis & Kamiya, 1970). As
the field began to mature, manufacturers began to
introduce products capable of measuring and feeding back EEG signals as their primary purpose.
A major limitation of early EEG feedback
devices was that they filtered the desired band
to indicate its presence but had no provision to
ensure that out-of-band signals did not also contribute the feedback. For example, low-frequency
signals due to theta waves, eye movement, motion
artifact, or other non-alpha phenomena, if sufficiently large, could still produce enough output
to trigger the reward. Similarly, high-frequency
signals, including EMG and other artifacts, could
also produce output within the desired training
band, again imprecisely rewarding the trainee. As
a result, early alpha trainers produced inconsistent results that contributed to a general lack of
acceptance as useful professional tools. Early “recreational” alpha trainers, circa 1975, were primitive and not only trained alpha but also rewarded
various artifacts such as muscle twitches and eye
movements. (Interested readers can find photos of
an early recreational alpha trainer, circa 1975, and
an Autogenics 120 analog EEG trainer at www.
marksschwartzphd.com).
During this time, professional biofeedback
trainers were also being developed and applied.
These remained entirely analog, and provided
a display meter, and generally simple tones. A
great deal of research was conducted using these
devices, so that by 1978, dozens of studies including EEG biofeedback had appeared in the literature (Butler, 1978).
As EEG feedback equipment became more
refined, and in particular when digital computers
began to be used, it became possible to introduce
“inhibit” bands, which were used to block feed-
9
back from occurring when these signals were present. The ability to withhold feedback when excessive slow or fast waves were present was a key step
in the refinement of EEG biofeedback, and made
it possible to produce useful and consistent clinical results. The use of these outer “guard” bands
became very common and produced a generation
of feedback trainers that accurately rewarded the
desired EEG frequency components, without providing false feedback due to artifacts (Ayers, Sams,
Sterman, & Lubar, 2000).
Early PC-based EEG biofeedback systems were
implemented on platforms such as the Apple and
IBM PC, and used simple “text-based” operating systems. Supplementary graphics and sound
were generally very simple, yet effective. With the
introduction of Windows and the Apple Macintosh, software became increasingly sophisticated.
It became possible for programmers to incorporate
advanced signal processing, graphics, video, multimedia, interactive games, and other capabilities,
further enriching the feedback and improving
responses.
Computers were used for delivering EEG biofeedback as early as the late 1970s. However, the
processing speed was insufficient to keep up with
the signal processing demands, which limited
their utility It was not until the second and third
generations of processors, when math coprocessors
became available in the mid-1980s, that it was possible, for example, to perform a 256-point fast Fourier transform (FFT) in substantially less than a
second. In the 1990s, computer speed became fast
enough to provide real-time signal processing and
adequate displays for useful training.
During the evolution of these techniques, certain aspects became paramount and hotly debated.
Among these was “response time,” which was
construed to mean “the delay between the time
something happens in the brain and the time it
appears on the screen.” While apparently straightforward, this definition cannot really be applied
as such. In the world of real-time signals, digital
filters, frequency transforms, and such, signals do
not simply “come and go.” Rather, they wax and
wane, have varying amplitudes and time courses,
and responses are analog, or graded, not simply
on or off. As a result, it is necessary to consider
filter response and other factors in evaluating system capabilities. In an early Windows-based EEG
biofeedback system (Ayers) there was one enhance
band (labeled “Facilitate”) and one inhibit band
(see examples at www.marksschwartzphd.com).
10
The response curve shown demonstrates that
in order to respond to the narrow bandwidth of
the 15.0–18.0 Hz range, the filter requires several
cycles of the input wave in order to respond. This
response characteristic is identical to that seen
with analog filters. In other words, while digital
processing provides benefits in the form of programmability, flexibility, and exotic displays, it
cannot violate the basic laws of physics, and the
ability to respond to EEG waves is in principle the
same in digital as it is in analog systems.
Currently, most EEG biofeedback systems are
PC-based. Thus, the hardware typically consists
of an amplifier/digitizer (“encoder”) that transmits EEG data to the PC in a digital form. From
then on, the system depends entirely on the PC
software, which can consist of thousands of lines
of software code, developed over many programmer years of effort. EEG biofeedback has thus followed the trend of many other industries that have
become dominated by software issues, and follow
an aggressive and rapid evolution spurred on by
continuous competition and the continual entry
of new developers.
With the flexibility of computerized EEG biofeedback (Collura, 1995), seemingly rigid rules
have been stretched and even broken. For example, with the introduction of multiple frequency
bands for analysis and the ability to either reinforce or to inhibit any of them, a variety of creative
protocols emerged.
Based on principles of learning theory, different
methods of adjusting thresholds exploit different
aspects of the nervous system having to do with
perceived rewards, motivation, level of difficulty,
and so on. Where computerized EEG biofeedback
systems have excelled is in the use of complex
rules to compute and deliver feedback, and the
control of engaging and meaningful displays such
as animation, video, games, and various types of
specially designed software. The use of computerized signal processing has also allowed the introduction of a plethora of alternative methods and
approaches, embodying physiological and mathematical concepts including nonlinear systems,
chaos, coherence and stability, synchrony, selfadaptive systems, and normative databases.
Some EEG biofeedback systems appeal to concepts generally derived from “quantum physics,”
“subtle energy,” and other seemingly esoteric areas.
While these include systems that are well studied
and published, there are others that appeal more
to an article of faith than to peer-reviewed studies.
I. ORIENTATION TO BIOFEEDBACK
One element that has carried forward from initial systems, and continues to be in contention, is
the issue of “monopolar” versus “bipolar” recording. Depending on whether the EEG is referenced
to a neutral site or to another active site, the type
of information available is profoundly different,
and impacts the ability to train synchrony, connectivity, and other brain properties (Collura,
2009; Fehmi & Collura, 2007).
Early Investigations Leading to Neurofeedback
Our aims in this section are to (1) describe some
of the early work that led to the field of research
and therapeutic application of EEG biofeedback/
neurofeedback and (2) address the question of
what internal behaviors or private experiences are
involved in learning to produce changes in specific EEG measures with the aid of neurofeedback.
My (Kamiya) research with the EEG was
conducted to pursue questions concerning relationships between the EEG of persons and their
consciousness. This interest in EEG research
developed when I was working in the sleep laboratory of Nathaniel Kleitman and his student and
research assistant William Dement, at the University of Chicago. It was from that laboratory that
Aserinsky and Kleitman (1953) and Dement and
Kleitman (1957) published the pioneering papers
that indicated dreaming during sleep usually was
accompanied by specific changes in the sleeping
person’s EEG and eye movements as monitored
by the electrooculogram (EOG). Their papers did
much to put private experience on the scientific
map. Kleitman generously offered me the use of
his laboratory to conduct some studies of my own.
Dement taught me the technology of EEG and
EOG recording of sleeping subjects. I completed
a study on other physiological concomitants of
drowsiness and sleep (Kamiya, 1961). My student
and colleague, Johann Stoyva, joined me in the
laboratory. In addition, in response to confusion
and disagreements in the field on the problem of
how to interpret the occasional fact that reports
of dreaming would occur despite the absence of
their EEG and EOG indicators, and the absence
of dream reports when the EEG and EOG indicate
dreaming had occurred, we published an analysis
of the logic of the relations between verbal reports
and physiological indicators as convergent indicators of private events such as dreaming (Stoyva
& Kamiya, 1968). This problem is worth mentioning here because it arises in connection with
1. History and Definitions
the validity of the evidence of any sort of private
experience, not just dreams. In the course of preparing a subject for all night recording with EEG
and EOG electrodes, I always conducted a test of
the EEG on the polygraph for a minute while the
subject was still awake to make sure that all the
electrode contacts with the scalp provided clean
traces. It was during these tests that I noticed the
irregularly timed appearance and disappearance of
the EEG alpha rhythms. I wondered if they were
related to any features of the consciousness of
the person. How this interest led to the development of methods for studying that possibility, and
eventually to the adoption of the method by others in the treatment of neurological disorders, is
described in what follows.
There are wide variations in the characteristics
of the trains of alpha rhythms. The question that
motivated me was whether there are subjective
concomitants associated with the moments when
alpha rhythms are present as opposed to when
they are absent. Might there be a difference in
the feel or mental activity between the two EEG
states in the relatively short-term alternations
between the two that occur several times a minute? Considering what is known about conditions
affecting alpha would help provide hints toward
an answer. (For more on alpha rhythms, see www.
marksschwartzphd.com).
Later, when I joined the Department of Psychiatry at the University of California at San Francisco (UCSF) in 1961 and moved my laboratory
equipment there, I added an improvement over
the on-or-off character of the feedback in earlier
studies. A continuously graded tone volume from
silent to loud now reflected the 1-second moving
average amplitude of alpha, thus improving the
information in the signal. The participant now
could monitor his or her performance more accurately. We also changed the score presented every
minute from total time of alpha above threshold
to the average amplitude of alpha for the minute.
With these improvements in the feedback,
the performance of the trainees improved, and
interest level was maintained. I believe the several reported failures by other investigators to
replicate the results we had obtained showing
increases in the average trainee of alpha relative
to initial baseline scores were, in many cases, due
to inadequate equipment. But a major part of the
reported failures of replication to train increases
in alpha amplitude was due simply to insufficient
total duration of training, as discussed by Hardt
11
and Kamiya (1976) and Ancoli and Kamiya
(1978). Several of the reported failures simply
reflected the stopping of training after one or
two sessions. In our laboratory we found that the
first and second training sessions, with each session lasting about 45 minutes, resulted in average
scores actually lower than the initial session baselines for each session. Plotting the average performance over six sessions of training, we saw a substantial drop in alpha relative to baseline scores
in the first session, followed by a gradual increase
in performance across sessions until the third session, when the trial scores and session baseline
scores were about the same. It was not until the
fourth session was reached that the trainees had
increased their performance sufficiently to exceed
their baseline for that session.
At least one factor, probably the major one, to
account for the puzzling drop in performance relative to the session initial baseline and slow recovery across three sessions is that the challenge to
find a mental state, feeling, and so forth, requires a
busy mind in search mode, but that reduces alpha
activity. Many a trainee has commented on the
fact that trying to solve how to increase the tone
level served only to reduce it. It is also possible that
the lack of progress is perceived as failure by the
trainee, and the resulting ego-threat and anxiety
cause more alpha reduction. Some of the trainees’
comments may best describe the situation: “The
harder I try, the more the tone goes away”; “I gave
up trying to increase the tone, and damned if it
didn’t get louder”; “I seem to do best when I just
wait and let it come on by itself and be happy
when it does.”
Overall, it seems that even though verbal
descriptions of the two states tend to agree among
trainees, and tend to support descriptions made
by earlier investigators, I believe that the use of
everyday language has its limits as a way of characterizing the subjective experiences associated
with the two states, particularly the state of alpha
dominance. In short, the results indicate that it
is far easier to detect a difference between these
two brain states than it is to be able to describe
what comprises the difference. However, it is also
possible that learning to discern that there is a difference between alpha and non-alpha is only a first
stage of coming into awareness of what the factors
behind the difference are.
The best hint yet of an answer to the question posed by our results may come from another
experiment that I had completed earlier. Because
12
it might be that the apparent rise in alpha in the
feedback experiment with the tone was at least to
some degree due to the trainee becoming accustomed to the laboratory over repeated session,
I gave new trainees alternating blocks of alphaincrease trials with alpha-decrease trials. I used
five 1-minute increase trials, providing a quantitative score of alpha output after each minute. Then
I delivered five 1-minute trials to decrease their
alpha tone, also with a score of alpha output after
each minute.
The subjects now learned to increase and
decrease their scores quite efficiently (Kamiya,
1968). It seemed more helpful in sharpening their
differentiation of the two in their verbal reports
than the task given them earlier to only increase
alpha amplitude. It is possible that the improvement reflects an increased opportunity to sense
the difference between the two states by having
them alternate within close temporal sequence the
internal behaviors or mental states that are instrumental in producing the two states.
Applications of the Method to Clinically Relevant
Physiological Measures
The field of EEG feedback or neurofeedback has
not been particularly concerned with research
issues such as the ones I (Kamiya) have been
describing. Instead, rather quickly it became clear
that the method of feedback training could yield
some immediately practical results in the clinic
as a method of treating neurological and psychological disorders that were known to be related to
specific characteristics in the EEG. The fields of
neurofeedback for the treatment of epilepsy and
attention-deficit/hyperactivity disorder (ADHD)
are two of the best examples. Sterman led the
way in neurofeedback for treating epilepsy when
he and Wyrwicka (Wyrwicka & Sterman, 1968)
reported that in cats, the Sensory Motor Rhythm,
their name for a burst of synchronized EEG activity over the motor cortex in an awake animal,
could be brought under control by operant conditioning. The same sensory motor rhythm, when
brought under operant control by humans, was
found to suppress epileptic seizures (Sterman &
Friar, 1972). Lubar and Shouse (1976) reported
that a hyperkinetic child could be treated successfully with sensory motor rhythm training, thus
starting the use of the method for the treatment of
ADHD in many different laboratories and clinics,
including that of Michael and Lynda Thompson
(1998). Thompson and Thompson (2003) have
I. ORIENTATION TO BIOFEEDBACK
since published a comprehensive book, The Neurofeedback Book, that has become a standard reference, as well as an aid for training therapists in
the methods of EEG. The investigators who have
worked and published their results in these fields
for several decades now, led by Sterman, Lubar, the
Thompsons, and several others have developed
protocols that will eventually revamp the thinking of the medical fields toward neurofeedback as a
treatment alternative to traditional pharmaceutical or surgical approaches.
Cybernetics
The term “biofeedback” is a shorthand term for
external psychophysiological feedback, physiological feedback, and sometimes augmented proprioception. The basic idea is to provide individuals
with increased information about what is going on
inside their bodies, including their brains.
The field that deals most directly with information processing and feedback is called cybernetics.
A basic principle of cybernetics is that one cannot control a variable unless information is available to the controller. The information provided
is termed “feedback” (Ashby, 1963; Mayr, 1970).
Another principle of cybernetics is that feedback makes learning possible. Annett (1969)
reviewed the evidence for this principle. In applied
biofeedback, individuals receive direct and clear
feedback about their physiology. This helps them
learn to control such functions. For example, from
a surface EMG instrument, persons receive information concerning their muscle activity. This
helps them to reduce, increase, or otherwise regulate their muscle tension.
From a cybernetic perspective, operant conditioning is one form of feedback. It is feedback
provided in the form of positive or negative results
of a particular behavior. The point is that another
significant contribution to the development of
applied biofeedback is an information-processing model derived from cybernetic theory and
research. Proponents of this model in the field of
biofeedback include Brown (1977), Anliker (1977),
Mulholland (1977), and Gaarder and Montgomery
(1981).
Cultural Factors
Several cultural factors have contributed to the
development of applied biofeedback. The gradual
13
1. History and Definitions
merging of the traditions and techniques of the
East and West is one major factor. The rise in
popularity of schools of meditation was an expression of a cultural change and provided a context
in which applied biofeedback developed. Yogis and
Zen masters reportedly alter their physiological
states significantly through meditation. Related
phenomena presumably occur in some forms of
biofeedback experiences. Therefore, some have
referred to biofeedback as the “yoga of the West”
and “electronic Zen.”
Within the United States, there are other
cultural factors adding to a Zeitgeist encouraging
biofeedback applications. These are the heightened costs of health care and the resulting need
for more efficacious and cost-effective treatments.
In addition, it is commonly recognized that pharmacotherapy, with its many benefits, is of limited
value for certain patients. Some patients cannot
take medications because of untoward side effects;
many patients avoid compliance; others prefer not
to consume medications; and some physicians
deemphasize pharmacotherapy.
Perhaps even more significant is the current
popular public health emphasis on prevention.
The movement toward wellness has continued
to grow since the 1960s. Practitioners of holistic health also emphasize self-regulation and
self-control. The result of these emphases is that
more people are involving themselves in lifestyle
changes to regulate their health. These changes
include enhancing physical fitness, avoiding caffeine and nicotine, reducing or stopping alcohol
use, and pursuing better weight control. More people are thus assuming increased responsibility for
their physical, as well as their mental and spiritual,
well-being. In addition, more people are accepting
responsibility for their recovery from illness. Many
believe that biofeedback therapies facilitate and
fit well into these efforts at greater self-regulation,
wellness, and growth.
Professional Organizations
Homer’s epic poem The Odyssey served as a metaphor for the past, present, and future of biofeedback
and applied psychophysiology. From the title of
this epic, an “odyssey” has come to mean any long
series of wanderings, especially when filled with
notable experiences, hardships, and the exploration of new terrain. Just as Homer’s Odysseus
experienced setbacks but was ultimately successful
in his journey to reach home, the journey of psychophysiological self-regulation with biofeedback
has experienced and will continue to experience
setbacks and successes. The Biofeedback Society
of America (BSA) was entering its 20th year, thus
completing one full generation of development,
when similar words were first delivered (M. S.
Schwartz, 1988). Twenty years constitute one generation, or the average period between the birth
of parents and the birth of their offspring. Thirteen years then remained until the year 2001, the
date of the famous book and movie 2001: A Space
Odyssey. However, our field does not seek the universality of something as monolithic as Arthur C.
Clarke’s and Stanley Kubrick’s odyssey.
The Association for Applied Psychophysiology
and Biofeedback and Its Various Names
How the Journey Began
The Biofeedback Research Society (BRS) was
formed in 1969, largely by a handful of research
psychophysiologists. After 6 years, the BRS
became the BSA, with both an experimental
and an applied division. Age 6 is about the age
at which children go through the transition from
home to school; similarly, the scope of the organization and the field broadened into applied areas.
This change in name reflected the growth and
importance of applied aspects.
How the Journey Continued
Professional Developments
Also adding to the development of applied biofeedback are the organizations of professionals
engaged in research and clinical, educational, and
performance enhancement applications. Issues
considered here include the professional organizations themselves, the status of the literature in
this field, the professional journal of the primary
organization (and the journal’s name), and finally,
the scope of the field.
At age 19, as a result of the field’s expanding scope,
the BSA went through its second transformation—into the Association for Applied Psychophysiology and Biofeedback (AAPB; www.aapb.
org). This is about the age at which many students
graduate to institutions of higher learning. The
organization returned to some of its roots in psychophysiology at the same interval. The consistency with the journey metaphor first struck M.
S. Schwartz (1988) then, as Odysseus also took 20
years to return home.
14
As later reported by M. S. Schwartz (1999a),
the name . . . change was a hotly debated topic. Many
argued for a need to expand the implied scope of the
organization. One factor was that most practitioners
utilized a wider array of therapy methods than biofeedback. Presentations at the annual meetings of
the BSA encompassed much more than biofeedback.
Researchers at universities . . . maintained that the
term biofeedback alone was not viewed as sufficiently
credible by some individuals and that this hampered
their abilities to publish their research in some quality journals and to obtain external research funding.
The researchers further contended that the term
“biofeedback” was insufficient for them to obtain the
kind of recognition they needed in their academic
departments. Thus, both applied practitioners and
researchers were contending that a name change was
needed.
Psychophysiology was the birthplace of the field
of biofeedback, and so it was time to return to these
roots. The emphasis was placed on the term applied
to distinguish it from [its] grandparent organization and field, the Society for Psychophysiological
Research.
Many members of the BSA . . . argued for dropping the term biofeedback but the supporters of
the term successfully argued for the preservation of
the term. . . . The term “applied psychophysiology”
reflected the evolution of science and clinical practice. (p. 3)
The AAPB continues to be a productive, intellectually stimulating, useful, scientifically sound,
and vibrant organization. There are several Interest Groups, Sections, and Divisions, including Sections for Applied Respiratory Psychophysiology,
Educational, International, Mind–Body, Optimal
Functioning, and Performing Arts Psychophysiology. There also now is a section for the U.S.
Stress Management Organization, which is part
of the International Stress Management Association (ISMA) with another interesting history dating from 1973, with illuminary founders Edmund
Jacobson, F. J. McGuigan, and Marigold Edwards.
Prior names for the international organization
included the International Stress and TensionControl Association and the International Stress
Management Association (ISMA). The Neurofeedback Division and the sEMG/SESNA (Surface Electromyography Society of North America)
Division reflect the two major modalitites and
areas of biofeedback. Each of these has major
tracks at the Annual Meeting of the AAPB.
Disagreement occasionally still arises about the
most appropriate name for both the membership
I. ORIENTATION TO BIOFEEDBACK
organization AAPB and its journal (see below).
Some occasionally argue for dropping the term
“biofeedback,” but those who advocate retaining
the term “biofeedback” in the names of the organization and journal focus on the established place
of this term in the minds of professionals and the
lay public, as well as on its history, brevity, and
ease of communication.
Other Related Membership Organizations and Groups
Sponsoring Meetings
Another national membership organization, the
American Association of Biofeedback Clinicians,
started in 1975 but went out of existence in the
late 1980s. This left the BSA, now the AAPB, as
the major organization with a major emphasis on
biofeedback.
Biofeedback’s impact is growing and spreading
beyond the borders of the United States, as evidenced by the rise of the Biofeedback Foundation
of Europe (BFE) (www.bfe.org). This excellent,
international organization has hosted an annual
meeting, featuring indepth workshops and scientific sessions, since 1996.
Since 1995, with the resurgence and expansion of EEG biofeedback, a specialty organization, the International Society for Neurofeedback
and Research (ISNR; www.isnr.org) has become
a major organization in this area. The ISNR was
formed in response to the need for a group that
was undividedly focused on EEG biofeedback.
There had been previously created, within the
AAPB, an “EEG Division” that attempted to serve
the needs of this community. However, the influence of those primarily interested in peripheral (or
“traditional”) biofeedback was considered by some
to be diluting these efforts, and it motivated certain individuals to create a new entity. Like the
AAPB, its name and focus has evolved over time,
but much more quickly. ISNR is an outgrowth of
the Society for the Study of Neuronal Regulation,
founded in 1993, whose name was shortened in
1998 to Society for Neuronal Regulation for simplicity, and then changed again in 2002, to the
International Society for Neuronal Regulation. In
2006, it was renamed ISNR “to better reflect the
fact that members of the society now came from
all parts of the globe, not just North America and
that research is a critical function of the society”
(www.isnr.org). The ISNR also provides publications, research support, education, and an annual
meeting.
15
1. History and Definitions
Each of the aforementioned organizations has
excellent websites with extensive and useful information.
The Biofeedback Certification International Alliance
The Biofeedback Certification International Alliance (BCIA), previously known as the Biofeedback
Certification Institute of America, is a professional
organization that has greatly influenced the continued development of the field. As its name indicates, the BCIA maintains a credible credentialing program. Before 1979, credentialing was in the
hands of a few state biofeedback societies. These
societies, well-meaning as they were, suffered from
the understandable problems of small groups of
professionals who typically had little or no training
and experience with the complexities of credentialing. Thus, there was considerable variability in
the credentialing across states. In most states, there
was no credentialing at all or even the hope of any.
Ed Taub, then president-elect of the BSA, had
the foresight and wisdom to inspire the development of an independent, credible, nationwide
credentialing program. The BSA sponsored and
supported the official establishment of the BCIA
(named by Bernard Engel, later the first chair of
the BCIA board) in January 1981. Three months
later, when Engel became President of the BSA, he
graciously relinquished the chair of BCIA to M. S.
Schwartz. The BCIA evolved with more stringent
criteria for education, training, experience, and
recertification. Professionals continue to seek and
earn the BCIA credential as the only credible one
of its kind.
In recent years, another so-called credentialing
organization arose in association with instrumentation its members refer to as “biofeedback,” but all
credible professionals known to at least the first
author consider this group or what its members call
biofeedback to be inaccurate or to lack credibility
and not worthy of mention in this chapter or book.
Although the BCIA holds primacy in credentialing, educational opportunities exist in many
undergraduate and graduate courses in biofeedback. Private training programs and workshops are
offered by national, state, and regional professional
organizations, as well as some biofeedback companies/distributors. There are also many companies
manufacturing biofeedback instrumentation, and
several “distributor” companies selling and servicing a variety of instruments from different manufacturers.
The Journey of a Family or Separate Journeys?
All professionals in this field share some joint
responsibility and custody for the young adult we
call “biofeedback and applied psychophysiology.”
Some professionals proceed on their own individual journeys; they seek their own destinations,
their own Ithacas, instead of common ones. However, the AAPB continues as the leading administrative, facilitative, educational, and coordinating member organization dedicated to integrating
professional disciples and conceptual frameworks
that involve varied scientific and applied areas of
applied psychophysiology and biofeedback. It is
the nuclear family for biofeedback.
Status of the Literature in the Field
The number of publications is one barometer of
the history, growth, and possibly the future of a
field. The first bibliography of the biofeedback literature (Butler & Stoyva, 1973) contained about
850 references. The next edition, 5 years later,
listed about 2300 references (Butler, 1978). Thousands more have appeared since then (Hatch,
1993; Hatch & Riley, 1985; Hatch & Saito, 1990).
There are dozens of papers published each year in
non-English-speaking countries, including Russia
(Shtark & Kall, 1998; Shtark & Schwartz, 2002;
Sokhadze & Shtark, 1991), and many others are
published in Europe, Israel, and elsewhere.
Note that there are dozens of papers published
each year in non-English-speaking countries. For
example, the important Japanese literature was
still in its early stages in 1979, but rapidly increased
in the 1980s (Hatch & Saito, 1990; Shirakura,
Saito, & Tsutsui, 1992). Their leading journal
on biofeedback, Japanese Journal of Biofeedback
Research, is nearing its 40th volume. There is also
a rich history of research publications and clinical applications in Russia and other countries that
were formerly part of the USSR (Shtark & Kall,
1998; Shtark & Schwartz, 2002; Sokhadze &
Shtark, 1991). This foreign literature is not well
known in the United States.
The Primary Journal, Its Name,
and Other Publications
A measure of the maturity of a field is the existence
and quality of its primary professional journal(s).
The journal Biofeedback and Self-Regulation, published by Plenum Press, was started in 1976. The
journal’s name was changed to Applied Psycho-
16
physiology and Biofeedback as of Volume 22, in
1997. The editors, board, and publisher noted that
“the journal has long had a broader focus than the
title implied, and this new name more accurately
reflects its expanded scope” (Andrasik, 1997, p. 1).
Frank Andrasik has been the Editor-in-Chief
since 1995, having followed many notable prior
editors—Johann Stoyva, the first editor, Al F. Ax,
coeditors Edward B. Blanchard & Mary R. Cook
from 1984 until 1992, and Robert R. Freedman
until 1995. It is still the major publication in this
field. However, AAPB also publishes another very
useful and important publication, called simply
Biofeedback. For the past several years, with Donald Moss, as Editor-in-Chief, this has become an
excellent quarterly publication. Another noteworthy journal is the Journal of Neurotherapy, which
focuses on EEG biofeedback/neurofeedback.
Definitions of Biofeedback
and Applied Psychophysiology
Historical Review of Definitions
The history of biofeedback has witnessed many
definitions. Olson (1987, 1995) noted 10 definitions starting from 1971. In the second and third
editions of this text, Schwartz and Schwartz (2003)
elaborated and discussed various historical definitions; the models from which they were derived;
and the issues, elements, and factors involved in
prior definitions. For example, whether or not the
specific feedback signals as such result in changes
and at what level does the signal become biofeedback, per se, was a focus of much debate in the
1980s. See the invigorating exchange and debate
between Furedy (1987) and Shellenberger and
Green (1987), a valuable and appreciated attempt
to moderate and create perspective by Rosenfeld
(1987), and the review and discussion of this by
Schwartz and Schwartz (2003).
Some persons might still consider these topics
interesting. However, we decided to deemphasize
these topics in this edition partly in view of the
2008 official definition (AAPB, BCIA, and ISNR)
(Schwartz, 2010) presented and discussed later in
this chapter. This was done chiefly to reduce confusion and not detract from the official definition.
Increased information and patient education
are common elements in all models. We suggest
a conceptualization that includes different levels
and types of information received by patients during biofeedback sessions. This discussion acknowledges the contributions of G. E. Schwartz (1982,
I. ORIENTATION TO BIOFEEDBACK
1983), who emphasized the contextual, organistic,
multicategory, and multicausal approach to understanding biofeedback.
Schwartz and Schwartz (2003) presented and
discussed their multilevel patient education model
involving seven levels or facets of information
about biofeedback. Readers are referred to the
third edition and to www.marksschwartzphd.com
for a full discussion of this model. This model proposed that patient education is an active ingredient
of biofeedback, regardless of the discipline within
which it is used. This component is not explicitly
included in the new and official definition but it is
implicitly “in conjunction with changes in thinking” (Schwartz, 2010, p. 90).
Toward the First Official Definition
of “Biofeedback”
By Olson’s (1995) definition, a competent therapist is an important part of biofeedback therapies.
Moreover, computerized biofeedback is like having
a high-tech electronic chalkboard for teaching and
a built-in ability to measure progress. It is up to
the therapist to use this technology to be the best
possible teacher and communicator. In essence,
biofeedback, used in the broad sense of signals,
explanations, and patient education, provides missing or deficient information in the intervention context. This information is helpful for the patient/client,
the therapist, or the interaction.
One does not evaluate a school book when it is
presented to students by itself. Some students have
the following: sufficient motivation, sufficient
capabilities, no significant interference, sufficient
times and places to study, other resources to use as
references, an experiential background conducive
to independent learning, confidence in their ability, and a teacher for help if they reach an impasse.
Therefore, some students do well with self-study
and never need to go to class. Others need classroom instructions and review of the text. Some
of these others need extensive text review—paragraph by paragraph, page by page, and chapter by
chapter. Some learn the material sufficient for
earning an average grade. Others seek or “need” a
grade of A. Some never learn much, if any at all.
None of this is news. However, the point here is
that we do not attribute the problem to the book
unless it is written poorly and/or not tailored well
to the student.
In Schwartz and Schwartz (2003) a comprehensive definition was offered that involved additions
to Olson’s (1995) definition.
1. History and Definitions
The 2008 Official Definition
In mid-2007, the leadership of AAPB5 started the
process and coordinated the creation of the Task
Force6 on Nomenclature, a task force to develop
an agreed-upon definition of “biofeedback” that
would be endorsed by the three major organizations, the AAPB, the BCIA, and the ISNR. The
task force’s diligent work on this challenging
project over several months culminated in a definition that was then submitted to the Boards of
the three organizations that had contributed task
force members. The Boards voted their agreement
in 2008, and the definition became the first, official, agreed-upon definition in the field. The story
of this process may be found in M. S. Schwartz
(2010).
Biofeedback is a process that enables an individual
to learn how to change physiological activity for the
purposes of improving health and performance. Precise instruments measure physiological activity such
as brainwaves, heart function, breathing, muscle
activity, and skin temperature. These instruments
rapidly and accurately “feed back” information to the
user. The presentation of this information—often
in conjunction with changes in thinking, emotions, and behavior—supports desired physiological
changes. Over time, these changes can endure without continued use of an instrument. (Approved May
18, 2008, by the AAPB, the BCIA, and the ISNR).
A Definition of “Applied Psychophysiology”—
Sort Of
Defining the term “applied psychophysiology” still
remained a need, goal, and challenge as of 1998,
several years after the name change for AAPB and
its journal. As noted by M. S. Schwartz (1999a,
p. 4), “One can only surmise that everyone
apparently knew what applied psychophysiology
meant. . . . What everyone apparently knew, no
one had written. What everyone apparently knew,
was unclear.” It was the broader term, a rubric
term, that subsumes biofeedback.
J. Peter Rosenfeld (1992), in his AAPB presidential address, was the first to address a definition of “applied psychophysiology.” He identified
some of its elements “and touched on elements
of a definition” (M. S. Schwartz, 1999a, p. 4).
Sebastian Striefel (1998), a later president of the
AAPB, again raised the question of a definition of
“applied psychophysiology” in his 1998 presidential address. At the same meeting, “Paul Lehrer,
chairperson of the AAPB Publication Committee,
17
convened an ad hoc committee to deal with a wide
array of topics. . . . One of these topics was . . . the
lack of a formal . . . definition of ‘applied psychophysiology’ ” (M. S. Schwartz, 1999a, p. 4). The
committee assigned the task of establishing an
operational definition for the term. Apparently, no
one thought to establish a task force. The AAPB
asked one person to develop a definition (M. S.
Schwartz, 1999a, 1999b).
A provisional definition was drafted and a
paper documenting the rationale for each component was written. An array of notable and diverse
professionals provided their critiques to the provisional definition in the initial paper by M. S.
Schwartz (1999b). The author of the definition
then prepared a response to the panel of independent critical reviewers (M. S. Schwartz, 1999b).
The development of a definition that is acceptable
to everyone is unlikely. Amendments and modifications were expected. The published discussions
of the key elements, examples of topics included
and excluded, rationales for these choices, critiques, and responses are best read in their original form. There is still no formal and agreed-upon
definition of “applied psychophysiology”—only a
tentative and certainly unofficial operational definition (M. S. Schwartz, 1999a, p. 5) presented here
only for historical interest and, we hope, to motivate others to refine and shorten it.
Applied psychophysiology reflects an evolving scientific discipline and specialty involving understanding
and modifying the relationship between behavior
and physiological functions by a variety of methods
including noninvasive physiological measures. The
term “applied psychophysiology” is a rubric encompassing evaluation, diagnosis, education, treatment,
and performance enhancement.
Applied psychophysiology includes a group of
interventions and evaluation methods with the
exclusive or primary intentions of understanding and
effecting changes that help humans move toward and
maintain healthier psychophysiological functioning.
Applied psychophysiology involves helping people
change physiological functioning and psychological functioning (measured, theoretical, and potential) and/or to achieve sensorimotor integration and
motor learning within physical rehabilitation.
The group of interventions use all forms of biofeedback, relaxation methods, breathing methods,
cognitive-behavioral therapies, patient/client education, behavioral changes, hypnosis, meditative techniques, and imagery techniques (some commentators
would add: when directed at changing physiological
functioning). In some situations, dietary and other
biochemical (nonmedication) changes and some
18
truth detection research and applications may be
considered under the rubric of applied psychophysiology.
Evaluation methods use all forms of physiological measurements. The physiological functioning
includes but is not limited to accurately measured
changes in skeletal muscles, all autonomic physiology, breathing measures, biochemistry, electroencephalographic activity, both normal and abnormal and imaging techniques. Autonomic measures
include electrodermal, skin temperature, blood pressure, heart rate, gastrointestinal motility, and vasomotor.
The interventions need to be part of or have
implications for applications to humans. These
could, but do not need to, involve the raw procedures
and/or symptoms of medical and psychophysiological
disorders.
Glossary
Alpha wave activity. EEG activity (8–12 hz) commonly,
but not always, thought to be associated with an alert
but relaxed state.
Autonomic nervous system (ANS). The part of the ner-
vous system that is connected to all organs and blood
vessels, and transmits signals that control their functioning. It consists of two branches, the sympathetic
and parasympathetic, which usually produce opposite
responses. Once thought to be totally involuntary, it
is now known to be under some significant voluntary
control, although less so than the CNS.
Central nervous system (CNS). The part of the nervous
system including human thought, sense organs, and
control of skeletal muscles. Once believed to be
totally separate from the ANS, it is now known to
interact with the ANS.
Classical conditioning. Originating with Pavlov, the
type of conditioning or learning that assumes that
certain stimuli (unconditioned stimuli, or UCSs)
evoke unconditioned or unlearned responses (UCRs)
(e.g., acute pain evokes crying, withdrawal, and fear),
and that other, previously neutral stimuli (conditioned stimuli, or CSs) associated with the pairing of
these events develop the capacity to elicit the same
or similar responses or conditioned responses (CRs).
Curarized animals. Animals intentionally paralyzed by
the drug curare to control for body movements during visceral conditioning, such as biofeedback of
heart rate.
Cybernetics. The science of internal body control sys-
tems in humans, and of electrical and mechanical
systems designed to replace the human systems.
Electroencephalography (EEG). The measurement of
electrical activity of the brain.
I. ORIENTATION TO BIOFEEDBACK
Electromechanical. A term describing devices that mea-
sure mechanical aspects of the body (e.g., position of
a joint or degree of pressure or weight placed on it),
rather than a property of the body (e.g., its direct
electrical activity or temperature). Examples of these
mechanical aspects include degrees that a person’s
knee bends after knee surgery, steadiness of the head
of a child with cerebral palsy, and the weight pressure
placed on a leg and foot by someone after a stroke.
Instruments transform these mechanical forces into
electrical signals.
Electromyography (EMG). The use of special instru-
ments to measure the electrical activity of skeletal
muscles. In recent years, also called “surface electromyography” and sometimes abbreviated as SEMG.
Extinction. The behavioral principle predicting that
abruptly and totally stopping all positive reinforcements after specified behaviors will lead to the
behavior’s no longer occurring.
Fading. Gradually changing a stimulus that controls a
person’s or animal’s performance to another stimulus.
As a behavioral procedure, it does not always mean
disappearance of a stimulus.
Fight or flight. Walter Cannon’s well-known concept of
the body’s psychophysiological arousal and preparation for fighting or fleeing actual or perceived threatening stimuli.
Galvanic skin response (GSR). A form of electrodermal
activity—increased resistance of the skin to conducting tiny electrical currents because of reduced sweat
and dryness. Older term less ofen used now, but still
accepted. Opposite of “skin conductance” (SC).
Insight-oriented psychodynamic theories and therapies. A wide range of psychological theories and
therapies, starting from the time of Sigmund Freud. A
basic assumption is that patients need to gain insight
into the psychological origins and forces motivating
their current psychological problems and behaviors
before they can achieve adequate relief of symptoms.
Instrumental conditioning. Same as operant condition-
ing (see below). The behavioral theories and therapies originated by B. F. Skinner. For example, reinforcers are said to be instrumentally linked to the
recurrence of behaviors.
Observational learning. Learning that takes place by
means of the organism’s observing another organism
doing the task to be learned.
Operant conditioning. The same as instrumental condi-
tioning (listed earlier), originating with B. F. Skinner. “Operant” means that a response is identified
and understood in terms of its consequences rather
than by a stimulus that evokes it. Stimuli and circumstances emit responses rather than evoke them, as in
classical conditioning.
19
1. History and Definitions
Proprioception. Perception mediated by sensory nerve
terminals within tissues, mostly muscles, tendons,
and the labyrinthal system for balance. They give
us information concerning our movements and position. Examples include (1) the sense of knowing
when we are slightly off balance; and (2) the ability to perceive (even with eyes closed) the difference
between, and approximate weights of, objects weighing 5 ounces and 7 ounces held in each hand.
Psychophysiology. The science of studying the causal
and interactive processes of physiology, behavior, and
subjective experience.
Reinforcers. Events or stimuli that increase the prob-
Zeitgeist. The spirit or general trend of thought of a
time in history. Often used to refer to a time in history when new ways of thinking and technologies are
more likely to be accepted by the culture in question.
Acknowledgments
We gratefully acknowledge and will always be thankful to R. Paul Olson for his earlier contributions to this
chapter. His creativity and scholarly writing of the original version of this chapter continue to be the model we
follow.
ability of recurrence of behaviors they follow.
Schedules of reinforcement. Usually, forms of intermit-
tent reinforcement of an operant behavior. A common schedule in life, and most resistant to extinction, is a variable-ratio schedule—one in which the
number of times a reinforcement follows a specific
behavior varies randomly, so the person or animal
never knows when the reinforcer will occur. This
contrasts with variable-interval, fixed-interval, and
fixed-ratio schedules.
Sensorimotor rhythm. An EEG rhythm (12–14 Hz)
recorded from the central scalp and involving both
the sensory and motor parts of the brain, the sensorimotor cortex. Used in the EEG biofeedback of some
persons with seizure disorders.
Shaping. A behavioral principle from operant con-
ditioning, referring to procedures designed to help
learning of complex new behaviors by very small
steps. Also known as “shaping by successive approximations.”
Single motor units. Individual spinal nerves or neu-
rons involved in movement. Biofeedback training of
single spinal motor neurons was a major advance in
the late 1950s and early 1960s. This training requires
fine-wire EMG electrodes.
Skeletally mediated mechanical artifacts. Artifacts in
instrumentation-recorded signals that are caused
by intentional body movements. Examples include
moving a body part such as the head or neck during
recordings of resting muscle activity, or clenching the
teeth during EEG recordings.
Slow-wave activity. EEG activity (3–8 Hz) included in
the frequency range often called theta activity, also
reported as 4–7 Hz.
Vasomotor. Affecting the caliber (diameter) of a blood
vessel.
Visceral learning. Learning that takes place by body
organs, especially those in the abdominal cavity,
such as the stomach and bowels.
Visceral reflexes. Reflexes in which the stimulus is a
state of an internal organ.
Notes
1. Although the term “applied psychophysiology” is
now usually given first in this pairing, the order is reversed
here to relect the emphasis on biofeedback.
2. The 25th anniversary meeting of the primary professional membership organization, the Association for
Applied Psychophysiology and Biofeedback (AAPB), was
held in 1994. The commemorative AAPB Silver Anniversary Yearbook published for that meeting contains articles
about the history and development of the biofeedback field
and the organization. Reading it is enriching and informative. It is available from the AAPB, 10100 West 44th
Avenue, Suite 304, Wheat Ridge, CO 80033; phone: 303422-2615; fax: 303-422-889; website: www.aapb.org.
3. Note that this sentence appeared in the first edition
of this book in early 1987. It does not seem to be a coincidence that the Biofeedback Society of America (BSA)
went through the process of changing its name to include
board meetings or other public or private meetings concerning the name change. The term was written into an
early draft of this chapter several years before 1987.
4. “Health psychology” is a more recent field with similar roots and ties to behavioral medicine. The focus is more
on prevention and health enhancement.
5. The AAPB Executive Board and, specifically and
most notably, Aubrey Ewing, then President-Elect and Alan
Glaros, the President, along with Executive Director Francine Butler, were the prime movers on the project. They
coordinated with the leadership of two other major organization in this field, the BCIA and the ISNR.
6. See Schwartz (2010) for a list of the members.
References
Ancoli, S., & Kamiya, J. (1978). Methodological issues in
alpha biofeedback training. Biofeedback and Self-Regulation, 3(2), 159–183.
Andrasik, F. (1997). Editorial. Applied Psychophysiology and
Biofeedback, 22, 1.
Andrews, J. M. (1964). Neuromuscular re-education of the
20
hemiplegic with aid of electromyograph. Archives of
Physical Medicine and Rehabilitation, 45, 530–532.
Anliker, J. (1977). Biofeedback from the perspective of
cybernetics and systems science. In J. Beatty & H.
Legewie (Eds.), Biofeedback and behavior. New York: Plenum Press.
Annett, J. (1969). Feedback and human behavior. Baltimore,
MD: Penguin Books.
Aserinsky, N., & Kleitman, N. (1953). Regularly occurring
periods of eye motility and concomitant phenomena
during sleep. Science, 118, 274–284.
Ashby, W. R. (1963). An introduction to cybernetics. New
York: Wiley.
Ayers, M. E., Sams, M. W., Sterman, M. B., & Lubar, J.
(2000). When to Inhibit EEG activity instead of reinforcing and inhibiting simultaneously. Journal of Neurotherapy, 4(1), 83–90.
Basmajian, J. V. (1963). Conscious control of individual
motor units. Science, 141, 440–441.
Basmajian, J. V. (1978). Muscles alive: Their functions
revealed by electromyography (4th ed.). Baltimore, MD:
Williams & Wilkins.
Basmajian, J. V. (Ed.). (1983). Biofeedback: Principles and
practice for clinicians (2nd ed.). Baltimore, MD: Williams
& Wilkins.
Basmajian, J. V. (Ed.). (1989). Biofeedback: Principles and
practice for clinicians (3rd ed.). Baltimore, MD: Williams
& Wilkins.
Basmajian, J. V., Kukulka, C. G., Narayan, M. G., &
Takebe, K. (1975). Biofeedback treatment of foot drop
after stroke compared with standard rehabilitation
technique: Effects on voluntary control and strength.
Archives of Physical Medicine and Rehabilitation, 56,
231–236.
Beatty, J., Greenberg, A., Deibler, W. P., & O’Hanlon, J.
F. (1974). Operant control of occipital theta rhythm
affects performance in a radar monitoring task. Science,
183(4), 871–873.
Benson, H. (1975). The relaxation response. New York: Morrow.
Bernstein, D. A., & Borkovec, T. D. (1973). Progressive
relaxation training. Champaign, IL: Research Press.
Bernstein, D. A., Carlson, C. R., & Schmnidt, J. E. (2007).
Progressive relaxation: Abbreviated methods. In P. M.
Lehrer, R. L. Woolfolk, & W. E. Sime (Eds.), Principles
and practice of stress management (3nd ed., pp. 88–122).
New York: Guilford Press.
Binder-MacLeod, S. A. (1983). Biofeedback in stroke rehabilitation. In J. V. Basmajian (Ed.), Biofeedback: Principles and practice for clinicians (2nd ed., pp. 73–89).
Baltimore, MD: Williams & Wilkins.
Birk, L. (Ed.). (1973). Biofeedback: Behavioral medicine. New
York: Grune & Stratton.
Blumenthal, A. L. (1977). The process of cognition. Englewood Cliffs, NJ: Prentice Hall.
Brown, B. (1977). Stress and the art of biofeedback. New York:
Harper & Row.
Brudny, J. (1982). Biofeedback in chronic neurological
I. ORIENTATION TO BIOFEEDBACK
cases: Therapeutic electromyography. In L. White & B.
Tursky (Eds.), Clinical biofeedback: Efficacy and mechanisms (pp. 249–288). New York: Guilford Press.
Bruner, J. S. (1966). Toward a theory of instruction. Cambridge, MA: Belknap Press of Harvard University.
Butler, F. (1978). Biofeedback: A survey of the literature. New
York: Plenum Press.
Butler, F., & Stoyva, J. (1973). Biofeedback and self-control:
A bibliography. Wheat Ridge, CO: Biofeedback Society
of America.
Cannon, W. B. (1932). The wisdom of the body. New York:
Norton.
Carrington, P. (1977). Freedom in meditation. Garden City,
NY: Doubleday/Anchor.
Carrington, P. (1978). Clinically Standardized Meditation
(CSM) professional pack. Kendall Park, NJ: Pace Educational Systems.
Carrington, P. (1998). The book of meditation. Kendall Park,
NJ: Pace Educational Systems.
Carrington, P. (2007). Modern forms of Mantra Meditation. In P. M. Lehrer, R. L. Woolfolk, & W. E. Sime
(Eds.), Principles and practice of stress management (3nd
ed., pp. 139–168). New York: Guilford Press.
Collura, T. F. (1995). History and evolution of computerized electroencephalography. Journal of Clinical Neurophysiology, 12(3), 214–229.
Collura, T. F. (2009). Towards a coherent view of brain
connectivity. Journal of Neurotherapy, 12(2–3), 99–110.
Dement, W., & Kleitman, N. (1957). The relation of eye
movements during sleep to dream activity: An objective
method for the study of dreaming. Journal of Experimental Psychology, 53, 339–346.
Dewan, E. D. (1967). Occipital alpha rhythm, eye position
and lens accommodation. Nature, 214, 299–304.
DiCara, L., & Miller, N. E. (1968a). Instrumental learning of systolic blood pressure responses by curarized rats:
Dissociation of cardiac and vascular changes. Psychosomatic Medicine, 30, 489–494.
DiCara, L., & Miller, N. E. (1968b). Instrumental learning
of peripheral vasomotor responses by rat. Communications in Behavioral Biology, 1, 209–212.
DiCara, L., & Miller, N. E. (1968c). Instrumental learning
of vasomotor responses by rats: Learning to respond differentially in the two ears. Science, 159, 1485–1486.
DiCara, L., & Miller, N. E. (1968d). Long term retention
of instrumentally learned heart-rate changes in the
curarized rat. Communications in Behavioral Biology, A2,
19–23.
Dworkin, B. R., & Miller, N. E. (1986). Failure to replicate
visceral learning in the acute curarized rat preparation.
Behavioral Neuroscience, 100, 299–314.
Fehmi, L., & Collura, T. F. (2007). The effects of electrode
placement upon EEG biofeedback training: The monopolar/bipolar controversy. Journal of Neurotherapy, 11(2),
45–63.
Fehmi, L. G., & Fritz, G. (1980, Spring). Open focus: The
attentional foundation of health and well being. Somatics, pp. 24–30.
1. History and Definitions
Fehmi, L., & Robbins, J. (2009). The open-focus brain: Harnessing the power of attention to heal mind and body. Boston: Trumpeter/Shambhala.
Forem, J. (1974). Transcendental meditation. New York: Dutton.
Furedy, J. J. (1987). Specific versus placebo effects in biofeedback training: A critical lay perspective. Biofeedback
and Self-Regulation, 12, 169–184.
Gaarder, K. R., & Montgomery, P. S. (1981). Classical biofeedback: A procedural manual for behavioral medicine
(2nd ed.). Baltimore, MD: Williams & Wilkins.
Greenfield, N. S., & Sternback, R. A. (1972). Handbook of
psychophysiology. New York: Holt, Rinehart & Winston.
Hardt, J., & Kamiya, J. (1976). Some comments on Plotkin’s
self-regulation of the electroencephalographic alpha.
Journal of Experimental Psychology: General, 105(1),
100–108.
Harlow, H. F., & Harlow, M. K. (1962). Social deprivation
in monkeys. Scientific American, 207, 136–146.
Harris, A. H., & Brady, J. V. (1974). Animal learning: Visceral and autonomic conditioning. Annual Review of
Psychology, 25, 107–133.
Hart, J. T. (1968). Autocontrol of EEG alpha [Abstract].
Psychophysiology, 4, 506.
Hassett, J. (1978). A primer of psychophysiology. San Francisco: Freeman.
Hatch, J. P. (1993, March). Declining rates of publication
within the field of biofeedback continue: 1988–1991. In
Proceedings of the 24th annual meeting of the Association
for Applied Psychophysiology and Biofeedback, Los Angeles. Wheat Ridge, CO: Association for Applied Psychophysiology and Biofeedback.
Hatch, J. P., & Riley, P. (1985). Growth and development
of biofeedback: A bibliographic analysis. Biofeedback and
Self-Regulation, 10(4), 289–299.
Hatch, J. P., & Saito, I. (1990). Growth and development
of biofeedback: A bibliographic update. Biofeedback and
Self-Regulation, 15(1), 37–46.
Hothersall, D., & Brener, H. (1969). Operant conditioning of changes in heart rates in curarized rats. Journal of
Comparative and Physiological Psychology, 68, 338–342.
Jacobson, E. (1938). Progressive relaxation. Chicago: University of Chicago Press.
Jacobson, E. (1978). You must relax. New York: McGrawHill.
Jacobson, E. (1982). The human mind: A physiological clarification. Springfield, IL: Thomas.
Jacobson, E., & McGuigan, F. J. (1982). Principles and practice of progressive relaxation: A teaching primer [Cassette].
New York: BMA Audio Cassettes.
Kamiya, J. (1961). Behavioral, subjective, and physiological aspects of drowsiness and sleep. In D. W. Fiske & S.
R. Maddi (Eds.), Functions of varied experience (pp. 145–
174). Homewood, IL: Dorsey Press.
Kamiya, J. (1968). Conscious control of brain waves. Psychology Today, 1, 56–63.
Kamiya, J. (1969). Operant control of the EEG alpha
rhythm and some of its reported effects on conscious-
21
ness. In C. T. Tart (Ed.), Altered states of consciousness
(pp. 489–501). New York: Wiley.
Kimmel, H. O. (1979). Instrumental conditioning of autonomically mediated responses in human beings. American Psychologist, 29, 325–335.
Krippner, S. (1972). Altered states of consciousness. In J.
White (Ed.), The highest state of consciousness (pp. 1–5).
Garden City, NY: Doubleday.
Kristeller, J. (2007). Mindfulness meditation. In P. M. Lehrer, R. L. Woolfolk, & W. Sime (Eds.), Principles and
practice of stress management (3rd. ed., pp. 393–427).
New York: Guilford Press.
Langley, L. L. (1965). Homeostasis. New York: Van Nostrand Reinhold.
Lichstein, K. L. (1988). Clinical relaxation strategies. New
York: Wiley.
Lubar, J. F. (1982). EEG operant conditioning in severe
epileptics: Controlled multidimensional studies. In L.
White & B. Tursky (Eds.), Clinical biofeedback: Efficacy and mechanisms (pp. 288–310). New York: Guilford
Press.
Lubar, J. F. (1983). Electroencephalographic biofeedback
and neurological applications. In J. V. Basmajian (Ed.),
Biofeedback: Principles and practice for clinicians (2nd ed.,
pp. 67–90). Baltimore, MD: Williams & Wilkins.
Lubar, J. F. (1991). Discourse on the development of EEG
diagnostics and biofeedback treatment for attentiondeficit/hyperactivity disorders. Biofeedback and Self-Regulation, 16, 201–225.
Lubar, J. F., & Shouse, M. N. (1976). EEG and behavioral
changes in a hyperkinetic child concurrent with training of the sensory motor rhythm (SMR): A preliminary
report. Biofeedback and Self-Regulation, 3, 293–306.
Luthe, W. (Ed.). (1969). Autogenic therapy (Vols. 1–6). New
York: Grune & Stratton.
Marinacci, A. A., & Horande, M. (1960). Electromyogram
in neuromuscular re-education. Bulletin of the Los Angeles Neurological Society, 25, 57–71.
Mayr, O. (1970). The origins of feedback control. Cambridge,
MA: MIT Press.
McGuigan, F. J., & Lehrer, P. M. (2007). Progressive relaxation: Origins, principles, and clinical applications.
In P. M. Lehrer, R. L. Woolfolk, & W. E. Sime (Eds.),
Principles and practice of stress management (3rd ed.,
pp. 57–87). New York: Guilford Press.
Miller, N. E. (1978). Biofeedback and visceral learning.
Annual Review of Psychology, 29, 373–404.
Miller, N. E., & Brucker, B. S. (1979). A learned visceral
response apparently independent of skeletal ones in
patients paralyzed by spinal lesions. In N. Birbaumer
& H. D. Kimmel (Eds.), Biofeedback and self-regulation
(pp. 287–304). Hillside, NJ: Erlbaum.
Miller, N. E., & DiCara, L. (1967). Instrumental learning of
heart rate changes in curarized rats: Shaping and specificity to discriminative stimulus. Journal of Comparative
and Physiological Psychology, 63, 12–19.
Mitchell, L. (1977). Simple relaxation: The physiological
method for easing tension. New York: Atheneum.
22
Mitchell, L. (1987). Simple relaxation: The Mitchell method
for easing tension (rev. ed.). London: Murray.
Moss, C. S. (1965). Hypnosis in perspective. New York: Macmillan.
Mulholland, T. (1977). Biofeedback as scientific method. In
G. E. Schwartz & J. Beatty (Eds.), Biofeedback: Theory
and research (pp. 9–28). New York: Academic Press.
Nowlis, D. P., & Kamiya, J. (1970). The control of electroencephalographic alpha rhythms through auditory feedback and the associated mental activity. Psychophysiology, 6, 476–484.
Olson, R. P. (1987). Definitions of biofeedback. In M. S.
Schwartz (Eds.), Biofeedback: A practitioner’s guide. New
York: Guilford Press.
Olson, R. P. (1995). Definitions of biofeedback and applied
psychophysiology. In M. S. Schwartz & Associates
(Eds.), Biofeedback: A practitioner’s guide (2nd ed.). New
York: Guilford Press.
Olton, D. S., & Noonberg, A. R. (1980). Biofeedback: Clinical applications in behavioral medicine. Englewood Cliffs,
NJ: Prentice Hall.
Orne, M. T. (1979). The efficacy of biofeedback therapy.
Annual Review of Medicine, 30, 489–503.
Ornstein, R. E. (1972). The psychology of consciousness. San
Francisco: Freeman.
Pelletier, K. R., & Garfield, C. (1976). Consciousness: East
and west. New York: Harper & Row (Harper Colophon
Books).
Peper, E., & Shaffer, F. (2010). Biofeedback history: An
alternative view. Biofeedback, 38(4), 142–147.
Pi Suñer, A. (1955). Classics of biology. New York: Philosophical Library.
Progoff, I. (1980). The practice of process meditation. New
York: Dialogue House Library.
Rosenfeld, J. P. (1987). Can clinical biofeedback be scientifically validated?: A follow-up on the Green–Shellenberger–Furedy–Roberts debates. Biofeedback and SelfRegulation, 121(3), 217–222.
Rosenfeld, J. P. (1992). New directions in applied psychophysiology. Biofeedback and Self-Regulation, 17, 77–87.
Rosenthal, T. L., & Zimmerman, B . J. (1978). Social learning and cognition. New York: Academic Press.
Schwartz, G. E. (1982). Testing the psychosocial model:
The ultimate challenge facing biobehavioral medicine.
Journal of Consulting and Clinical Psychology, 50(6),
1040–1053.
Schwartz, G. E. (1983). Social psychophysiology and
behavioral medicine: A systems perspective. In J. T.
Cacioppo & R. E. Perry (Eds.), Social psychophysiology:
A sourcebook. New York: Guilford Press.
Schwartz, G. E., & Beatty, J. (Eds.). (1977). Biofeedback:
Theory and research. New York: Academic Press.
Schwartz, G. E., & Weiss, S. M. (1978). What is behavioral
medicine? Psychosomatic Medicine, 39(6), 377–381.
Schwartz, M. S. (1988). The biofeedback odyssey: Nearing
one score and counting (Presidential address). Biofeedback and Self-Regulation, 13(1), 1–7.
Schwartz, M. S. (1999a). What is applied psychophysiol-
I. ORIENTATION TO BIOFEEDBACK
ogy?: Toward a definition. Applied Psychophysiology and
Biofeedback, 24, 3–10.
Schwartz, M. S. (1999b). Responses to comments and closer
to a definition of applied psychophysiology? Applied Psychophysiology and Biofeedback, 24, 43–54.
Schwartz, M. S. (2010). A new improved universally
accepted official definition of biofeedback: Where did
it come from? Why? Who did it? Who is it for? What’s
next? Biofeedback, 24, 88–90.
Schwartz, N. M., & Schwartz, M. S. (2003). Definitions
of biofeedback and applied psychophysiology. In M. S.
Schwartz & F. Andrasik (Eds.), Biofeedback: A practitioner’s guide (3rd ed., pp. 27–39). New York: Guilford
Press.
Selye, H. (1974). Stress without distress. Philadelphia: Lippincott.
Selye, H. (1976). The stress of life (rev. ed.). New York:
McGraw-Hill.
Selye, H. (Ed.). (1983). Selye’s guide to stress research (Vol.
II). New York: Scientific and Academic Editions.
Shaffer, F. (2010). Biofeedback tutor (CD). Available
online at www.biosourcesoftware.com (also at www.aapb.
org).
Shealy, C. N. (1977). Ninety days to self-health. New York:
Dial Press.
Shellenberger, R., & Green, J. (1987). Specific effects and
biofeedback versus biofeedback-assisted self-regulation
training. Biofeedback and Self-Regulation, 12(3), 185–
209.
Shirakura, K., Saito, I., & Tsutsui, S. (Eds.). (1992). Current biofeedback research in Japan. Tokyo: Sginkoh Igaku
Shuppan.
Shtark, M. B., & Kall, R. (1998). Biofeedback-3: Theory and
practice. Novosibirsk, Russia: CERIS.
Shtark, M. B., & Schwartz, M. S. (2002). Biofeedback-4:
Theory and practice. Novosibirsk, Russia: CERIS.
Silva, J. (1977). Silva mind control method. New York: Simon
& Schuster.
Slattery, P., & Taub, E. (1976). Specificity of temperature
self-regulation to feedback loci. Biofeedback and SelfRegulation, 1, 316.
Slaughter, J., Hahn, W. W., & Rinaldi, P. (1970). Instrumental conditioning of heart rate in the curarized rat
with varied amounts of pretraining. Journal of Comparative and Physiological Psychology, 72, 356–359.
Smith, J. C. (1989). Relaxation dynamics. Champaign, IL:
Research Press.
Smith, J. C. (1990). Cognitive-behavioral relaxation training.
New York: Springer.
Smith, J. C. (2001). Advances in ABC relaxation: Applications and inventories. New York: Springer.
Smith, J. C. (2005). Relaxation, meditation, and mindfulness:
A mental health practitioner’s guide to new and traditional
approaches. New York: Springer.
Smith, O. C. (1934). Action potentials from single motor
units in voluntary contraction. American Journal of
Physiology, 108, 629–638.
Sokhadze, E. M., & Shtark, M. B. (1991). Scientific and
1. History and Definitions
clinical biofeedback in the USSR. Biofeedback and SelfRegulation, 16(3), 253–260.
Sterman, M. B. (1982). EEG biofeedback in the treatment
of epilepsy: An overview circa 1980. In L. W. White &
B. Tursky (Eds.), Clinical biofeedback: Efficacy and mechanisms (pp. 311–334). New York: Guilford Press.
Sterman, M. B., & Friar, L. (1972). Suppression of seizures
in epileptic following sensorimotor EEG feedback training. Electroencephalography and Clinical Neurophysiology,
33, 89–95.
Stoyva, J., & Kamiya, J. (1968). Electrophysiological studies of dreaming as the prototype of a new strategy in
the study of consciousness. Psychological Review, 75,
192–205.
Striefel, S. (1998). Creating the future for applied psychophysiology and biofeedback: From fantasy to reality.
Applied Psychophysiology and Biofeedback, 23, 93–106.
Stroebel, C. (1982). The quieting reflex. New York: Putnam.
Tarlar-Benlolo, L. (1978). The role of relaxation in biofeedback training: A critical review of the literature. Psychological Bulletin, 85, 727–755.
Tart, C. T. (Ed.). (1969). Altered states of consciousness: A
book of readings. New York: Wiley.
Taub, E. (2010). What psychology as a science owes Neal
Miller: The example of his biofeedback research. Biofeedback, 38(3), 108–117. (This article originated as
a presentation to the Claude Bernard Society, at the
annual meeting of the Association for Applied Psychophysiology and Biofeedback, in San Diego, CA, March,
26, 2010.)
Taub, E., Miller, N. E., Morris, R. L., & Robinson, W.
(1993). Anatomical specificity in thermal biofeedback
23
responding. Paper presented at the meeting of the Association for Applied Psychophysiology and Biofeedback,
New Orleans, LA.
Taub, E., & School, P. J. (1978). Some methodological considerations in thermal biofeedback. Behavioral Research
Methods and Instrumentation, 10, 617–622.
Thompson, L., & Thompson, M. (1998). Neurofeedback
combined with training in metacognitive strategies:
Effectiveness in student with ADD. Applied Psychophysiology and Biofeedback, 23(4), 243–263.
Thompson, M., & Thompson, L. (2003). The neurofeedback
book: An introduction to basic concepts in applied psychophysiology. Wheat Ridge, CO: Association for Applied
Psychophysiology.
Trowill, H. A. (1967). Instrumental conditioning of heart
rate in the curarized rat. Journal of Comparative and
Physiological Psychology, 63, 7–11.
Wickramasekera, I. E. (Ed.). (1976). Biofeedback, behavior
therapy and hypnosis: Potentiating the verbal control of
behavior for clinicians. Chicago: Nelson Hall.
Wickramasekera, I. E. (1988). Clinical behavioral medicine:
Some concepts and procedures. New York: Plenum.
Wolf, S. L., & Binder-MacLeod, S. A. (1983). Electromyographic biofeedback in the physical therapy clinic. In J.
V. Basmajian (Ed.), Biofeedback: Principles and practice
for clinicians (2nd ed., pp. 62–72). Baltimore, MD: Williams & Wilkins.
Wolpe, J. (1973). The practice of behavior therapy (2nd ed.).
New York: Pergamon Press.
Wyrwicka, W., & Sterman, M. B. (1968). Instrumental
conditioning of sensory motor cortex EEG spindles in
the waking cat. Physiology and Behavior, 3, 703–707.
Chapter 2
Entering the Field and Assuring Competence
Fredric Shaffer and Mark S. Schwartz
Biofeedback, neurofeedback, and applied psychophysiology constitute a multidisciplinary and
heterogeneous field of many professional disciplines and types of applications. Educational and
training opportunities in the field range from
courses at universities, webinars, and individual
workshops to comprehensive biofeedback training
programs. The Biofeedback Certification International Alliance (BCIA) provides accreditation
for programs that teach courses based on BCIA’s
Blueprints of Knowledge in biofeedback (BF),
neurofeedback (NF), heart rate variability biofeedback (HRVB), and pelvic muscle dysfunction
biofeedback (PMDB). Since BCIA only recognizes regionally accredited academic institutions,
the courses offered by universities are already
accredited through those regional boards; however, BCIA does monitor their course content to
assure it appropriately covers Blueprint content.
For many professionals, the sources of education
are the annual meetings, webinars, and workshops
of the Association for Applied Psychophysiology
and Biofeedback (AAPB), the Biofeedback Federation of Europe (BFE), and the International
Society for Neurofeedback and Research (ISNR);
BCIA clinical update and mentoring webinars;
and workshops sponsored by state and regional
societies. Additional sources include professional
organizations, societies dedicated to specific disorders, commercial training programs that offer
multiday programs, and self-paced distance educa-
tion; articles published in Applied Psychophysiology
and Biofeedback (AAPB), Biofeedback (AAPB),
and NeuroRegulation (ISNR); and mentoring from
experts in the field.
General Suggestions for Entering the Field
and Maintaining Competence
The development and maintenance of clinical competence require active participation in a
variety of educational and training experiences.
Responsible professionals seek continuing education and training. Mentors and others involved
with the education and training of professionals in
their setting have a responsibility to budget time
and financial resources for this type of continuing
education. To follow are some general suggestions
for ways to obtain and maintain competence. We
urge our readers to consider them seriously and to
try as many as are feasible.
Training and Continuing Education
1. Enroll in carefully selected workshops, commercial programs, and academic courses. BCIA’s
website (www.bcia.org) lists accredited programs
and regionally accredited universities that teach
its Blueprints of Knowledge. AAPB, BFE, ISNR,
state and regional biofeedback societies, and vendors list didactic and continuing education pro-
24
25
2. Entering the Field and Assuring Competence
grams on their websites. For feedback about the
quality of the presentations, ask sponsors and presenters for the names of those who have attended
in the past, and talk to them. Inquire about the
equipment used and impartiality toward other
manufacturers. Many of these courses are accredited by professional organizations like the American Psychological Association (APA).
2. Read recommended books, journal articles,
multimedia tutorials, manuals, AAPB, BFE, and
ISNR publications, and patient education booklets. ISNR offers an extensive online neurofeedback bibliography (www.isnr.net). Five indispensable publications are BCIA’s (2015) Professional
Standards and Ethical Principles; Khazan’s (2013)
The Clinical Handbook of Biofeedback: A StepBy-Step Guide for Training and Practice with
Mindfulness; Peper, Tylova, Gibney, Harvey, and
Combatalade’s (2008) Biofeedback Mastery: An
Experiential Teaching and Self-Training Manual;
Tan, Shaffer, Lyle, and Teo’s (2016) Evidence-Based
Practice in Biofeedback and Neurofeedback (3rd ed.);
and Thompson and Thompson’s (2015) The Neurofeedback Book: An Introduction to Basic Concepts
in Applied Psychophysiology (2nd ed.). A superb
BFE publication is Peper and Gibney’s (2000)
Healthy Computing with Muscle Biofeedback. Study
the references listed in BCIA’s Core Reading
Lists and Advanced Reading Lists. Furthermore,
review recorded presentations from webinars and
national meetings.
3. Study the BCIA Blueprints of Knowledge
and prepare for and attain BCIA certification.
BCIA offers certification for clinicians and technicians in biofeedback and neurofeedback, pelvic
muscle dysfunction biofeedback exclusively for clinicians, and a certificate of completion in heart
rate variability biofeedback for all applicants who
have completed appropriate didactic training.
4. Regularly read the principal journals in this
field (Applied Psychophysiology and Biofeedback,
and NeuroRegulation), the core magazine Biofeedback, and other journals that publish pertinent
articles. Subscribe to abstracting services such as
the National Center for Biotechnology Information (NCBI) at the U.S. National Library of Medicine (NLM).
Professional Networking
1. Learn from experienced professionals with
a track record of success. Review their patient
education documents and didactic handouts.
When feasible, consult with them to discuss treatment of selected patients and observe their clinical approaches. While face-to-face opportunities
may be limited in some geographic areas, many
BCIA Board-certified clinicians provide distance
mentoring within their region and internationally. This is especially important outside of North
America. BCIA certificants who provide mentoring services are listed on the BCIA website. Also,
in a number of professional listserves in diverse
biofeedback and neurofeedback specializations,
one can communicate with many experts in the
field.
2. Attend the annual meetings of AAPB, BFE,
ISNR, state/regional biofeedback societies, and/or
clinical organizations such as the American Academy of Pain Management (www.aapainmanage.
org) and the EEG & Clinical Neuroscience Society (www.ecnsweb.com). These meetings are the
best chance to attend a wide variety of symposia,
panels, and workshops. They also present an excellent chance to talk with professionals in this field.
These meetings are high caliber and attended by
many clinicians and researchers who are interesting, competent, academically sound, and encouraging. Many of the presentations are accredited by
organizations like the American Medical Association (AMA) and American Psychological Association (APA).
AAPB’s website: www.aapb.org.
BCIA’s website: www.bcia.org.
BFE’s website: www.bfe.org.
ISNR’s website: www.isnr.net.
3. Become involved in a state or regional biofeedback society. AAPB lists these societies at
www.aapb.org.
4. Invite highly credible and experienced professionals who are good therapists, educators, and/
or researchers to your professional setting. Institutions or other groups of professionals can cooperate to absorb the costs.
Practice Considerations
1. Beginners should usually limit the number
of biofeedback modalities they initially offer. For
example, clinicians might start with heart rate
variability (HRV), respiration (RESP), and skin
temperature (TEMP) biofeedback for the treatment of stress. These modalities are easier to teach
to clients, are widely used in stress management
26
protocols, and provide a relatively inexpensive
entry into biofeedback practice. New professionals should be encouraged to “test drive” the field
without having to “offer their first born.” After
they have mastered these modalities, they can
add neurofeedback and surface electromyographic
(SEMG) biofeedback, which are more expensive
and present a steeper learning curve. Beginners
should not try to master all the major modalities
because this approach often compromises a professional’s learning and unduly complicates assessment and therapy sessions.
2. If you plan to incorporate surface electromyographic (SEMG) biofeedback into your practice, learn surface muscle anatomy in order to
place sensors accurately. Study muscle kinesiology to help in interpreting SEMG measurements.
For neurofeedback, learn the International 10-20
System and the Modified Combinatorial System
for precise electrode placement. Review brain
anatomy and physiology, and the main electroencephalographic (EEG) generators, to understand
the significance of EEG values.
3. Perform tracking tests before recording data
to ensure that the biofeedback display mirrors the
client’s behavior. For example, masseter SEMG
amplitude should increase when a client clenches
his or her jaw. Know the typical range of values for
each biofeedback modality to identify false values
quickly. Use your own body as a benchmark. For
example, if a person’s resting heart rate is typically
65 beats per minute, a clinician should be suspicious of readings over 85 beats per minute. Whenever possible, examine raw waveforms to ensure
signal fidelity. Learn to identify and then remove
or minimize artifacts in physiological recordings.
Most importantly, acquire a healthy skepticism
about the accuracy of measurements.
4. Understand how to mitigate the risk of
infection through client and provider hand-washing, cleaning exposed surfaces and reusable sensors, and using disposable sensors.
5. Initially, limit the number of disorders for
which biofeedback and/or neurofeedback services
are offered. A clinician needs time to read the
clinical literature and develop expertise in assessment and treatment. Select more prevalent disorders that are of interest, that are likely to generate
referrals and be reimbursed by insurance and supported by efficacy research, and that are appropriate for one’s professional training and license.
Always work within your scope of practice as dictated by your state-issued health care license and
I. ORIENTATION TO BIOFEEDBACK
prior experience and expertise. Consult Tan et al.’s
(2016) Evidence-Based Practice in Biofeedback and
Neurofeedback (3rd ed.) for an overview of recent
outcome studies. If there are several competing
practices in the area, a clinician should offer services within his or her scope of practice and competence that are unique.
6. Once sufficient clinical expertise has been
developed, be prepared and willing to accept
patients with difficult problems, to invest more
time with these patients, and to adjust therapeutic
goals accordingly. Even some moderate improvement can be very satisfying to such patients and to
the referral source. Referral sources will probably
appreciate a practitioner’s willingness to accept
such patients.
7. Review sample assessment and therapy protocols from credible and experienced professionals.
Standardized assessment and therapy protocols
may be useful for some applications, but tailoring
assessment and treatment to individual patients
can also be successful and cost-effective. Practitioners can always adapt the protocols of other therapists to fit their own needs, preferences, and situations. Collaboration and professional networking
promote clinical competence.
Technical Considerations
1. Be familiar with a few equipment manufacturers before purchasing instruments, and discuss
instrumentation with professionals experienced
with different manufacturers and models. The
AAPB, BFE, and ISNR annual meetings, state
and regional biofeedback meetings, and some
professional society meetings can provide such
exposure to new equipment. Explain your needs
and ask for a demonstration before purchasing
equipment. A number of independent distributors
sell instruments from several manufacturers. Shop
around and get good advice about what will best
meet the anticipated needs of the clinical setting
and be most cost effective. Manufacturers’ online
forums can provide valuable information about
product performance and manufacturer support
from professionals who own the products under
consideration.
2. Since most professionals require training
and regular technical support to make full use
of their systems, select vendors who excel in customer support. This is one of the most overlooked
issues when purchasing equipment. Check with a
prospective vendor’s customers about the speed,
27
2. Entering the Field and Assuring Competence
supportiveness, and effectiveness of their technical support. Since even hardware from respected
manufacturers may malfunction during normal
use, choose vendors and manufacturers who will
quickly repair or replace the equipment. This will
reduce “down time,” thus minimizing clients’ or
research participants’ cancellations.
3. EMG instrumentation that allows simulta-
neous recording from at least two sites is critical
for evaluation and therapy. Likewise, neurofeedback providers increasingly use the quantitative
EEG (QEEG), which analyzes and topographically displays the EEG power spectrum of signals
recorded from multiple electrodes, to guide treatment planning and enrich ongoing assessment.
4. Develop a good working relationship with
the vendor and/or manufacturer to quickly resolve
problems and take advantage of the system’s features.
Mentoring
1. BCIA has moved from a supervision model
to a mentoring model. While supervisors provide
comprehensive oversight of their supervisees’
actions and assume legal liability for patient care,
mentors function as educators and are not liable
for their mentees’ actions. An unlicensed provider
may work with both a supervisor and a mentor; the
supervisor takes responsibility for client care and
the mentor provides biofeedback guidance.
2. Mentored providers or technicians should
attain the basic training and knowledge base that
qualify them for BCIA certification.
3. Mentors and their mentees should maintain
close and frequent communications about patients
and services. It is important that mentoring be
provided by professionals with sufficient biofeedback and/or neurofeedback experience and expertise to match a mentee’s clinical background and
needs.
4. Competent use of biofeedback and neurofeedback in a clinical setting begins with a fundamental understanding of the symptoms and
conditions to be treated. Recognizing the limits
of one’s expertise and collaborating with other
health care professionals involved in a case can
help ensure treatment success. Interprofessional
communication should be clear and reliable, and
include interpretations of psychophysiological and
clinical data that are accurate, scientifically valid,
and pertinent to the present problem.
5. Mentors should maintain their own proficiency by regularly treating patients with biofeedback, neurofeedback, and/or other applied psychophysiological therapies. There is no substitute for
this type of direct experience.
Education and Training Programs
Selecting courses, workshops, and training programs is often difficult. One source of information about training programs is the BCIA website
(www.bcia.org). BCIA provides separate listings
with contact information for colleges/universities
offering biofeedback coursework, didactic education/accredited trainers, and distance education
programs. Remember that BCIA only accepts
accredited training to satisfy its certification
requirements. When you evaluate educational and
training programs, the following steps are essential:
1. Assess
the presenter’s experience. This
includes the number of years using biofeedback, neurofeedback, and other applied psychophysiological therapies; the percentage of
time devoted to providing these services; and
the number of patients treated. Evaluate the
presenter’s qualifications and experience to
teach about the specific topic, as well as the
number of workshops, courses, and other presentations that he or she has provided. Finally,
consider the presenter’s reputation as a clinical practitioner and potential conflicts of
interest, which should be listed in a disclosure
statement.
2. Select a program sponsored or accredited by a
credible organization, such as the American
Psychological Association (APA) or the Biofeedback Certification International Alliance
(BCIA).
3. Consider comments by previous attendees of
the specific education and training program.
4. Check the time available for the topics listed
in the program. A minimum of 1 hour is often
necessary to cover even very specific topics.
Half-day and full-day workshops are often
required to explore topics thoroughly. It is
desirable for presenters to know the needs and
preferences of enrollees a few weeks ahead of
time.
5. Ask about the meaning of the term “handson experience.” Will the presenter observe
enrollees preparing a subject, attaching
28
electrodes and thermistors, adjusting the
instruments, and providing a few minutes of
physiological monitoring and biofeedback?
Someone who wants to observe and briefly
become familiar with an instrument does not
need much hands-on experience. However, if
someone wants to learn more about using the
instruments in assessment and therapy, then
more time with the instruments is preferable
and necessary.
6. Confirm that the presenter has clearly identified specific goals that are relevant, and that
the level is appropriate for one’s experience.
7. Verify which instrumentation will be available for demonstration or use. For vendorsponsored courses, identify which brands of
equipment the vendor is authorized to sell.
8. Ask about time for audience questions and
discussion.
9. Consider the cost–benefit ratio of sponsoring or attending a continuing education
program. Experienced and talented professionals deserve and have the right to expect
reasonable compensation for their educational services. Promotional materials, space,
administrative factors, transportation for the
presenter, and daily expenses all add to the
cost. It is also necessary to consider preparation time, even if the instructor has presented
the same or similar content before.
10. Check whether the instructor has earned
BCIA Board certification. While this is not
necessary, it is one piece of useful information about the presenter’s broader knowledge,
skills, and involvement in the field.
11. Some manufacturers offer training on their
equipment. Check with the manufacturer of
the equipment and then ascertain the qualifications of the presenters in the applications of
interest.
University-based educational opportunities are
available at various regionally accredited universities and colleges, at both the undergraduate and
graduate levels. Examples of such institutions and
their opportunities include the following:
• The graduate program in Clinical Psychology at Brigham Young University offers speciality
training in clinical health psychology and biofeedback as part of its APA-approved clinical psychology program. Students receive both didactic and
hands-on training as part of their doctoral studies.
I. ORIENTATION TO BIOFEEDBACK
• California School of Professional Psychology
at Alliant University in San Diego, California,
offers training in clinical psychophysiology and
biofeedback as a part of its APA-approved Clinical
Psychology PhD program Health Track. Students
get both didactic and hands-on training as a part
of their doctoral studies.
• East Carolina University in Greenville,
North Carolina, was the first to offer a university-based distance education course in biofeedback and now offers courses both on campus and
through a global classroom.
• San Francisco State University’s Biofeedback
and Self-Regulation Laboratory and academic
program offers undergraduates a minor and a certificate in holistic health. The program includes
a semester course on biofeedback, a 4-hour-perweek class, which consists of a 2-hour lecture and
2-hour laboratory practicum; a semester course on
the basic theory and technique of autogenic training, and a self-generating therapeutic approach
for clinical and nonclinical applications; and an
independent study for biofeedback research and
mentoring. These classes are also available to nonmatriculating students through the San Francisco
State University College of Extended Learning.
• Saybrook University offers Basic and Intermediate training and education in biofeedback
and an Advanced Biofeedback/Neurofeedback
Practicum course taught by BCIA-certified faculty
as part of its master’s and doctoral programs in the
College of Integrative Medicine and Health Sciences. These courses combine residential and distance learning components. The same college also
offers a 10-credit Certificate in Biofeedback and
Neurofeedback. Saybrook’s Clinical Psychology
degree program offers a PhD in Psychology, with
a Psychophysiology Specialization, that combines
two required “in person” training sessions each
year with distance education.
• Sonoma State University in Sonoma, California, has offered a unique upper-division biofeedback professional training sequence in the
Psychology Department, open to all university
students since 1981. This sequence teaches the
BCIA Blueprint and comprises three classes: Biofeedback and Somatics, Biofeedback Practicum,
and Biofeedback Internship.
• Southwest College of Naturopathic Medicine
in Tempe, Arizona, has offered formal elective
courses in biofeedback and neurofeedback since
29
2. Entering the Field and Assuring Competence
2005. Both courses cover the BCIA Blueprint and
are offered once per year as an elective.
• Truman State University in Kirksville, Missouri, offers an undergraduate, 3-hour Applied
Psychophysiology course and a 1-hour Research
Practicum in its Center for Applied Psychophysiology that may be a part of the bachelor’s degree
in psychology. Truman’s undergraduate research
team has explored questions in applied psychophysiology since 1977 and annually presents its
studies at national and international meetings.
• Widener University’s Institute of Graduate
Psychology, in Chester, Pennsylvania, teaches
biofeedback to about 10 graduate students and 8
online nonmatriculated students each semester,
and offers clinical supervision and BCIA mentoring.
BCIA anticipates the development of new programs to keep up with the need to educate clinicians. Check BCIA’s website (www.bcia.org) for a
current listing of university programs. While many
universities may include an introduction to biofeedback as part of a more comprehensive course,
it is important to look for coursework that is specific to biofeedback or neurofeedback and based
on BCIA’s Blueprints of Knowledge.
Certification of Biofeedback Professionals
Rationale
The primary reason for certification is to set uniform educational and training standards to use
biofeedback and neurofeedback competently in
clinical, optimal performance, and/or research
practice. Biofeedback and neurofeedback are
modalities that do not belong to a single profession.
BCIA was created in 1981 to set minimal standards for service providers with varied training
experience in diverse professions. At that time, as
now, no profession had identified specific training and experience criteria for biofeedback and
neurofeedback providers. BCIA is the oldest and
most widely recognized international board for
the certification of biofeedback and neurofeedback professionals. BCIA’s mission is indispensable: “BCIA certifies individuals who meet education and training standards in biofeedback and
progressively recertifies those who advance their
knowledge through continuing education” (www.
bcia.org).
Attaining BCIA certification has many advantages for providers of biofeedback and neurofeedback services (e.g., researchers, practitioners, technicians, and presenters of biofeedback educational
and training programs). Certification is valuable
for both mentors and mentees, although there
are certainly competent practitioners who are not
BCIA Board certified. Certification is not a guarantee of competence, and it was never intended to
guarantee a full range of competencies. However,
certification provides a useful index of fundamental
knowledge and basic instrumentation proficiency.
There are ten compelling reasons why most
practitioners using biofeedback and neurofeedback
should seriously consider attaining and maintaining BCIA certification:
1. Certification reflects involvement in this field
and increases professional credibility.
2. It attests that the certified individual has met
specified criteria to use biofeedback and/or
neurofeedback competently.
3. It increases “market value” and mobility for
many.
4. It gives employers a credible index of training.
Many employers require or give preference to
applicants who have earned a BCIA certification.
5. Some reimbursement systems view the BCIA
credential as an important criterion for reimbursement.
6. BCIA advocates for its certificants and the
field when interest groups attempt to restrict
who may offer biofeedback and neurofeedback
services; and explains the field and BCIA’s
(2015) Professional Standards and Ethical Principles to state health care regulators and insurers.
7. BCIA provides free “Find a Practitioner” and
job board listings.
8. BCIA awards scholarships for neurofeedback (Eugenia B. Davis and Celeste DeBease
Scholarships) and biofeedback (Francine Butler), which waive all certification application
fees, to promote the entry of young professionals into the field.
9. A credible certification program is a cornerstone and important sign of the maturation of
the field. It improves the image of biofeedback
to health care professionals, referral sources,
and others outside the field. We should not
undervalue its importance.
10. Preparing for certification and maintaining
it through continuing education (CE) and
30
I. ORIENTATION TO BIOFEEDBACK
recertification every 3 or 4 years involve considerable studying and learning—a benefit for
applicants, certificants, and clients/patients.
BCIA identifies open access articles that certificants may read and take brief tests over for
inexpensive CE credit. Also, BCIA’s clinical
update and mentoring webinars provide additional CE opportunities.
The certification process acts to deter the
least competent practitioners and is an incentive
for increasing competence. Epstein and Hundert
(2002) defined competence as “the habitual and
judicious use of communication, knowledge, technical skills, clinical reasoning, emotions, values,
and reflection in daily practice for the benefit
of the individual and community being served”
(p. 226). It is an objective and acceptable criterion
for would-be practitioners to assess their entrylevel competence. Since biofeedback and neurofeedback are unregulated fields, certification helps
both the public and professionals select qualified
providers.
BCIA Certification Programs
BCIA now offers biofeedback, neurofeedback, and
pelvic muscle dysfunction biofeedback (PMDB)
certification for clinicians and technicians. For
biofeedback, the requirements for the clinical
track include an earned degree in a health carerelated field, anatomy and physiology courses,
didactic training in a core curriculum, mentored
self-regulation training, clinical biofeedback experience (which includes SEMG, electrodermal, and
thermal modalities, and may also encompass an
introduction to EEG, heart rate variability [HRV],
and respiration), and biofeedback case conferences. The applicant must also successfully complete a psychometrically validated examination on
the materials covered in the Blueprints of Knowledge, which provides a detailed outline of information needed to enter the biofeedback field and
prepare for the examinations.
BCIA added certification in neurofeedback in
1997 and in pelvic muscle dysfunction biofeedback in 2004. Each of these certification programs
involves a corresponding health care-related
degree, training, and experience requirements, a
detailed set of Blueprint Knowledge Statements,
a core reading list, and agreement to adhere to
BCIA’s Professional Standards and Ethical Principles.
All certificants require a license for independent
practice when treating a medical and/or psycho-
logical disorder. They should read applicable state
Practice Acts and understand how they regulate
their advertising and provision of services. The
BCIA website (www.bcia.org) provides a complete
description of the certification process, certification requirements, and helpful resources.
Briefly, the Blueprint of Knowledge areas for
biofeedback certification (2015 revision) include
the following:
I. Orientation to Biofeedback
II. Stress, Coping, and Illness
III. Psychophysiological Recording
IV. Research Methodology
V. Surface EMG (SEMG) Applications
VI. Autonomic Nervous System (ANS) Applica-
tions
VII. Respiratory Applications
VIII. Intervention Strategies
IX. Professional Conduct
The Blueprint of Knowledge areas for neurofeedback certification (2014 revision) include the
following:
I. Orientation to Neurofeedback
II. Basic Neurophysiology and Neuroanatomy
III. Instrumentation and Electronics
IV. Research Evidence Base for Neurofeedback
V. Psychopharmacological Considerations
VI. Patient/Client Assessment
VII. Developing Treatment Protocols
VIII. Treatment Implementation
IX. Current Trends in Neurofeedback
X. Professional Conduct
The Blueprint of Knowledge areas for PMDB
certification (2011 revision) include the following:
I. Applied Psychophysiology and Biofeedback
II. Pelvic Floor Anatomy, Surface EMG Assess-
ment of Pelvic Floor Musculature, and Clinical Practice Procedures
III. Clinical Disorders: Bladder Dysfunction
IV. Clinical Disorders: Bowel Dysfunction
V. Clinical Disorders: Chronic Pelvic Pain Syndromes
The Blueprint of Knowledge areas for the heart
rate variability certificate of completion (2015)
include the following:
I. Cardiac Anatomy and Physiology
II. Respiratory Anatomy and Physiology
31
2. Entering the Field and Assuring Competence
III. Autonomic Nervous System Anatomy and
Physiology
IV. Heart rate Variability
V. HRV Instrumentation
VI. HRV Measurements
VII. HRV Biofeedback Strategies
VIII. HRV Biofeedback Applications
Improved BCIA Examinations
BCIA task forces carefully evaluate and revise the
three certification exams on a regular basis to keep
them current and relevant. During the last examination revisions, the reading lists were simplified,
making them more compact, accessible, inexpensive, and closely tied to actual examination items.
The reading lists consist of current textbooks and
articles recommended by experienced educators.
Several of these articles may be downloaded for
free. Readers can find a current reading list and
access articles from the BCIA website (www.bcia.
org).
During regular examination evaluation and
revision, experts review each examination to
ensure that their questions adequately represent
key Blueprint concepts, are sourced to a suggested
reading list, and conform to the highest psychometric standards. They replace outdated examination questions with new ones contributed by
leaders in the field and validated by certificants.
While BCIA has maintained its high standards,
applicant performance on BCIA examinations has
improved over the last several years. More candidates than ever are passing with higher scores.
More than 85% of candidates pass on their first
attempt.
Certification by Prior Experience
As the field has grown and matured, the BCIA
Board discovered that many experts in the biofeedback and neurofeedback fields, who did not
become certified earlier in their careers, had
already surpassed BCIA entry-level certification requirements. BCIA now offers Certification by Prior Experience (CPE) in biofeedback,
neurofeedback, and PMDB. BCIA designed CPE
in biofeedback and neurofeedback for licensed
professionals who have completed a minimum
of 100 hours of continuing education and more
than 3,000 patient/client hours using biofeedback,
formal study of anatomy/physiology, and formal
mentoring. The PMDB CPE program requires a
minimum of 72 hours of continuing education
and 3,000 hours of patient care using SEMG for
elimination disorders. CPE makes BCIA certification more accessible to experienced professionals
who have exceeded the entry-level requirements
for traditional certification and are leaders in the
field. While the majority of these role models did
not financially need CPE, they attained it for “the
good of the field.”
BCIA Certification 2.0
Web-based technology has made BCIA certification and recertification increasingly accessible and
affordable:
1. Professionals can now enroll in distance learn-
ing courses to obtain didactic and continuing
education and to satisfy the anatomy and physiology requirement.
2. Distance mentoring allows professionals in
rural communities and foreign countries to
receive clinical training by BCIA-approved
mentors. Cutting-edge technology allows both
parties to see each other via webcams, to view
each other’s desktops, and allow a mentor to
demonstrate software features on the mentee’s
computer.
3. Web-based examination administration allows
applicants to take proctored examinations at
their local library or university at their convenience.
BCIA Certification Goes Global
BCIA has accomplished certificants in 32 countries that include Australia, Austria, Canada,
Japan, Mexico, the Netherlands, Norway, South
Africa, and Spain. BCIA has two international
affiliates, BCIA-Australia and BCIA-Mexico, who
assist BCIA in setting education and training
standards, reviewing applications, and providing
guidance for finding qualified mentors and legally
using biofeedback and neurofeedback within the
laws that regulate health care.
There are many reasons for the growth of international BCIA certification. First, several of the
best instructors have offered their didactic programs abroad and have subsequently provided
distance mentoring for their students. Second,
BCIA’s University Initiative has facilitated the
creation of the first BCIA-accredited neurofeedback curriculum offered entirely in French by
Vincent Paquette and Johanne Levesque at the
Institut of Neurofeedback du Quebec. Third,
32
I. ORIENTATION TO BIOFEEDBACK
BCIA has worked hard to eliminate obstacles
to international certification through distance
learning and distance mentoring, and by providing online access to continuing education and
testing. Finally, AAPB, BCIA, BFE, and ISNR
have enthusiastically reached out to international
biofeedback professionals and provided extensive
opportunities for collaboration.
currently members of AAPB, BFE, and/or ISNR
should join or rejoin. Those who are not certified
by BCIA should consider this credential for themselves. Those currently certified should continue
to stay certified. Those who were certified in the
past, but are not currently certified, should return.
Our field ensures its integrity when practitioners
are “more than qualified, BCIA Board certified!”
BCIA Recertification
Acknowledgments
BCIA has made recertification easier and more
affordable than it has ever been. Certificants can
now read online articles from Biofeedback, Applied
Psychophysiology and Biofeedback, NeuroRegulation,
and other open access journals and then complete
brief online examinations to earn continued education credit at minimal cost. They can receive
credit for scholarly activities such as professional
presentations, teaching, and writing. They can
also attend educational webinars offered by organizations such as the AAPB, BCIA, and ISNR in
the comfort of their home or office.
Summary
In this chapter we have provided ideas and suggestions for persons entering the biofeedback field. For
those already in the field, these ideas and suggestions may help them maintain and enhance their
competence. Biofeedback is a broad, heterogeneous, and complex field. Practitioners need infusions of new knowledge, ideas, and skills. Deciding
when and where these infusions are to take place
and determining who is to provide them are not
always easy. In this chapter, we have offered some
guidance.
AAPB, BCIA, BFE, and ISNR have become
international resources and centers for continued maturation of the field. Those who are not
We wish to thank Celeste DeBease, Aubrey Ewing,
Zachary Meehan, Brian Milstead, Donald Moss, Randy
Neblett, and Erik Peper for their invaluable contributions to this chapter. We especially want to recognize
the exceptional contributions of BCIA Executive Director Judy Crawford to this chapter and for her decades of
dedication to the field we love.
References
BCIA Board. (2015). Professional standards and ethical principles of biofeedback. Wheat Ridge, CO: BCIA.
Epstein, R. M., & Hundert, E. M. (2002). Defining and
assessing professional competence. Journal of the American Medical Association, 287, 226–235.
Khazan, I. Z. (2013). The clinical handbook of biofeedback:
A step-by-step guide for training and practice with mindfulness. Malden, MA: Wiley-Blackwell.
Peper, E., & Gibney, K. H. (2000). Healthy computing with
muscle biofeedback: A practical manual for preventing
repetitive motion injury. Woerden, The Netherlands:
BFE.
Peper, E., Tylova, H., Gibney, K. H., Harvey, R., & Combatalade, D. (2008). Biofeedback mastery: An experiential teaching and self-training manual. Wheat Ridge, CO:
AAPB.
Tan, G., Shaffer, F., Lyle, R. R., & Teo, I. (2016). Evidencebased practice in biofeedback and neurofeedback (3rd ed.).
Wheat Ridge, CO: AAPB.
Thompson, M., & Thompson, L. (2015). The neurofeedback
book: An introduction to basic concepts in applied psychophysiology (2nd ed.). Wheat Ridge, CO: AAPB.
P a r t II
Instrumentation
Chapter 3
A Primer of Traditional
Biofeedback Instrumentation
C. J. Peek
Monitoring Psychophysiological Arousal:
The Central Focus of Biofeedback
My aim in this chapter is to put into ordinary
language the basic technical matters of practical importance in biofeedback. Technical concepts are introduced through analogy or heuristic
description, such that they can become a usable
part of the reader’s biofeedback language. This
chapter contains many judgments on the practical
importance of things encountered in biofeedback,
and to that extent represents my own views on the
subject, especially in matters where no definitive
conceptual, empirical, or practical view holds sway
in the field.
This chapter is focused on basic electronic and
measurement concepts for EMG, temperature,
and electrodermal biofeedback. Electroencephalographic biofeedback is covered in Chapter 6 of this
volume. This chapter focuses on the “front end,” at
which electrodes, basic electronics, and feedback
modes interact with clinicians and clients. Other
chapters address the “back end,” where computers
and myriad forms of feedback and data recording
are devised for clinical biofeedback or research.
Although computerized biofeedback is common
and sophisticated, many biofeedback users still
work with freestanding biofeedback instruments
or simple devices, such as those used in this chapter as vehicles for illustrating basic electronic and
measurement principles. Moreover, users with
sophisticated computer systems still need to under-
A major application for biofeedback is to provide
tools for detecting and managing psychophysiological arousal. As health care fields matured,
by the early 1970s it became clear that frequent,
excessive, and sustained psychophysiological tension and overarousal cause or exacerbate many
health problems. Interest in detecting and managing these states intensified. By the same time,
improved biomedical electronics had made it practical to monitor previously “invisible” physiological processes associated with overarousal.
The natural combination of these developments
in health and technology found expression in the
new field of biofeedback, in which the languages
and concepts of psychology, physiology, and electronics freely intermingle. The terms “stress,”
“anticipation,” “autonomic arousal,” and “muscle
fibers” are found in the same sentences as “electromyography” (EMG), “microvolts,” “bandwidths,”
and “filters.” Such hybrid sentences usually contain at least some mystery to those (i.e., most of us)
who are not fluent in all these languages. Probably
the greatest mystery among biofeedback devotees
and beginners is the language of electronics. Of
the three languages spoken in biofeedback, this is
the least similar to ordinary language.
35
36
stand the fundamental principles and methods for
detecting and measuring physiological processes,
even at times when much may be automated and
invisible behind the computer screen.
This chapter is therefore created as a primer.
Its focus is practical rather than comprehensive. It
is simplified rather than highly technical. It stays
with basic principles and methods that are foundational to highly technological approaches rather
than delving into today’s advanced technology. It
is heuristically presented, with emphasis on principle, as well as fact. It also addresses basic instrument functions that may be more hidden to users
in this era of computers but remain important to
understand even now.
Correlates of Arousal: Three Physiological
Processes of Interest in Biofeedback
Three physiological processes commonly associated with overarousal are skeletal muscle tension,
peripheral vasoconstriction, and electrodermal
activity. These three, especially the first two, are
the most common biofeedback modalities. This
is no surprise, as these processes have been recognized all along as intimately involved in anger,
fear, excitement, and arousal.
This association can be seen by recalling common expressions or idioms found in everyday language. For example, when a person is said to be
“braced” for an onslaught, one gets a picture of
muscles “at the ready.” The person is tense and
may have fists “clenched” and jaw “set”; in a word,
the person is “uptight.” If this tension were unrealistic or simply habitual, commonplace advice
would be to loosen up,” “relax,” or “let go.”
The expression “my blood ran cold” evokes the
connection between fear and cold extremities, as
does the image “cold hands, warm heart.” In both is
the recognition that having cold hands is a sign of
emotional responsivity—in other words, the common knowledge that peripheral vasoconstriction is
a sign of arousal. In referring to electrodermal activity, a person might illustrate fear with the image of
“a cold sweat” or of “sweating bullets.” A picture of
calm and ease is drawn by the term “no sweat.”
As these idioms illustrate, people already know
that muscle tension, peripheral vasoconstriction,
and electrodermal activity are related to arousal.
The systematic study and modification of these
processes are in the domain of biofeedback. Biofeedback devices exist to aid in the study and especially in the modification of these processes.
II. INSTRUMENTATION
Biofeedback Equipment
Terminology
A piece of biofeedback hardware may be referred to
as “instrument,” “machine,” “device,” “equipment,”
“apparatus,” “unit,” and even “gadget” or “gizmo.”
Most of these terms are used interchangeably and
with little or no uniformity or consistency; often
the choice is based simply on preference or whim.
This is not offered as a criticism, for people often
have many terms for things that are interesting
to them. It may simply be a case of the ancient
Chinese proverb, “A child who is loved has many
names.”
In any case, it is worthwhile to outline the connotations for the more popular terms for biofeedback hardware. “Instrument” is the most formal
of the terms, denoting a measuring device for
determining the value of a quantity under observation. Many items of biofeedback hardware do
not qualify as “instruments” under this definition, since actual measurements are not being
made; only changes or relative magnitudes are
being monitored. For example, “mood rings” and
other simple biofeedback “gadgets” or “gizmos” are
not considered instruments. The terms “apparatus,” “equipment,” and “device” leave unspecified
whether measurement is made, and hence are
safe general terms, although “device” implies the
performance of a highly specific function. The
term “unit” is even more neutral, claiming nothing more than reference to an entity. The term
“machine” denotes a mechanism that transmits
forces, action, or energy in a predetermined manner. Those familiar with electronics see electronic
equipment abstractly transmitting forces, motion,
and energy within their circuits, and hence often
use the term “machine” in describing biofeedback
equipment.
In this chapter, most of these terms are used,
and (as in common practice) they are used more
or less interchangeably. Nothing beyond the ordinary meanings and connotations is intended.
What Biofeedback Instruments Are
Supposed to Do
A biofeedback instrument has three tasks:
1. To monitor (in some way) a physiological pro-
cess of interest.
2. To measure (objectify) what is monitored.
3. To present what is monitored or measured as
meaningful information.
37
3. A Primer of Biofeedback Instrumentation
The following sections outline how access is
gained to three important psychophysiological
processes in biofeedback.
EMG: An Electrical Correlate of Muscle Contraction
A biofeedback device cannot measure muscle contraction in a simple, direct way. When a muscle
contracts, it tries to pull its two anchor points
together; this is what is meant by “muscle contraction.” It is a kinetic phenomenon involving force
and sometimes movement. Practically speaking,
this is not easily monitored. One cannot insert a
strain gauge between one end of a muscle and its
anchor point to measure grams of pull. (Force and
movement gauges, called “goniometers,” are used
as muscle contraction monitors in physical medicine applications, but these are not sensitive to the
levels or locations of muscle contraction involved
in relaxation and low-arousal applications of biofeedback.)
Because muscle contraction itself is inaccessible, some aspect or correlate of it will have to
do. Biofeedback exploits the electrical aspect of
muscle contraction. Muscle contraction results
from the more or less synchronous contraction of
the many muscle fibers that constitute a muscle.
Muscle fibers are actuated by electrical signals
carried by cells called “motor units,” and muscle
contraction corresponds to the aggregate electrical activity in these muscle fibers. This electrical
activity can be sensed with fine wire or needle
electrodes that actually penetrate the skin above
the muscle. More commonly, it is sensed with surface electrodes that contact the skin above the
muscle, where there exist weakened electrical signals from muscle fibers beneath the skin. This is
the preferred biofeedback method for monitoring
muscle contraction, because it is practical and corresponds well to actual muscle contraction. Note
that this electrical method, surface EMG, does not
directly monitor muscle contraction, but monitors
an electrical aspect of muscle contraction that
bears a more or less regular relationship to muscle
contraction.
The important point is this: Surface EMG
(hereafter referred to simply as EMG) is the preferred method for monitoring muscle contraction,
but it does not directly measure muscle contraction. An EMG device does not read out in units of
force or movement, such as grams or millimeters.
Instead, it measures an electrical correlate of muscle contraction and reads out in electrical units
(microvolts—millionths of a volt). This is because
it is making an electrical, not a kinetic, measurement. This explains the initial puzzlement that
often comes over the biofeedback novice upon
learning that muscle contraction is measured in
volts—electrical units that, at face value, seem to
have little to do with muscle contraction.
Peripheral Temperature: A Correlate
of Peripheral Vasoconstriction
A biofeedback device cannot directly measure
the changing diameter of peripheral blood vessels or the smooth muscle activity that brings
about these changes. Therefore, some correlate of
vascular dilation will have to do. Dilated vessels
pass more warm blood than constricted vessels do.
Therefore, surrounding tissue tends to warm and
cool as vascular diameter increases and decreases,
providing a good correlate of vascular diameter.
This effect is most pronounced in the extremities
(especially the fingers and toes), where changes
in vascular diameter are pronounced, and where
the relatively small amount of surrounding tissue
warms and cools rapidly in response to changes in
the blood supply.
Here again the physiological process of interest
(peripheral vasoconstriction) is inaccessible, but
an accessible correlate (peripheral temperature)
is a practical indicator. Biofeedback devices typically read out in degrees Fahrenheit as the indirect indicator of peripheral vasoconstriction. This
shows that only indirect access to peripheral vasoconstriction is possible in biofeedback.
Finger Phototransmission: Another Correlate
of Peripheral Vasoconstriction
A second indirect way to gain access to peripheral
vasoconstriction takes advantage of the fact that a
finger or toe having less blood in its vessels allows
more light to pass through than an extremity with
more blood. That is, pale skin passes light more
readily than infused skin. A small light is shone
through the flesh of a finger and is reflected off the
bone back to a light sensor. The variation in light
intensity at the sensor, and the resulting electrical
signal, indicate variation in blood volume.
This device is commonly called a “photoplethysmograph” and is sometimes used in biofeedback. It monitors pulse, and (with appropriate
circuitry to average out the pulses) can give an
indication of relative blood volume, another correlate of vasoconstriction. Such devices read out
only in relative units. That is, they read changes
38
but are not anchored to some outside standard
reference point. Photoplethysmography is not
employed nearly as often as peripheral temperature to indicate peripheral vasoconstriction.
Further description of photoplethysmography is
beyond the scope of this chapter. For more information, see Jennings, Tahmoush, and Redmond
(1980). For more on the psychophysiology of
peripheral blood flow, see Cacioppo, Tassinary,
and Berntson (2000).
Skin Conductance Activity: A Correlate of Sweat
Gland Activity
Sweat gland activity is another physiological process that is not directly accessible. One cannot tell
whether a sweat gland is “on,” how much sweat
is being secreted, or how many such glands are
active. However, sweat contains electrically conductive salts that make sweaty skin more conductive to electricity than dry skin. Hence skin conductance activity (SCA) corresponds well to sweat
gland activity. SCA, along with other electrical
phenomena of the skin, is known as electrodermal activity (EDA); historically, it has also been
known as “galvanic skin response” (GSR). A skin
conductance device applies a very small electrical pressure (voltage) to the skin, typically on the
volar surface of the fingers or the palmar surface
of the hand (where there are many sweat glands),
and measures the amount of electrical current that
the skin will allow to pass. The magnitude of this
current is an indication of skin sweatiness and is
read out in units of electrical conductance called
“microsiemens,” formerly referred to as “micromhos.”
Here again, an electrical unit (conductance)
serves as the indirect measure of a physiological
phenomenon (sweat gland activity). This explains
what might initially seem odd: that EDA is measured in electrical units that at face value have
nothing to do with sweat gland activity.
Objectification and Measurement
As described earlier, direct monitoring of muscle
contraction, peripheral vasoconstriction, and
sweat gland activity is not feasible. Therefore, biofeedback devices gain access indirectly through
monitoring more accessible correlates of these
physiological processes. This means that a biofeedback reading should be taken as a convenient
indication of a physiological process but should
II. INSTRUMENTATION
be understood as separate from the physiological
process itself. Practitioners must distinguish the
physiological process beneath the skin from the
instrumentation schemes outside the skin used to
gain access to it. This distinction is important for
understanding measurement, objectification, artifact, and the interpretation of biofeedback data.
To compare a person’s biofeedback readings from
one occasion to another, or to compare readings
between different individuals, an objective scale
permitting such measurement is advantageous.
“Measurement” takes place when the device is calibrated to and reads out in standardized quantitative units that show how the monitored process is
varying. An example is a thermometer calibrated
to the Fahrenheit temperature scale.
On the other hand, what I call “indication of
relative magnitude” takes place when an observable signal such as a meter reading is made to
correspond to a particular process (e.g., skin temperature or muscle contraction) but the correspondence is not displayed in standardized quantitative
units. An example would be a homemade thermometer that reads out on an arbitrary “warm–
cool” scale of 1–10.
The advantage of measurement is that different
observers can make direct quantitative comparisons. Without measurement, observers can only
compare relative magnitude or change. Measurement tends to increase replicability of procedures
and comparability of results. Measurement in this
sense is often not possible in biofeedback, due to
the lack of clearly defined and/or widely accepted
standardized scales for measurement.
For example, EMG devices typically have meters
or scaled outputs that give readings in microvolts.
These numbers appears to give objectivity to the
readings and to permit actual measurement. In
fact, however, there is no widely accepted and standardized scale for EMG microvolts. Consequently,
different equipment gives different readings for the
very same degree of muscle contraction. Therefore, EMG readings can be compared only when
the same (or very similarly designed) equipment
is used for all the readings, including methods of
digital processing (Bolek, 2013). Explanations for
this will become clear later, when I describe the
design of the EMG device. The important point
to remember now is that EMG readings are better thought of as indications of relative magnitude
than as measurements. The same is true for skin
conductance readings. Skin temperature readings,
however, are measurements, as long as the temper-
39
3. A Primer of Biofeedback Instrumentation
ature device is properly calibrated to a temperature
scale such as degrees Fahrenheit.
Operation of the EMG Instrument
The EMG instrument picks up weak electrical signals generated during muscle action. Each muscle
consists of many muscle fibers, with “motor neurons” electrically connected to higher levels of the
nervous system. Muscle contraction occurs when
these motor neurons carry electrical activating
signals to the muscle fibers. A small part of this
electrical energy leaves the muscle and migrates
through surrounding tissue. Some of this energy
becomes available for monitoring at the surface of
the skin. The tasks of an EMG machine are as follows:
1. To receive the very small amount of electrical
energy from the skin.
2. To separate EMG energy from other extrane-
ous energy on the skin, and to greatly magnify
the EMG energy.
3. To convert this amplified EMG energy into
information or feedback that is meaningful to
the user.
Receiving EMG Energy from the Skin: Electrodes
Surface electrodes and the wires attached to them
complete the electrical pathways from the skin
to the EMG machine. Electrodes come in many
forms. Snap-on one-use disposable electrodes are
common today, while older equipment often used
reusable electrodes (usually small metallic discs
mounted on plastic or rubber) attached with tape
or double-stick adhesive washers. Other electrodes
come on strips or on a headband for simultaneous application of the three electrodes generally
required for EMG biofeedback. Some electrodes
are permanently attached to electrode cables,
whereas others are made to snap onto the cable,
permitting changes of electrodes without changing the cable.
Many electrodes are made of simple materials,
such as nickel-plated brass or stainless steel; others
are made of rare materials, such as gold or silver
chloride over silver (silver/silver chloride). The
precious metal electrodes have historically been
the materials of choice for physiological monitoring, because the materials do not interact significantly with skin or other substances with which
they are in contact. However, the simpler and
cheaper electrodes have been found to be quite
satisfactory for biofeedback EMG applications and
are now in widespread use. Modern equipment can
well tolerate imperfect electrodes and skin preparation, and thus has greatly reduced the need for
precious metal electrodes for EMG biofeedback.
Electrode Cream or Gel
Some EMG electrodes are made for use with an
electrode gel or cream. This conductive substance
flows into the irregularities of the skin and the
electrode, establishing a stable and highly conductive connection between them (see Figure
3.1). Modern equipment is often very resistant to
artifact from differences in electrode contact and
employs flat electrodes that do not require electrode gel. Electrode gel was more common in the
past when cupped, reusable electrodes were in
more common use.
Skin Preparation
A standard part of electrode application is to
remove skin oil, makeup, or dead skin cells,
because these impede the travel of bioelectric signals from the skin to the electrode. Most manufacturers suggest using alcohol wipes for this purpose. In the past, some manufacturers suggested
using an abrasive skin cleaner, although modern
equipment and technology has made the use of the
alcohol swab sufficient and far more common. The
risk in underpreparing the skin is that EMG readings will be erroneously high if the skin remains
very oily or covered with makeup. The risk of
overpreparing the skin, particularly with the abrasive compounds, is that the skin (and client) will
become irritated. There is nothing to be gained by
actually scrubbing skin unless there is significant
dirt, oil, or makeup present. As said earlier, some
equipment operates with simple metal electrodes
requiring little or no electrode gel or elaborate skin
FIGURE 3.1. EMG electrode and gel.
40
II. INSTRUMENTATION
cleaning. This is not to say, however, that one can
just forget about skin preparation.
Separating EMG Energy from Extraneous
(“Noise”) Energy
“Noise” is the general term for unwanted or extraneous signals. In EMG machines, there are two
kinds of noise: electrical interference and internally generated noise.
Electrical Interference and the Differential Amplifier
The environment is continuously saturated with
electrical energy transmitted through space from
power lines, motors, lights, and electrical equipment. Human bodies and EMG electrodes pick
up this energy. The EMG apparatus receives these
unwanted electrical noise signals in addition to
the desired bioelectric signals from the muscles.
The EMG unit must therefore find a way to reject
the noise so that only EMG signals remain.
Interference is rejected in an ingenious way,
using an electrical subtraction process in a “differential amplifier.” The electrodes establish three
independent pathways from an area of the skin to
the EMG instrument. One pathway, called the
“reference,” is used by the instrument as a point of
reference from which the minute electrical pressure (voltage) exerted from the other two “active”
electrodes is gauged. (Remember that any electrical pressure or voltage measurement is defined as a
pressure difference between one point and another
point. There is no such thing as a voltage measurement without respect to some second point of reference.) This results in two “sources” feeding the
instrument, each using the reference electrode as
the point of reference (see Figure 3.2). Note that
the reference electrode can be placed nearly anywhere on the body, but it is shown in Figure 3.2
between the two active electrodes for the sake of
illustration, and because it is a common arrangement.
The differential amplifier requires these two
sources in order to separate the EMG energy
from the extraneous energy. To see why, we must
remember that this extraneous energy is the hum
or noise transmitted through space from power
lines, motors, and appliances that is picked up
by the body acting as an antenna. Most of this
extraneous electrical noise energy rises and falls
rhythmically at 60 cycles per second. At any given
moment, this energy is in exactly the same place
in its rhythm (“in phase”) at any point on the body
and at any point that an electrode can be placed.
Hence it is possible for the differential amplifier to
continuously subtract the voltage at source 1 from
that at source 2. This cancels the noise voltage.
Only slightly simplified, this is illustrated graphically in Figure 3.3; it is assumed that the muscle is
at rest and giving off no EMG signals. The following steps explain Figure 3.4:
1. Electrical interference is received by the body
acting as an antenna.
2. The interference is in the same place in its
rhythm for both active electrodes.
3. Therefore, the active inputs (from source 1
and source 2) of the differential amplifier “see”
exactly the same interference signal at any
given moment (interference is in the “common
mode”).
4. Because the output of the differential amplifier
is proportional to the difference between the
signals at its two active inputs (from sources 1
and 2).
5. And the interference signals are always identical (restatement of point 3).
6. Therefore, the output of the differential amplifier is zero for electrical interference.
FIGURE 3.2. Active and reference EMG electrodes.
3. A Primer of Biofeedback Instrumentation
41
FIGURE 3.3. Differential amplifier eliminating the electrical interference picked up by the body acting as an
antenna.
But What about EMG Signals?
Suppose that motor neurons now signal the resting muscle to contract. Each electrode receives signals most strongly from the area of muscle directly
beneath it. Because electrodes are spaced along
the muscle, they each receive a different pattern
of EMG signals. Here is an analogy: If two microphones were placed in a room full of speaking
people, each one would pick up a different pattern
of sounds, even if the overall loudness of sound in
each microphone were the same. Similarly, at any
given moment, the electrodes feed “differential”
EMG signals superimposed on the previously discussed identical “common-mode” signals.
As the differential amplifier continuously subtracts the signal at source 2 from that at source 1
(thus amplifying only differences between them),
the common-mode noise signals will be canceled,
while the differential EMG signals will always leave
a remainder to be amplified and ultimately displayed
on a meter. The operation of the differential EMG
amplifier with the desired EMG signals is shown
graphically in Figure 3.4 and is summarized below.
1. Different EMG signals arrive at the two
e­lectrodes as the muscle beneath them contracts.
2. Therefore, sources 1 and 2 feed “differential”
EMG signals to the inputs of the differential
amplifier.
3. At the same time, identical (“common-mode”)
interference signals are superimposed on the
differential EMG signals.
4. Thus, the inputs (from sources 1 and 2) “see”
composite signals that have an identical component (common-mode noise) and a differing
component (differential EMG signals).
5. Because the output of the differential amplifier is proportional to the difference between
the signals at its two inputs (from sources 1
and 2).
6. A portion of the signals are identical (common-mode) and a portion are different (differential-mode) (restatement of point 4).
7. So the output of the differential amplifier is
zero for common-mode interference and high
for differential-mode EMG signals.
FIGURE 3.4. Differential amplifier “canceling” the common-mode interference while amplifying differential
EMG signals.
42
II. INSTRUMENTATION
The Chemist’s Balance Analogy
The operation of the differential amplifier can be
illustrated by another analogy. Imagine a sensitive
chemist’s balance scale, with its two pans, center
fulcrum, and a set of weights. With no weights in
the pans, the scale balances. With equal weights
in the pans, it also balances. Even if we stretch
our imaginations to envision the weights constantly changing (but always remaining equal in
both pans), the scale will remain balanced. However, a fly landing on one pan during this process
will upset the balance, and the pointer will move
off center. Moreover, if two flies of equal weight
hop up and down, one on each pan, each with its
own idiosyncratic rhythm, the pointer will move
from side to side. The deflection indicates, at any
given moment, the difference in weight on the two
pans. Only differences in total weights can lead to a
pointer deflection.
By now, the reader will recognize the differential amplifier as an electronic version of the chemist’s balance. Figure 3.5 and Table 3.1 spell out the
correspondence between the two.
The preceding discussion of the differential
amplifier makes it easier to see why deteriorated
electrodes or high-resistance contact with the skin
can lead to erroneously high EMG readings. For
example, if one of the active electrodes makes poor
contact with the skin, it will feed a reduced signal
to the differential amplifier. Since the other electrode is feeding a full-sized signal to the differential
amplifier, the common-mode noise signals applied
to the two inputs are of different size. Therefore,
when the subtraction process takes place, there is a
noise remainder that artificially elevates the reading. Figure 3.6 illustrates this.
The ratio of differential signal amplification
to common-mode signal amplification for a particular differential amplifier is the “common-mode
rejection ratio.” “Input impedance” is the differential amplifier specification that indicates its level
TABLE 3.1. Correspondence between Chemist’s
Balance and Differential Amplifier
Chemist’s balance
Differential amplifier
1. Pans
1. Inputs
2. Pointer
2. Output
3. Fulcrum
3. Reference
4. Equal wights in the pans
4. Common-mode
→ balance (pointer
remains straight)
5. Different weights in
the pans → imbalance
(pointer deflects)
6. Two equal weights in the
pans
and
Two unequal weights
in the pans and →
imbalance (pointer
reflects the difference
between the unequal
weights only, as equal
weights cancel out)
signals → zero
output
5. Differential signals
→ nonzero output
6. Common-mode
signals
and
differential signals
only, as equal
signals cancel out)
of protection from inaccuracy due to unequal electrode contact. Both are quite high for reputable
instruments. Further discussion of these specifications is beyond the scope of this chapter.
Internal Noise: Filters and Bandwidth
The task of removing extraneous signals is still not
complete. Electrical “filters” further reduce interference from power lines and limit the noise inevitably generated within the circuits of the EMG
amplifier itself. These filters are comparable to
tone or “equalization” controls on a stereo amplifier, except that they are usually set in one position
at the factory. Their purpose is to make the EMG
amplifier sensitive to some frequencies (or pitches)
of incoming signals and insensitive to others.
FIGURE 3.5. Graphic representation of data in Table 3.1.
43
3. A Primer of Biofeedback Instrumentation
FIGURE 3.6. Unequal common-mode noise inputs leading to a noise remainder.
Speech or music consists of a wide range of
frequencies or pitches, all combined to give us
the familiar sounds. Tone controls alter these by
increasing or decreasing bass and treble, depending on the listener’s preference. For example, turning down the treble may improve the sound of a
particularly hissy tape or scratchy old phonograph
record by reducing some of the high-frequency
scratch and hiss sounds. Turning down the bass
may improve the sound of an amplifier that has a
boomy bass or hums. In both cases, a modification
of the amplifier’s “frequency sensitivity” or “bandwidth” or “bandshape” is being made.
There are reasons to do something similar
with an EMG device. For example, much of the
electrical interference or noise from power lines
is concentrated at a narrow pitch of 60 cycles or
vibrations per second (hertz [Hz]). Anyone with
a stereo or electric guitar with a bad input cord
knows this humming or buzzing sound. To further
reduce this noise signal, a special filter can make
the EMG amplifier much less sensitive to this
pitch. Sometimes, especially in older equipment,
the entire bass response of the amplifier is “rolled
off” to further reduce any 60 Hz electrical interference remaining after the differential amplifier,
even though significant EMG energy appears in
the “bass” portion.
There is also good reason to limit the EMG
amplifier’s “treble” frequency sensitivity. All
amplifiers unavoidably generate high-pitched
noise within their own circuits that sounds like
hiss. The EMG amplifier’s treble response is typically “rolled off” (e.g., above 1000 Hz) to diminish
internal noise contributions to EMG readings.
The EMG instrument’s range of sensitivity
between the bass frequency limit and the treble
limit is called the “bandwidth.” Like speech and
music sounds, EMG signals are comprised of a
range of frequencies or pitch. They tend to vary
from about 10 to 1000 Hz. The graph in Figure 3.7
shows two idealized bandwidths superimposed on
a hypothetical EMG frequency distribution. This
shows that even with treble and bass limits, an
EMG amplifier is sensitive to significant amounts
of EMG energy.
In both cases, the amplifier’s bandwidth (range
of sensitivity) includes a significant area of EMG
energy. However, the wide bandwidth shown in
Figure 3.7 includes more EMG energy (and noise)
than the narrower bandwidth. This means that
(other things being equal) the instrument set with
a wider bandwidth will give higher readings than
the one with the narrower bandwidth. A stereo
can also illustrate this. Turning the bass and treble
controls all the way down narrows the bandwidth
and produces not only a different tone but also less
volume of both sound and noise.
Other things being equal, the wider the bandwidth, the higher the readings for both EMG and
noise. The proportion of the reading that is noise
(the “signal-to-noise ratio”) may be the same in
both cases, but the levels of both EMG and noise
will be higher with a wide-bandwidth EMG amplifier.
EMG biofeedback devices are made with different bandwidths, due to differing design philosophies. The most important message here is that
different bandwidths lead to different readings. This
has to be taken into account when one is comparing readings and noise specifications between
different models of EMG equipment. For example,
an instrument with a lower noise or sensitivity
specification may not really be any more sensitive
or noise resistant than another; it may just have a
narrower bandwidth. Readers who wish to know
more about EMG frequency distribution, filters,
and bandwidth are referred to Mathieu and Sul-
44
II. INSTRUMENTATION
FIGURE 3.7. Two hypothetical bandwidths.
livan (1990) and to Bolek (2013) for how digital
sampling and processing affects readings.
Converting EMG Energy to Information
At this point in our story, EMG energy has been
picked up from the skin and separated from extraneous noise energy. The resulting signal is proportional to the electrical activity of the motor neurons in the muscle being monitored, and is often
referred to as “raw EMG.”
Raw EMG
Raw EMG resembles auditory static; it is a rushing
sound that rises and falls in loudness in proportion
to muscle contraction. This “raw” or “raw filtered”
EMG is one form of audio feedback. Commercial
EMG units usually do not provide raw EMG audio
output. Instead, they generate an audio tone or
series of beeps, or a signal that is turned into some
form of feedback via a computer. The pitch or repetition rate is made proportional to the amplitude
or “loudness” of the raw EMG, and therefore to the
muscle contraction. Raw EMG amplitude can also
be displayed on a meter or other visual display.
Smoothing and Integration
Smoothing and integration are two ways of quantifying EMG energy over time. “Smoothing” refers
to continuously averaging out the peaks and valleys of a changing electrical signal. “Integration”
refers to measuring the area under a curve over a
time period. Both require processing the raw EMG
signal, as described in the next section.
Alternating Current and Pulsating
Direct Current
Raw EMG is an alternating current (AC) signal.
Alternating current pushes alternately back and
forth or “vibrates” like a reed in the wind or a
swinging clock pendulum, as represented in Figure
3.8. The curve represents the change in electrical
pressure over time, first in one direction and then
in the opposite direction. The “+” represents pressure in one direction, and the “–” represents pressure in the other direction. The centerline represents the point of zero voltage, analogous to the
position of the reed at rest or the clock pendulum
in its straight-down position. The height of a wave
represents its peak amplitude or peak voltage. Figure 3.8 shows an electrical signal “vibrating” at a
specific frequency (number of oscillations per second, or hertz).
Not only is the electrical signal oscillating, but
the amplitude or magnitude of the oscillations first
builds to a high point and then diminishes. It is
the measurement of this overall increase and then
decrease that is significant for EMG biofeedback.
The first step in accomplishing this is to “flip” the
negative peaks up above the zero line with the positive peaks, a process called “rectification.” Without rectification, the sum of the negative peaks
and positive peaks would always equal zero (they
would cancel each other). Without rectification, it
would be hard to recognize overall trends in magnitude unless one was viewing the oscillations on
an oscilloscope or computer screen or listening to
the raw EMG over a speaker. Figure 3.9 shows the
rectified EMG wave. The negative peaks have been
electronically “flipped” up with the positive peaks,
45
3. A Primer of Biofeedback Instrumentation
FIGURE 3.8. Gradually increasing, then decreasing alternating voltage.
so that all the peaks are positive. This means that
the electrical signal now pushes in just one direction; hence the term “direct current” (DC). In this
case, the signal is pulsating DC.
Smoothing the EMG Signal
for Moment‑to‑Moment Quantification
Electronic smoothing can be performed on the
rectified EMG signal leading to an “average” EMG.
The outputs of smoothing circuits are then used to
drive an analog or digital display, as well as audio
feedback circuitry (see Figure 3.10 and discussion
below). Electronic smoothing is essential for digital meters, because unlike old-style mechanical
meters with needles, they have no mechanical
inertia to smooth out the pulses.
Electronic smoothing or filtering circuits also
afford a wide choice of time constant or response
time. The designer has wide choice of how fast the
digital or analog display or other feedback modality will respond to momentary changes in EMG
level, referred to as “tracking time.”
The most common form of smoothing or filtering found in small, self-contained EMG equipment employs a fixed time constant (or tracking
time) suitable for general purpose use. Some EMG
instruments and computer-based systems have
selectable tracking times, which require the user
to decide how much smoothing of the curve is
desired. Long tracking time leads to a smoother
output that is less responsive to momentary ups
and downs in the EMG level. Unsmoothed output may seem too jumpy for relaxation training,
and overly smoothed output may cover or delay
information. There is no generally agreed-upon
optimum tracking time. The choice is based on
application, technique, and subjective preference.
It does not appear that any one tracking time is
particularly advantageous for relaxation training.
This view is apparently shared by the manufacturers, who build their instruments with various fixed
or adjustable tracking times.
Integration for Cumulative EMG or Average
EMG over a Fixed Time Period
A second quantification scheme involves letting
the area under the EMG curve (in microvolt minutes) accumulate over a period of time, such that
the reading starts at zero and continually builds
until the time period ends, as shown in Figure
3.11. The accumulated area under the curve at the
end of the trial indicates the accumulated number
of microvolt minutes of EMG received over that
time. Dividing the accumulated microvolt minutes of integrated EMG by the accumulated time
in minutes yields the average level of EMG (in
microvolts) during that time. Then the timer and
integrator are reset to zero, and a new time period
or trial begins. Integration establishes relaxation
trials of many seconds (e.g., 30, 60, 120, or more
seconds). Comparisons can then be made over
multiple trials—something that is more difficult
FIGURE 3.9. Rectified alternating voltage.
46
II. INSTRUMENTATION
FIGURE 3.10. Rectified filtered or smoothed EMG.
FIGURE 3.11. Integration for cumulative EMG or average EMG over a fixed time period.
to do when only moment-to-moment EMG levels
are used. This method was used more commonly
in the past.
Audio Feedback
Audio feedback encodes the EMG level in auditory form and is very important in biofeedback,
because it transmits information without the need
for visual attention. A common way to do this is
to use the smoothed EMG signal to vary the pitch
of an electronic tone generator. The higher the
EMG level, the higher the pitch. Current equipment often goes far beyond use of simple tones—
allowing the therapist or researcher to import all
kinds of sounds that can be customized as audio
feedback.
The range of possibilities for audio feedback is
virtually limitless, and many forms have appeared.
There is no one optimum form of audio feedback. Preferences develop on the basis of purely
subjective criteria, as well as application require-
ments and user preferences for modalities such as
imported sound tracks.
Visual Feedback: Meters, Lights,
and Computer Displays
Over the years, meters, lights, and computer-based
displays have been used for visual EMG feedback.
In early years, analog meters with needles were
often used, calibrated in microvolts, just using a
relative scale. Digital meters also came into use.
Analog and digital meters each have advantages
and disadvantages for particular applications, as
described in earlier editions of this volume. Later,
computers and other electronic displays became
common—and are capable of an almost limitless
range of ways to provide visual feedback.
Objective Units of Measurement
Several factors besides degree of muscle contraction affect the number of microvolts an EMG
47
3. A Primer of Biofeedback Instrumentation
device displays. A brief review of the earlier section on objectification and measurement may be
helpful. The microvolt is the unit of EMG measurement; this is an electrical term used as a measure of muscle contraction. The microvolt is not
literally a measure of muscle contraction, but a
measure of an electrical correlate of muscle contraction. Therefore, microvolt readings involve
the characteristics of the electrical apparatus (the
EMG unit) that monitors and processes the EMG
signals. Because of differences in design philosophy, EMG devices differ from one another, and
so do the readings obtained for any given degree
of muscle tension at a given site on a given person. Consequently, microvolt readings are only
objectively comparable from one model to another
if the instruments are known to have the same
bandwidth and quantification method.
For the technically inclined, EMG instruments
are AC voltmeters that make objective AC voltage
measurements. However, the internal characteristics of bandwidth or bandshape and quantification
method affect these measurements. Accuracy, if
specified, is only at a given frequency within the
bandwidth. Because EMG voltages sensed by surface electrodes are composed of an ever-changing
blend of frequencies (see Figure 3.7), the bandwidth
or bandshape of any particular unit will affect the
readings. (This has been discussed in the section
on internal noise, filters, and bandwidth.)
Quantification method also affects EMG
instrument readings—the numbers on a display.
For those technically inclined: Early EMG instruments were calibrated in “average,” “root mean
square (RMS),” or “peak to peak” microvolts,
though there is little practical difference between
these quantification methods or in the action
of a meter or other display, just different scales.
Calibration is done using conveniently available
constant-amplitude AC signals called “sine waves”
rather than nonstandardized EMG signals (see
Figure 3.8).
• The term “peak-to-peak microvolts” refers to
the voltage difference between the positive
peaks and the negative peaks of the unrectified
AC sine wave. Quantification by the “averaging” method usually involves rectification and
smoothing, then moment-to-moment display, or
integration and division by time, both described
earlier.
• The “average” voltage of a sine wave after rectification as displayed on a meter is equal to just
less than one-third of the “peak-to-peak” value.
To convert from peak-to-peak values to average
values, divide by 3.14.
• Quantification by the RMS method involves
electronically making a mathematical computation on the filtered EMG signal to arrive at
an RMS voltage. RMS quantification reflects
the electrical power as contrasted with voltage,
carried by the signal. RMS values for EMG are
usually within 20% of average values.
In any case, the user of EMG equipment should
become familiar with the range of readings
obtained under various conditions, and should
be cautious about comparing microvolt readings
between units that are not known to have similar characteristics. The lesson here is that even
though EMG instruments are AC voltmeters,
EMG readings are not made on standardized scales
and are not standardized measurements of muscle
contraction. Variability exists between EMG
instruments, and there is no standardized scaled
correspondence between EMG microvolts and
muscle contraction.
Thresholds
A threshold control allows the user to set a particular EMG level as a criterion for some form of feedback (e.g., to turn on audio feedback only when
EMG exceeds the threshold). Visual feedback,
such as lights or a computer display, may indicate
when EMG exceeds or drops below the chosen
threshold level. Thresholds are adjusted over time
as training goals change.
Other Feedback Modes
The smoothed EMG level can be used to operate
virtually any feedback method, including lights,
sound, appliances, computers, or tactile feedback
devices. All forms of feedback are ways of encoding EMG level as meaningful information or consequences. Choice of feedback mode depends on
the requirements of the application and the people
using the feedback. Although complex or novel
feedback may be interesting, the best feedback
modes for a given application are the ones that get
the information across with a minimum of distraction and ambiguity. Simple, well-designed feedback often fits this criterion. Practitioners often
settle on a limited number of practical feedback
modes.
48
Safety
EMG equipment makes direct electrical connection to a person via surface electrodes, thereby
establishing a path for bioelectric signals between
the person and the instrument. Although this
path is intended for bioelectric signals, electricity
from other sources can also take this path under
some conditions. The presence of other currents
in the signal path is a risk. Consequently, great
care is taken in the design and manufacture of
top-grade biomedical instruments to minimize
the possibility of exposing patients to extraneous
electrical currents. Despite this, no equipment,
no matter how well made and installed, is 100%
immune from electrical hazards for all time.
The chance of risky electrical faults developing is small, especially in battery-operated equipment, but the manner in which an equipment user
sets up and maintains the equipment is at least as
important to patient safety as the soundness of the
equipment design. It is therefore the responsibility
of the professional using these instruments to be
aware of potential electrical hazards and to take
standard safety precautions in installing, using,
and maintaining the equipment. If there is any
question about the safety of a particular installation, the professional must consult the manufacturer of the equipment or a qualified biomedical
engineer or technician. This is particularly important when there are multiple instruments or any connections to power-line-operated equipment or accessories.
A good rule of thumb is to be skeptical of the
safety of all setups involving power-line-operated
auxiliary equipment, such as audio amplifiers,
computers, or other devices (especially if not part
of a commercially purchased biofeedback setup),
until the safety of the installation is positively
established. This is because the potential consequences of leakage current from the AC power
line can be extreme. For example, it takes only
0.009 amperes (9 milliamperes) or less to cause
a person to be unable to release his or her grasp
on an object through which the leakage current
flows. Respiration may be affected at approximately 18 milliamperes, and heart fibrillation (and
death) may occur at about 50 milliamperes. This is
hundreds of times less than the current required to
blow a standard household fuse or circuit breaker,
so they provide no protection. There are several
precautions to take, some of which require the
consultation of a biomedical engineer or technician:
II. INSTRUMENTATION
1. Each power-line-operated piece of auxiliary
equipment should be periodically evaluated technically and certified by a biomedical technician
for electrical safety. Power-line operated EMG
equipment should also be periodically tested for
leakage currents. Consult the manufacturer or
your biomedical technician.
2. Keep all patients or subjects out of arm’s
reach of all metal building parts, such as radiators
and plumbing.
3. Ground all equipment properly. Use a
“ground fault interrupter,” a device that senses a
diversion of electricity from the normal pathway
established by the two legs of the standard power
circuit. This device shuts down power to the equipment if more than about 5 milliamperes of current
is “lost” through non-normal pathways, such as
leakage current to ground through a person.
Troubleshooting with a “Dummy Subject”
High-grade EMG circuitry is quite reliable, but
electrodes, cables, and batteries may need frequent
service in heavily used installations. Diagnosing
failure of these parts is usually simple and requires
few tools.
Faulty electrodes or electrode contact usually
leads to spuriously high readings. Follow the electrode maintenance and application instructions
supplied with the instrument. If unexpected or
suspiciously high readings are observed, determine
whether the problem is in the electrodes, the electrode contact, the cable, or the EMG unit. Use a
“dummy subject,” which is nothing more than two
resistors that can be snapped to the electrode cable
in place of the normal electrodes. This simulates a
subject with zero EMG (see Figure 3.12).
The dummy subject supplies about the same
“input resistance” as actual electrodes on skin, but
generates no EMG signals. With the dummy subject in place, the readings should therefore be close
to the residual noise level of the instrument as
given in its specifications. For a fair test, hold the
electrode cable between the fingers at least a foot
away from the dummy subject as it dangles toward
the floor. This distance prevents excessive noise
from being coupled from your body to the dummy
subject. EMG readings with the dummy subject
typically vary as you twist the cable between the
fingers, much as television reception on “rabbit
ears” varies as one rotates the antenna.
If a dummy subject test done in the patient area
results in a reading near the instrument’s residual
49
3. A Primer of Biofeedback Instrumentation
FIGURE 3.12. Dummy subject.
noise specification, then it is safe to conclude that
electrical noise in the area is not overpowering.
This means that suspiciously high readings with the
real subject are not the result of failure of the EMG
unit or electrode cable. In this case, the fault is most
likely with the electrodes or electrode contact.
If the reading goes off scale and stays there
while the dummy subject is rotated, there is likely
to be a break in the electrode cable. Verify this by
substituting another cable. If the repeat test still
leads to off-the-scale or very high readings, then it
is likely that the fault is with the EMG unit itself,
or that the work area is saturated with electrical
noise. High dummy subject readings that do not
go off the scale may be attributable to excessive
noise from nearby electrical equipment. Check
this by moving the machine to other locations and
repeating the dummy subject test.
If the dummy subject test indicates that the
instrument and cable are working properly, but
abnormally high readings with the real person
being tested remain, consider removing the electrodes, cleaning the person’s skin again, and reapplying the electrodes.
Construct a dummy subject if you do not have
one already. Experiment with the dummy subjects
when you know that your instrument and cables
are working properly. You will then be in a better
position to judge test results with dummy subjects
when actual failures occur.
Battery Failure
Abnormally high or low readings may result from
battery failure. Most instruments have a built-in
battery check or battery-checking instructions in
the user’s manual. Use it whenever there is doubt
about the accuracy of the readings. Aging batteries
may pass the check and work fine early in a session,
deteriorate during the session, then “self-rejuvenate”
after a few idle hours. The usable time after these
self-rejuvenations gets shorter and shorter, until the
batteries are unable to power the equipment at all.
Summary
A summary block diagram of a hypothetical EMG
instrument with several outputs is presented in
Figure 3.13.
Operation of the Temperature
Biofeedback Instrument
Temperature biofeedback instruments measure
changing skin temperature, which is significant
because it is linked, through vasoconstriction, to
sympathetic arousal. Vasoconstriction affects perfusion of blood and therefore skin temperature,
particularly in the extremities (especially the fingers and toes). Typically, sympathetic arousal leads
to increased vasoconstriction, which leads to a
reduction in blood volume and hence to a cooling
effect at the skin.
Although this neurovascular phenomenon
involves the constriction and dilation of vessels,
the single term “vasoconstriction” is used here to
denote all changes in vascular diameter. For example, “reduced vasoconstriction” is used to express
the idea of vasodilation.
50
II. INSTRUMENTATION
FIGURE 3.13. Block diagram of hypothetical EMG instrument with several outputs.
The tasks of a temperature biofeedback instrument are as follows:
1. To let the skin heat a temperature-sensitive
probe.
2. To make the probe serve as a temperature-
sensitive electrical “valve” that modulates an
electric sensing current applied to the probe.
3. To display temperature-dependent variations
in probe current as temperature in degrees, and
to provide other temperature feedback or information meaningful to the user.
Letting the Skin Heat a Probe
A typical temperature probe is made of one or
more small pieces of heat-sensitive electrical
material (called “thermistors”), encased in electrically insulating material with wires protruding
for connection to the temperature unit. A temperature probe is not an electrode. It is specifically
designed to make only thermal contact, not electrical contact with the skin, where it is usually taped
or strapped. The probe accepts heat from the skin
and remains at nearly the same temperature as the
skin immediately beneath it. As the skin warms
and cools, the probe warms and cools accordingly—but with a slight delay, as probe temperature takes a little time to “catch up” with the skin
temperature.
The probe is attached to either side of a finger.
No single site is standard, nor has any particular
site been shown to be superior. However, consistency from session to session is important, because
temperature or speed of response may vary from
site to site. The dorsal surface (back side) of the
fingers is a common site. This permits the person
to rest the hand on the chair or lap without artificially warming the probe between the finger and
the chair or body. Furthermore, the dorsal surface
has fewer sweat glands, so the chance of evaporative cooling is less. It is no doubt possible to make
a case for the use of other sites as well, but consistency will probably remain more important than
the specific choice of finger site.
Making the Probe Serve
as a Temperature‑Sensitive Electrical Valve
The heat-sensitive probe acts like a “valve” for electricity applied to it from the instrument, analogous
to a water valve that gradually opens and closes to
regulate water flow. But in this case, probe temperature operates the “valve” and regulates the flow of
electricity. As the probe heats, its electrical resistance decreases, and more electric current flows.
As the probe cools, its resistance increases (the
“valve” closes a little), and less electric current
flows. In this way, probe (and skin) temperature is
encoded in the electrical flow through the probe.
51
3. A Primer of Biofeedback Instrumentation
Displaying Temperature
and Other Feedback
TABLE 3.2. Ohm’s Law: Voltage, Resistance,
and Current
The temperature instrument measures the current flow through the probe and displays this
quantity (properly scaled) as degrees or as other
feedback.
Electrical law
Internal Workings of Temperature Feedback Devices
Temperature feedback instruments can perform
the required operations in more than one related
way. Use of temperature biofeedback equipment
does not require detailed knowledge of internal
workings. However, it is important to understand
the basic scheme shared by all temperature feedback devices.
Hydraulic analogy
Units
Volt: Unit of electrical
pressure
Pounds per square inch:
Unit of water pressure
Ampere: Unit of electric
current flow
Gallons per minute:
Unit of water flow
Ohm: Unit of resistance to
electric current flow
Unspecified unit of
resistance to water flow
Circuit description
Pressure (in volts) pushes
the current (in amperes)
Quantification
Ohm’s Law
Temperature feedback devices operate on one
or another form of Ohm’s law. Georg Ohm was
the Bavarian scientist who, in 1827, specified the
quantitative relationships among three basic elements of an electrical circuit: voltage, resistance,
and current. In 1891, the Electrical Congress in
Paris agreed that electrical pressure would be
measured in volts, after Volta, an Italian; electrical flow volume would be measured in amperes,
after Ampère, a Frenchman; and resistance in
ohms, after Ohm, a German. Because there is a
convenient hydraulic analogy to Ohm’s law, the
law and the analogy are presented together in
Table 3.2.
Ohm’s Law and a Temperature
Feedback Device
According to Ohm’s law, the amount of current
flowing in a circuit powered by a constant voltage depends entirely upon the resistance in the
circuit. The resistance of the probe varies with
its temperature. Therefore, when the probe is the
only resistance element in a constant-voltage circuit, the current flow in the circuit is proportional
to the temperature of the probe. The quantitative
relationship between temperature and probe resistance is a property of the probe and varies greatly
from one model to another. For this reason, probe
models are usually not interchangeable. A suitable
current-sensing circuit with meter displays a reading in degrees. Figure 3.14 shows a hypothetical
temperature feedback device.
Pressure (in pounds per
square inch) pushes the
water flow (in gallons
per minute) through the
resistance of the pipes
Current = pressure/resistance; that is,
Amperes = volts/ohms (Ohm’s Law)
Algebraic formulas
Volts = amperes × ohms
Ohms = volts/amperes
Conventional abbreviations
Voltage: V or E
Current: I
Resistance: R
Parameters of Temperature Feedback Devices:
Ways They Differ from One Another
Temperature feedback devices come in a wide
range of performance and cost. The following
three parameters—response time, absolute accuracy, and resolution—provide a basis for judging or
comparing the performance of temperature feedback devices. An exploration of various device
factors that may influence feedback and training
results in computer biofeedback systems appears in
Otis, Rasey, Vrochopoulos, Wincze, and Andrasik
(1995).
Response Time
“Response time” indicates how rapidly the unit
responds to a change in skin temperature. It is
mostly a property of the probe; if a probe responds
quickly, feedback delay is minimal, and small
52
II. INSTRUMENTATION
FIGURE 3.14. A hypothetical temperature feedback device.
temperature changes are readily apparent. However, quick response time is usually gained at the
expense of increased cost and fragility. A very fastresponding probe is very small and light, encased
in a material that gains or loses heat very rapidly in
step with skin temperature. Larger, bulkier probes
are cheaper and more durable, but they take more
time to heat and cool as skin temperature changes.
A very fast-responding probe is not thought
necessary in most applications. To understand
why, recall that skin temperature is important
because it provides indirect access to peripheral
vasoconstriction. There is already a considerable
time delay between a change in vascular diameter
and the resultant change in skin temperature.
Probe response time adds a second delay to the
overall delay between the vascular event and the
resulting temperature event. One could argue that
because of these delays, it is important to minimize probe response time, so that further delay is
kept to a minimum. A counterargument is that
skin temperature is a relatively slow-changing
phenomenon to which rapid response time does
not add value for relaxation applications. Neither
view holds obvious sway. Successful thermal biofeedback appears to have been conducted with
temperature devices of widely differing response
times.
Absolute Accuracy
“Absolute accuracy” refers to how closely the
displayed temperature corresponds to the actual
probe temperature. Virtually any temperature
machine will follow temperature changes (delayed
by its particular response time), but there is variation between instruments in the accuracy of the
temperature readings. Although a given unit may
respond very sensitively to changes in temperature,
it is unlikely that readings will exactly equal the
true temperature of the probe; it may read up to
a few degrees higher or lower than the true temperature.
Furthermore, two identical units monitoring the same site will probably not give exactly
the same readings. This variability in absolute
accuracy is to be expected, and the error range
for a given unit is usually included in its specifications. Absolute accuracy of ±1°F is considered
sufficient. Absolute accuracy is a tradeoff against
cost, because ensuring a high degree of absolute
accuracy tends to be very expensive. And practical
advantages of highly accurate temperature equipment for clinical biofeedback are not evident. Successful biofeedback takes place with widely differing degrees of absolute accuracy—and includes
devices that are not calibrated to the Fahrenheit
or Celsius standard at all, giving only relative indications of warming and cooling.
The question of accuracy arises for temperature
feedback equipment because there exist standardized temperature scales (Fahrenheit and Celsius).
Remember that although temperature is measured
on a standardized scale, vasoconstriction is not. An
absolutely accurate temperature reading does not
imply an absolutely accurate gauge of vasoconstriction, much less sympathetic arousal.
Resolution
“Resolution” refers to the smallest temperature
change that the instrument can discern and
display. Resolution affects length of feedback
delay. For example, a digital unit that resolves
53
3. A Primer of Biofeedback Instrumentation
to 1°F will feed back that a temperature change
has taken place when a 1° change has occurred.
Since temperature change occurs over time, the
feedback will be delayed by however long it takes
for the temperature to change 1°F. A resolution
of 0.1°F will provide much more rapid feedback,
because it takes far less time for the temperature
to move 0.1° than 1°. Instruments could be built
to resolve 0.01°F, which would reduce feedback
delay even further. However, extremely high resolution also increases the risk of mistaking artifact
for vasoconstriction-caused temperature change.
For example, the effects of movement, a light
breeze, and room cooling are much more likely
to affect the readings from an instrument with
exceedingly fine resolution than from one with
coarser resolution. Furthermore, a high-resolution temperature instrument must be manufactured with much more exacting tolerances and
increased expense. Otherwise, it may create discernible change in the readings through “drift” in
its own circuits. An instrument with exceedingly
high resolution is more likely to display distracting information or artifacts superimposed on true
vasoconstrictive effects. A resolution of 0.1°F is
historically a typical resolution value for temperature instruments and appears to be a suitable general purpose value.
Digital and analog feedback have different
resolving power. For example, a digital meter with
three digits (10’s, 1’s, and 10ths) can resolve to
0.1°F. However, an audio tone (e.g., the sensitive
pulsed-tone feedback described in the EMG section) indicates even finer differences that occur
during the interval between changes of the 10ths
digit on the meter.
heat from the skin. Cool air may also directly cool
the probe.
Artifacts
Room Temperature and the Temperature
Feedback Instrument
Because peripheral temperature is an indirect
index of peripheral vasoconstriction, there are
several sources of misleading readings. In looking
for sources of artifact, the question to ask is this:
“What conditions lead to temperature changes
that are not linked to vasoconstrictive changes?”
Cool Room Temperature
Air temperature in the room where measurements
are being made may affect the readings. For a
given degree of vasoconstriction, skin temperature may be cooler in a cool room than in a warm
room, simply because the cool air absorbs more
Breeze
Moving air exaggerates the cooling effect mentioned earlier in two ways. First, breeze removes
heat from the skin more rapidly than does still air.
Second, breeze evaporates sweat more rapidly than
does still air.
Warm Room Temperature
Room temperature sets an approximate lower
limit for hand temperature. That is, a hand cannot cool very much below the temperature of the
air around it. This is because cooling takes place
through the dissipation of heat from the hand to
the air. As soon as the hand cools down to the
temperature of the air, there is no longer anyplace
for heat to go. The hand remains at about that
temperature regardless of further vasoconstriction, unless the skin cools a little further as sweat
evaporates.
Warm air effect is usually not a problem,
because room temperature is usually below 72°F
(close to the low end of the skin temperature range
for most persons). However, in the event of a high
room temperature, higher skin temperature will
be observed than in a cooler room, even with an
identical degree of vasoconstriction. For example,
using thermal biofeedback in a 90°F room will
lead to warmer hands for everyone, regardless of
the degree of vasoconstriction. In this case, even
the hand temperature of a cadaver, which has no
warm blood at all, would be 90°F!
Even if the temperature of the probe is held constant, temperature readings may change as the
temperature unit itself is heated and cooled. The
performance of electronic circuitry is vulnerable
to change or “drift” as surrounding air temperature
changes. This is a well-known phenomenon that
designers take into account. Such “temperature
compensation” is very important for temperature
instruments, because they are required to resolve
exceedingly small changes in electric current from
the probe. If temperature compensation is inadequate, then readings vary with room temperature,
as well as with skin temperature. This source of
54
artifact is not practically significant unless room
temperature is known to vary over a wide range.
Probe Contact and “Blanketing”
Changes in probe contact caused by movement
also affect temperature readings. If the probe
begins to lift from the skin when pulled by its
leads, lower readings are likely. The opposite
occurs when the probe is covered by a hand, clothing, or materials used to secure the probe to the
skin, all of which have the effect of “blanketing”
the probe. It is important to allow normal heat dissipation into the surrounding air rather than trapping the probe or hand under such materials.
Chill
If the person to be monitored comes in chilled
from the outside, cold hands are likely. Cold hands
should be allowed to restabilize indoors before
training begins. Otherwise, the natural warming
of the hands after being exposed to cold may be
mistaken for a training effect. It can take considerable time for skin temperature to stabilize after
coming in chilled from outdoors.
Testing for Absolute Accuracy
Test the accuracy of temperature instruments by
immersing the probe in a glass of water along with
a laboratory thermometer of known accuracy, then
stirring the water. Compare the readings after they
have stabilized. This test is useful when the accuracy of the instrument or probe is questioned, or
when the actual interchangeability of “identical”
probes is assessed.
If done carefully, this method can be used to test
for temperature drift in the temperature instrument itself. With probe temperature stabilized in
a thermos of water, heat and cool the instrument
while noting any change in its reading.
Other Feedback
Different models of stand-alone or computerized
temperature biofeedback instruments employ different variations on the basic audio and visual
feedback described in this chapter, as well as different levels of response time, absolute accuracy,
and resolution. A question remains about which
combinations of these parameters and feedback
modes are most effective for training various skills.
II. INSTRUMENTATION
There is some evidence that significant differences
may exist (Otis et al., 1995), although a systematic
research base on these many variations does not
exist.
Audio Feedback
Digital meters are often used for visual feedback,
because they resolve small differences over a very
wide range. Audio tones cannot provide the same
resolution over such a wide range. If a usable range
of audio pitches is simply distributed over the
working range of skin temperature, then persons
with very low or high skin temperature will have
to listen to feedback in the extremes of the audio
range. This will be uncomfortable to listen to for
long. Moreover, small changes in temperature will
lead to only slight changes in the pitch of the tone.
A good solution to this problem is to let the user
move the entire pitch range of audio tones up and
down the temperature range, so that high-resolution audio feedback in a comfortable pitch range
can be obtained, regardless of the actual skin temperature. Moving the audio range is accomplished
by turning a control that affects the pitch of the
audio feedback but not the meter readings. In this
way, the user adjusts the audio feedback for a comfortable pitch range around any temperature.
Some temperature machines have a control
that allows the user to select whether the pitch
rises or falls with temperature. This encourages
the user to fit the audio feedback to his or her
warming images. For example, some users feel that
the image of increasing blood flow through the
fingertips calls for an increasing audio pitch. Others find decreasing pitch more natural as relaxation occurs.
Derivative Feedback
Derivative or “rate” feedback is sometimes found
on temperature instruments. “Derivative” is a
mathematical term referring to rate of change.
In a temperature machine, this usually takes the
form of a light or tone that turns on when skin
temperature is changing at a certain rate. For
example, a red light turns on when the person’s
hand temperature is climbing at 1°F or more per
minute. Another light or tone might come on if
the person’s hand temperature is falling at that
rate. This establishes a target hand-warming rate
and permits a summary quantification, such as the
percentage of time above the target warming rate.
3. A Primer of Biofeedback Instrumentation
Safety
Because no electrodes are used, temperature biofeedback equipment may not pose the same electrical safety challenges as EMG equipment. The
probe is deliberately electrically insulated from
the subject, so the chances of a risky electrical
fault’s developing may be lower than with EMG
equipment. Nevertheless, temperature equipment
should not be considered exempt from the safety
precautions discussed earlier for EMG equipment.
If, for example, a probe fails (internally or through
a break in its insulation) so that it is no longer
insulated from the skin, it becomes in effect an
electrode. This increases the potential for electric
shock or leakage currents, particularly since the
temperature device is probably not specifically
designed to operate safely with a direct electrical
connection to a person. Therefore, to be as safe
as possible, follow the safety guidelines for EMG
equipment. Moreover, safety guidelines are best
thought of as applying to entire biofeedback installations, not just the individual units in isolation.
Electrodermal Biofeedback
Early History of Electrodermal Research
The early history of electrodermal research is an
interesting story recounted by Neumann and Blanton (1970). They begin the story with Galvani’s
discovery of the electrical processes in nerve and
muscle action, which quickly stimulated research
into the medical applications of electricity. By
1840, it was widely believed that electrical processes provided a basis for explaining disease and
generating diagnoses and therapies. The authors
note that this was strongly consistent with the
physicalistic thinking of the day, in reaction to the
vitalistic thinking of earlier times. By 1870, thensophisticated instrumentation and procedures had
been developed as part of electrophysiological
research methodology. (A fascinating collection of
such literature and instrumentation exists at The
Bakken Museum, Minneapolis, MN.)
As the field developed, investigators noted
that skin resistance varied over the body. Because
investigations focused on the physical effects of
electrical currents and static fields, the early workers noted that variations in skin resistance introduced variations in current flow through the body;
hence they viewed variations in skin resistance as
a source of artifact, and they built instruments
55
that controlled for this artifact. Most researchers
continued to regard variations in skin resistance as
artifact encountered while applying electric current or static fields for diagnostic or therapeutic
purposes.
But in 1879, Romain Vigouroux measured skin
resistance as an experimental variable in cases of
hysterical anesthesias. This, according to Neumann and Blanton (1970), is generally regarded
as the first observation of psychological factors in
electrodermal phenomena. In 1888, Vigouroux’s
colleague, Charles Fere, studied the effect of physical stimulation on skin resistance, noting increases
in current flow following stimulation. This, the
reviewers say, was the first study of what by 1915
was called galvanic skin response (GSR), and was
probably the first statement of an arousal theory.
It is noteworthy that by Fere’s time, the French
physicist D’Arsonval had developed silver chloride nonpolarizable electrodes for physiological
research, as well as a sophisticated “galvanometer”
(needle-type meter), a forerunner of modern meter
movements that still bear D’Arsonval’s name. The
German investigator Hermann linked GSR with
sweat gland activity in 1881, thus establishing a
physiological basis for the phenomenon. In 1889,
the Russian investigator Ivan Tarchanoff, while
investigating skin potentials, showed that not only
physical stimuli but also mental activity (e.g., mental arithmetic and the recollection of upsetting
events) led to skin potential changes. Moreover,
he linked this phenomenon to the distribution of
sweat glands and proposed that it was related to
the action of “secretory nerves.” Neumann and
Blanton (1970) report that Tarchanoff’s and Fere’s
papers were followed by “several years of oblivion.”
GSR was rediscovered in 1904.
At that time, a Swiss engineer, E. K. Mueller,
noticed that skin resistance changes with psychological events. He showed this to the Swiss neurologist Veraguth, and both believed this to be a
newly discovered phenomenon. Mueller went on
to assume the role of a psychological expert and
to address the technical problems of measurement, reliability of electrode design, and experimentation with the use of AC. By 1905, Veraguth
had finished some preliminary experiments when
he embarrassedly discovered the earlier work of
Tarchanoff and others.
Veraguth and Carl Jung were friends, and
somehow (each claimed to have suggested it to
the other), GSR was used in Jung’s word association experiments. Jung then provided most of the
56
impetus for further studies in this area. By 1907
he considered GSR, known to Veraguth and Jung
as “psychogalvanic reflex,” a means of objectifying heretofore invisible “emotional tones.” Jung
embarked on extensive studies and exported this
idea to friends in the United States. Neumann
and Blanton (1970) report that a “flood” of papers
in America appeared over the next two decades
and established this field as a major research area.
Since then, GSR has been recognized as a way
to gain objective access to psychophysiological
arousal.
This physiological variable has appeared in
countless psychological experiments, in clinical
practice, in “lie detector” equipment, and even in
toys and parlor games. Biofeedback has used it for
access to autonomic arousal. GSR is recognized
as distinctively sensitive to transitory emotional
states and mental events, while often remaining
more or less independent of other biofeedback
measures such as muscle tension and skin temperature. It is a complex variable, responsive to
a wide range of overt and covert activities and
external and internal stimulation. Its responsivity
to psychological content in actual or laboratory
human situations apparently prompted Barbara
Brown (1974) to dub GSR “skin talk.” This is an
apt metaphor that does justice to its psychological
responsivity, while legitimizing its often complex
and seemingly unpredictable variations and individual differences.
Like any actual language, “skin talk” must be
studied and experienced to be understood. EMG
and temperature biofeedback are, in comparison,
more easily understood by virtue of their less articulated response to mental events. That is, EMG
and temperature biofeedback tend not to reflect
mental events as quickly or with as much resolution as GSR. Electrodermal phenomena are often
regarded as more complex and less well conceptualized than other biofeedback measures because of
rapid responsiveness, individual variability, methodological challenges in measurement, and the
multiplicity of technical approaches. My purpose
in this section is to conceptualize the skin conductance phenomenon, and to describe and critique
some of the approaches to skin conductance measurement and instrumentation.
As revealed in the history provided earlier, two
forms of EDA have been studied. The most common is the exosomatically recorded activity of Fere,
Veraguth, and Jung, in which an external electric current is passed through the skin. Activity
is indicated by the skin’s electrical resistance (or
II. INSTRUMENTATION
its reciprocal, conductance). The second method,
that of Tarchanoff, is endosomatically recorded
activity (skin potentials), which involves monitoring voltage differences between electrodes at two
points on the surface of the skin. The endosomatic
method is not covered in this chapter, because it is
much less common in biofeedback than exosomatically recorded skin conductance. For more on the
endosomatic method, see Venables and Christie
(1980). For more on EDA, see Dawson, Schell, and
Filion (2000) and Boucsein (2012). Dawson et al.’s
work is a chapter in the third edition of Cacioppo
et al.’s (2000) Handbook of Psychophysiology. This
handbook is also recommended for basic information relevant to biofeedback.
Terms
GSR is historically a universally recognized term
for EDA, perhaps because the term has for a
long time referred to a variety of exosomatic and
endosomatic phenomena, and to both levels and
responses. Although the term GSR will probably
continue in widespread use, other terminology has
been suggested that is more descriptive of specific
electrodermal phenomena. Adopted from Venables and Christie (1980), the following nomenclature is used in this chapter.
Electrodermal activity (EDA), electrodermal
response (EDR), and electrodermal level (EDL)
are general terms for either exosomatic or endosomatic phenomena. EDL refers to baseline levels;
EDR refers to responses away from baselines; and
EDA, the most general term, refers to levels and/
or responses.
Skin conductance activity (SCA), skin conductance response (SCR), and skin conductance
level (SCL) specify the exosomatic method and
the conductance (in contrast to resistance) scale.
Again, SCL refers to baseline levels; SCR refers to
changes from baselines; and SCA refers to either
or both.
Parallel terms for skin resistance and skin
potentials are sometimes used: skin resistance
activity (SRA), skin resistance response (SRR),
and skin resistance level (SRL); skin potential
activity (SPA), skin potential response (SPR), and
skin potential level (SPL).
Table 3.3 clarifies the meaning of all these
terms and their interrelationships. Although the
table contains a dozen terms, this chapter is concerned only with SCA—that is, SCL and SCR.
These are clearly the prevalent forms of electrodermal biofeedback.
57
3. A Primer of Biofeedback Instrumentation
TABLE 3.3. Organization of Electrodermal Terms
Activity
Response
Level
Exosomatic
Endosomatic
or exosomatic
Conductance
Resistance
Endosomatic
EDA
EDR
EDL
SCA
SCR
SCL
SRA
SRR
SRL
SPA
SPR
SPL
Electrical Model of the Skin
The skin is electrically complex, and no one
claims to have perfect knowledge of the physiology
of EDA. But the following electrical model of the
skin brings out the essential features of practical
importance in biofeedback.
The skin on the palm or volar surface of the
hand contains up to 2000 sweat glands per square
centimeter. Each sweat gland, when activated, can
be considered a separate electrical pathway from
the surface of the skin, which normally has high
resistance, to deeper and more conductive layers
of the skin. This is shown in Figure 3.15, based on
Venables and Christie (1980).
Each resistor represents the conductive pathway
of a sweat gland. For illustrative purposes, a sweat
gland is considered “on” or “off.” When it is “on,”
it forms a low-resistance path from the skin surface to deeper layers. When it is “off,” it makes a
very high-resistance pathway. In Figure 3.15, some
glands are shown “on” and others are shown “off.”
The inner layers of skin are highly conductive,
but the outer layer is highly resistive. This means
that the resistors are electrically tied together at
the deeper layers within the skin but are electrically isolated from each other at the surface. This
presents an opportunity to monitor sweat gland
activity electrically. If two electrodes are placed
over skin laden with sweat glands, and a voltage is
applied to the electrodes, a circuit is formed, and
an electric current will flow. The size of the current will depend (according to Ohm’s law) on the
resistance of the skin, which in turn depends on
the number of sweat glands turned “on.” See Figure 3.16 for an illustration.
As more and more sweat glands turn “on,” more
and more conductive pathways switch into the
circuit, and (since some current flows through
each pathway) more and more total current flows.
In this case, Ohm’s law determines current flow,
just as it does in temperature instruments. The
difference is that the skin (instead of a temperature probe) acts as a variable resistor that regulates
current flow through the circuit. The meter measures current flow in the circuit, and the reading is
proportional to sweat gland activity. (Review this
circuit in the section on temperature biofeedback
instruments by substituting “skin resistance” for
“probe resistance” in the explanation of Ohm’s
law.)
Scales and Measurement:
Resistance and Conductance
At this point, I distinguish resistance from conductance and explain why conductance is the preferred measurement unit. “Resistance” and “conductance” are defined as reciprocals of each other,
FIGURE 3.15. Electrical model of the skin. Based on Venables and Christie (1980).
58
II. INSTRUMENTATION
FIGURE 3.16. Basic skin conductance current loop.
and they represent the same basic electrical property of materials. As discussed earlier, the ohm is
the unit of resistance. The unit of conductance is
the “mho” (“ohm” spelled backward); it is defined
as the reciprocal of resistance (i.e., 1 divided by
resistance). Therefore, resistance is also the reciprocal of conductance (1 divided by conductance).
These are two scales for measuring the same phenomenon (see Table 3.4).
A newer term for micromhos (one millionth of
a mho) is “microsiemens” after the German inven-
TABLE 3.4. Correspondence between
Conductance and Resistance
Conductance
Resistance
Units
Mho
Micromho (millionth)
Ohm
Megohm (million)
Conversion formulas
Conductance =
1/resistance
Mho = 1/ohm
Micromho = 1/megohm
Resistance =
1/conductance
Ohm = 1/mho
Megohm = 1/micromho
Sample correspondences
1 micromho ~ 1 megohm
10 micromhos ~ 0.1 megohm
100 micromhos ~ 0.01 megohm
Range of skin conductance values
Approx. 0.5 micromho to
50 micromhos
Approx. 0.02 megohm to
2 megohms
tor and industrialist Ernst Werner von Siemens.
This term appears in textbooks and is synonymous
with “micromhos.” I continue to use micromhos
as a synonym for microsiemens in this chapter,
because the older term may be more familiar to
many readers.
Although resistance and conductance scales
measure the same property, there is a good reason
to use the conductance measurement scale. Recall
that as sweat glands turn “on,” they add conductance pathways within the skin. This means that
conductance increases in a linear relationship to
the number of activated sweat glands. Resistance,
on the other hand, decreases in a nonlinear fashion as more and more sweat glands are activated.
This is shown graphically in Figure 3.17.
The linear relationship between sweat gland
activity and skin conductance is statistically preferable for scaling and quantification. This is why
skin conductance is now the standard unit. There
are times (e.g., when one is using Ohm’s law or
testing electrodes) when it is more convenient to
think in terms of resistance rather than conductance. Once the relationship between these two
scales is understood, shifting from one scale to the
other presents no problem.
Speaking of scales and measurement, note that
skin conductance is not a direct measure of sweat
gland activity (i.e., how many are turned “on”).
Rather, it is an indirect measure that, except for
artifact, correlates highly with sweat gland activity. That is, conductance is an electrical concept,
not a physiological concept; it is not a direct measure of how many sweat glands are in operation.
Because skin conductance results only when
an electrical voltage is imposed from outside, the
59
3. A Primer of Biofeedback Instrumentation
FIGURE 3.17. Comparison of skin conductance (left) and resistance (right) scales.
measurement apparatus is inextricably tied into
the skin conductance phenomena and contributes heavily to the observations. For the technically inclined, skin resistance or skin conductance
biofeedback instruments are designed to be ohmmeasuring or mho-measuring meters.
As such, they objectively measure whatever
electrical equivalent network is presented to
their inputs. They are characterized in part by
the means of applying electrical excitation to the
skin— either a steady-state voltage (DC) or an
alternating voltage (AC)—and by their readout in
either ohms or micromhos (or “microsiemens,” the
newer term for micromhos). If a calibrated readout
is provided, calibration is usually done by presenting a known value or values of simple electrical
resistors and by verifying that the unit displays
those values to within the specified accuracy of
the instrument.
The problem is that skin presents a far more
complex electrical network than simple calibration resistors. Sweat glands are not uniformly distributed in skin tissue, so sensing sites and electrode surface areas affect readings. If DC current
loops are used, electrode material may be very
important, because voltage may accumulate at the
skin–electrode interfaces, which then act like tiny
batteries and influence the readings. This is called
“electrode polarization” and is discussed later. The
use of silver/silver chloride electrodes will minimize but not eliminate this artifact. If AC current
loops are used, polarization effects are minimized,
but “reactive” components of the electrical equivalent network of the skin will cause an apparent
increase in skin conductance. (These and other
artifacts are discussed in a later section.) Finally,
the electrical resistance of skin tissue may vary
with the magnitude of the current in the current
loop.
In summary, biofeedback users should not
assume that each other’s or published quantified
SCA readings are actually comparable. Specifications of the conditions outlined earlier (plus the
technical knowledge required to interpret the
effects of these conditions) are necessary in order
to compare SCA readings from different contexts.
Parameters of SCA
The hypothetical 20-second SCA record in Figure 3.18 yields three primary and two secondary
parameters. Similar descriptions of measurement
and typical waveforms appear in Stern, Ray, and
Quigley (2001).
Primary Parameters
SCL or Tonic Level
SCL expressed in micromhos represents a baseline
or resting level. Although this level may change,
in a resting, quiescent person it is likely to hover
around a value identified as the tonic level. SCL
or tonic level is thought to be an index of baseline
level of sweat gland activity, an inferred indication of a relative level of sympathetic arousal. For
example, conductance values above 5–10 micromhos are thought to be relatively high, whereas
those below 1 micromho are thought to be low.
Remember that these estimates depend on a number of other variables and should be taken only as
a rule of thumb based on the use of 3/8-inch dry
electrodes on the volar surface of fingertips.
SCR or Phasic Changes
Phasic changes are noticeable episodes of increased
conductance caused by sympathetic arousal gen-
60
erated by a stimulus. For example, in the case of
the stimulus introduced after 5 seconds, there is a
1- or 2-second delay, then an increase in conductance that peaks, levels out, and falls back to the
baseline or tonic level. This is a phasic change, and
its magnitude (height) is expressed as the number
of micromhos reached above baseline. The size of
phasic changes is thought to be an indication of the
degree of arousal caused by stimuli (e.g., a startle or
orientation to novel internal or external stimulus).
SCR Half‑Recovery Time
“SCR half-recovery time” is defined as the time
elapsed from the peak of the phasic change to
one-half of the way back down to baseline. SCR
half-recovery time is thought to be an index of
a person’s ability to calm down after a transitory
excitation. It has been hypothesized that persons
with chronic overarousal may have difficulty
returning to relaxed baselines after even minor
stimulation.
Secondary Parameters
“SCR latency” is defined as the time from stimulus onset until the beginning of an SCR. “SCR
rise time” is defined as the time elapsed from the
beginning of an SCR to its peak. These parameters have carried little significance in biofeedback;
therefore, they are not discussed in detail here.
Normative Values for the Parameters
The hypothetical SCA record in Figure 3.18 shows
specific values for the parameters. These values are
II. INSTRUMENTATION
actual mean values taken from normative samples
of SCA records for tropical nonpatients, summarized in Venables and Christie (1980). However,
these are not necessarily representative of values
obtainable in ordinary biofeedback practice. Since
large individual differences in SCL and SCR are
common, readings far different from those cited in
Figure 3.18 should come as no surprise. Furthermore, potential sources of normative variation
include differences between patient and nonpatient groups, the effects of medications on SCL
and SCR, differing procedures for establishing
baselines and especially SCRs, and the great differences in instruments and electrodes likely to be
used. For a discussion of the effects of such variables as temperature, humidity, time of day, or season, see Venables and Christie (1973, 1980).
My advice to the reader is this: To increase
your confidence in norms, find or build normative
samples specific to the instruments you are using
and to the populations with which you are working. At this time, there is no solid substitute for
systematically accumulated experience with your
own patient group, purposes, and equipment. This
is not meant to be discouraging to the clinician or
disparaging to the field; it is only a reflection of
the present state of the art.
Scales and Measurement: The “Percentage
Increase” Scale for SCR Amplitude
Displaying SCR amplitude as an increase in the
number of micromhos is not the only alternative.
SCR amplitude can also be expressed as a percentage change from the tonic level. For example, an
SCR consisting of a 1-micromho change from 3 to 4
FIGURE 3.18. Parameters of skin conductance. SCA values shown are taken from Venables and Christie (1980).
3. A Primer of Biofeedback Instrumentation
micromhos is expressed as a 33% change. This has
the effect of relativizing the SCR to the baseline
from which it occurs. With this method, a change
from 6 to 8 micromhos is also a 33% change, and so
is a change from 1.5 to 2.0 micromhos.
The rationale for this scale is the assumption
that a given increase in autonomic arousal leads to
a given percentage increase in conductance over
the baseline level, and that this holds for all baseline levels. The following hypothetical examples
and the electrical model of the skin illustrate this.
Imagine that 200 sweat glands are turned on, giving an SCL of 2 micromhos. Now a stimulus comes
along that turns on an additional 100 sweat glands,
thus leading to a 1-micromho or a 50% increase.
Now imagine another case in which there are 600
sweat glands turned on for an SCL of 6 micromhos.
According to the percentage model, a stimulus with
the same arousing properties as in the first case will
lead again to a 50% increase in conductance by
turning on an additional 300 sweat glands, for a
3-micromho increase in conductance.
The assumption here is that changes in arousal
are better gauged as percentage increases in conductance over existing baselines than as absolute
increases in conductance with no regard to initial baselines. This is analogous in the economic
domain to expressing a year’s growth in the gross
domestic product as a percentage increase over the
previous year’s level, rather than as an increase in
the number of dollars.
Loudness perception provides a second analogy:
Achieving a given increase in perceived loudness
takes a larger absolute increase in loudness above
a noisy background level than above a quiet background level. If an SCR is some sort of “orienting
response,” it is plausible that to be psychophysiologically “noticeable,” a stimulus must lead to
a significant increase in conductance relative to
existing baseline arousal—parallel to what occurs
in loudness perception.
Pitch perception supplies a third analogy. The
difference in pitch between the note C and the
note A above it sounds the same in any octave. (It
is the musical interval of a sixth.) The difference
between middle C (256 Hz) and the A above it
(440 Hz) is 184 Hz, a 72% increase in frequency.
The difference between the next C (512 Hz) and
the next A (880 Hz) is 368 Hz, but it is also a 72%
increase in frequency. In this case, the percentage
increase in frequency, rather than the number of
vibrations per second, leads to the perception of
equal increases in pitch.
61
The absolute-micromho increase scale for SCR
amplitude rests on an assumption opposite to that
of the percentage increase scale, namely, that a
micromho increase in conductance indicates a
given increment in arousal, no matter where it
is observed on the continuum of possible initial
baselines—a fixed increment of arousal, regardless
of initial baseline. This assumption is also plausible.
There are, to my knowledge, no published data
or definitive conceptual arguments to support or
disconfirm either of the assumptions presented
earlier. Each of these scales has plausibility and
appeal, and it is apparently yet to be discovered
whether either has distinct practical advantages
or greater psychophysiological appropriateness.
However, I prefer the assumptions supporting
the use of the percentage increase scale for SCR
amplitude. This is because the method of relating the magnitude of changes to initial baselines
is appropriate and useful in perceptual contexts
that to me are analogous to SCR. In addition, my
informal observations suggest that persons with
low SCL baselines often show fewer micromhos
of SCR than persons with average SCL baselines.
For me, intrinsic plausibility and these informal
observations tip the balance toward the percentage increase scale for SCR amplitude. However,
at very high SCLs, the percentage increase scale
probably loses appropriateness, because most of
the available sweat glands are already turned on to
make the high SCLs.
Convenient scaling follows from the percentage
increase scale assumption. If the skin conductance
continuum is plotted along a line, a logarithmic
scale conveniently contains all possible SCL values while retaining a useful degree of resolution
for SCRs all along the line. This scale is illustrated
in Figure 3.19. It has the advantage of providing
adequate resolution at the low end while avoiding
excessive resolution at the high end. Recall that
the percentage increase scale supposes that the
difference between 1 and 2 micromhos is more
significant than the difference between 10 and
11 micromhos, and is equivalent to the difference
between 10 and 20 micromhos. On the logarithmic scale, equal distances along the line represent
equal percentage changes. That is, the distance
from 1 to 2 is the same as that from 10 to 20;
both are 100% changes. This means that an SCR
amplitude of any given percentage is represented
by the same distance along the line, regardless of
initial baseline.
62
II. INSTRUMENTATION
Skin Conductance Record Interpretation
The three primary parameters discussed earlier
help professionals describe actual skin conductance records and extract data from them. But
because records usually contain compounded
changes in both responses and levels, interpretation is often required to specify values for the
parameters. Below are paradigmatic descriptions
of complex skin conductance records and interpretive hypotheses.
Upward Tonic Level Shift
The sample record in Figure 3.20 reveals a phasic
change away from the beginning tonic level and
incomplete return to that level. Think of this as
an SCR that did not recover and led to a new and
higher tonic level from which subsequent phasic
changes depart. A hypothesis is that whatever
arousal led to the phasic change did not completely “wear off,” thereby leaving the person with
a new and elevated tonic level. Increase in conductance may be slow like “drift,” rather than rapid
like a typical SCR.
Downward Tonic Level Shift
The arousal leading to the new or elevated tonic
level discussed above may in time “wear off” or be
“relaxed away,” leading to a downward trend in
skin conductance. As shown in Figure 3.21, this
record has downward slope to it, although SCRs
may be superimposed. In this way, a new lower
tonic level may eventually be reached.
Stairstepping
With multiple excitatory stimuli, especially for
persons who show high-magnitude phasic changes
and slow recovery time, a phenomenon called
“stairstepping” may occur. As shown in Figure
3.22, this results when an excitatory stimulus
occurs before the phasic changes from previous
stimuli have had time to return to the prior tonic
level. The SCA may then “stairstep” higher and
higher. This stairstepping process could theoretically be implicated in development of overarousal.
Figure 3.23 illustrates how individuals who show
lower magnitude phasic changes and more rapid
return to baseline are less susceptible to stairstepping from repeated stimulation.
Nonresponsive Pattern
A “nonresponsive pattern” is an unusually flat conductance level (see Figure 3.24), which does not
respond to typically arousing stimuli even when
there is a reason to believe that arousal or emotion
is or should be present. A hypothesis for this pattern, when extreme, is inappropriate detachment,
overcontrol, or helplessness rather than relaxation
(Toomim & Toomim, 1975).
FIGURE 3.19. Logarithmic scale for SCA values.
FIGURE 3.20. Upward tonic level shift.
63
3. A Primer of Biofeedback Instrumentation
FIGURE 3.21. Downward tonic level shift.
FIGURE 3.22. Stairstepping.
FIGURE 3.23. Rapid return to baseline, reducing stairstepping.
Optimal Skin Conductance Patterns
Skin conductance is linked to arousal, but optimal SCA patterns are not necessarily the lowest or
flattest patterns. This is because persistent minimal arousal, overcontrol, inattention, or flattened
affect is not usually considered healthy or adaptive.
There is a time for minimizing arousal during deep
relaxation, in which a steady, low level of skin conductance may be desired, but uniformly invariant
or flat levels are not necessarily desirable.
For example, encountering a novel stimulus
calls for recognizing and treating it appropriately.
Habitual blunting of the arousal associated with
orientation or action is not thought to be healthy
or adaptive. However, after a person orients to
the novel stimulus and takes appropriate action,
arousal should drop to baseline levels, avoiding
unnecessary arousal or wasted energy. It is possible
for a person to react too vigorously to novel stimuli, so that the reaction is out of proportion. In this
case, the person is treating stimuli as more alarming, dangerous, or exciting than warranted, and is
paying a price in energy and physical tension.
SCA is not something to be minimized but
something to be optimized, and this requires judgment about what is appropriate for a given person
in a given circumstance. At this time, no one
64
II. INSTRUMENTATION
FIGURE 3.24. Nonresponsive pattern.
claims to know optimum tonic levels and SCRs,
or to be able to show that there is any such thing
as specifiable optimums. What is clear is that it is
possible to have overreaction and underreaction,
and that this holds for both the tonic levels and
phasic changes. Quick return to baseline after an
SCR may be consistently desirable except when it
is part of an underresponsive pattern.
Because of large individual differences in SCA
patterns and the lack of normative data under
various standard paradigms of stimulation and
measurement, it is difficult to specify clear and
widely accepted procedures for relaxation training with SCA. Useful SCA biofeedback requires
experience and judgment on the part of the clinician. The best way to acquire the “feel” of how
SCA works under various conditions is to observe
it within and between individuals, especially oneself. Those who work regularly with SCA are often
quick to point out its ambiguities and uncertainties, but, undiscouraged, are also eager to discuss
its unique responsiveness to transitory emotional
states and thoughts. Its apparent complexity and
ambiguity may conceal a wealth of valuable psychological as well as physiological information to
those who have the patience to learn and further
describe its patterns.
Operation of the Skin Conductance Instrument
Most Basic Constant DC Voltage Scheme
Figure 3.25 shows the most basic SCA monitoring
scheme. A constant voltage is impressed across
the two electrodes. The variable resistance of the
skin leads to a variable current through the circuit. A current amplifier monitors this current,
and, through proper scaling, drives a display that
reads out in micromhos or microsiemens. In this
most basic form, it is similar to temperature instruments, as shown in Figure 3.14. However, to be
practical, it must be refined.
SCL baselines are spread over a wide range, yet
it is important to distinguish small SCRs (e.g., a
5% change from any SCL) from all possible baselines. As an illustration, if the entire range of possible SCA values were made to fit on a meter face,
SCL values would show, but an SCR would barely
deflect the needle. This is the familiar issue of
“resolution,” discussed earlier in connection with
temperature biofeedback instruments. A digital
FIGURE 3.25. Most basic SCA monitoring scheme.
65
3. A Primer of Biofeedback Instrumentation
meter overcomes resolution problems simply by
having enough digits (e.g., 10ths or even 100ths).
However, a digital meter is not suitable for observing SCRs, because changing digits during an SCR
are hard to read. In contrast, the swing of a meter
needle or light bar up and then back down or computer display is much more meaningful for SCRs.
Such challenges are addressed in the designs of
EDA instruments. A description of how this was
typically addressed in early electrodermal biofeedback equipment appears in earlier editions of this
volume.
Simple SCR Devices
Simple SCR devices use a manual, noncalibrated
baseline adjustment and feed back SCR with an
audio tone or noncalibrated meter scale. These
devices quantify neither SCL nor SCR and are
more susceptible to artifact than full-sized instruments. Even so, they are very convenient and provide very interesting and useful information to a
person about patterns of SCR.
For example, the rise and fall of an audio tone
communicates a great deal about the person’s
responsivity in actual situations, even when quantified SCL or SCR is absent. These devices have
distinct advantages when it comes to ambulatory
use in real life. Pocket-size miniaturization, dry
finger electrodes, and an earplug for private feedback permit a person to wear the unit conveniently
while walking, talking, driving, phoning, writing,
thinking, reading, or carrying out other real activities. This provides insight into patterns of responsivity in active situations that are not obtainable
in the clinic setting. It is a very good way for a
person (including the therapist!) to discover his or
her own patterns of responsivity. In any application, the therapist involved must provide adequate
instruction in the use and limitations of the device
and in the interpretation of results.
Artifact
The following points about artifact are important
for all SCA devices.
Therefore, electrode size must be standardized in
order to assure comparability of quantified SCL
readings.
Movement
Because electrode size affects SCA, anything that
alters the effective contact area of an electrode
also alters SCA. Finger or hand movement causes
variations in contact pressure. The electrode may
lift slightly and diminish the contact area, or press
harder against the skin and increase the contact
area. These effects are more pronounced for dry
electrodes than for precious metal electrodes with
electrode gel. The practitioner should encourage
the monitored person to minimize hand movements and arrange the electrodes and cables for
a reasonably stable position. When hand movement cannot be avoided (e.g., monitoring while
the person is doing something with both hands),
corresponding sites on the toes could be used. This
would require exploring new norms for SCL and
SCR on those sites. Fortunately, movement artifact is usually easy to spot, because the resultant
patterns are often abrupt and uncharacteristic of
true SCA patterns, and because movement can
often be observed.
Skin Condition
Skin condition can affect conductance readings.
For example, if a person has a skin abrasion or a
fresh cut through the high-resistance skin surface,
a high-conductance path may be established from
the electrode to deeper layers of the skin and lead
to an increased SCL. If a person has developed a
callus, the high-resistance surface layer increases
in thickness and dryness, leading to a much lower
SCL and diminished SCR amplitude. Venables
and Christie (1980) note that SCL falls markedly
after a washing with soap and water, as residual
salt is removed. Because salt builds up over time
since the last wash, they recommend that persons
begin sessions with freshly washed hands. It is not
clear how important this is to clinical biofeedback,
but it is clear that this standardizing procedure is
not universally followed.
Electrode Size
Different-size electrodes lead to different readings.
A larger electrode covers more skin and therefore
places more sweat glands in the current loop. This
leads to a higher SCL than does a smaller electrode that places fewer sweat glands in the loop.
Room Temperature
There is some evidence (Venables & Christie,
1980) that SCA is affected when individuals feel
cold, and that warmer-than-usual office conditions
appear to produce what they call more “normal”
66
II. INSTRUMENTATION
responsivity. It is also plausible that the temperature-regulating function of sweating in an overly
warm room leads to increases in SCL that are not
psychophysiologically significant.
A similar prolongation may occur in very humid
climates even when dry electrodes are used. Artifactual prolongation of SCR recovery could lead
to results mistakenly interpreted as “stairstepping.”
Electrode Polarization Potentials
and Electrode Design
Use of AC to Control Electrode
Polarization Artifact
The exosomatic method involves the passage of
current through the skin via surface electrodes.
Polarization potentials develop at the skin–electrode interface as DC passes, and the polarization
effect builds up over time. The size of polarization
potential is variable and unknown. EDA units
have historically varied widely in their susceptibility to the effects of electrode polarization. But
in general, this is probably not a major problem,
especially with DC instruments that apply very
small electrical currents to the skin. Nevertheless, biofeedback clinicians who are interested in
EDA and the devices that have been employed to
assess it over the years should probably be aware
of the issues concerning electrode polarization and
methods that have been used to minimize it. A
somewhat technical discussion of this follows.
Dry electrodes are often used for EDA. They
are made from various materials, including lead,
zinc, chrome, stainless steel, gold, or silver-coated
fuzz, and are often secured by Velcro straps that
conveniently adjust to different finger sizes. They
are simpler, cheaper, and more convenient than
silver/silver chloride electrodes, especially in clinical practice. However, when used with DC EDA
equipment, the simple dry electrodes suffer from
polarization potentials to various degrees.
When polarized, the skin–electrode interface
is like a tiny battery charged by the passing current. Polarization voltage is thereby added to (or
subtracted from) the constant voltage applied by
the instrument. Because the polarization potential
(voltage) is variable, the voltage in the current
loop is no longer constant. Therefore, what appear
to be changes in SCL may be due in part to variable electrode polarization potentials. Drift in skin
conductance level due to the buildup of polarization potential causes artifact, but this effect may
not be all that significant, for practical purposes.
Nevertheless, silver/silver chloride electrodes
have sometimes been used, because they develop
minimal polarization potentials and therefore add
minimal polarization artifact. But they are more
expensive and less convenient than dry electrodes,
and the gel used with silver/silver chloride electrodes may prolong the recovery phase of SCRs.
Instrument designs have been evolved to circumvent the effects of polarization potentials. Use of
an AC rather than a DC current loop is described
in previous editions of this volume but is less relevant to biofeedback practitioners now than in
previous times.
Safety
Electrical safety precautions for SCA devices are
the same as those for EMG devices. Both are electrically connected to the person via electrodes;
therefore, the same stringent standards for design,
manufacture, installation, and maintenance
should be followed for SCA and EMG devices, and
for the entire installation of which any of these
instruments are a part. The passage of DC from
an electrode to the skin over a prolonged time may
lead to the formation of chemical by-products on
the skin if the voltage drop across the skin exceeds
about 3 DC volts, such as might be encountered
in “toy” or very early EDA gizmos (Leeming, Ray,
& Howland, 1970). This effect is normally negligible, but if the current passed is high enough and
is passed long enough, then skin irritation could
develop. This effect is unlikely to occur in modern skin conductance instruments, but very old
units, those that were made as novelties or toys, or
those that have developed leakage currents, may
be more likely to create this effect. As a rule of
thumb, a device that passes current of 10 microamperes or less per square centimeter of electrode
area in its current loop, and applies under 3 DC
volts to the skin, will not lead to the accumulation
of irritating chemicals on the skin.
Acknowledgments
My thanks to the late Wallace A. Peek, the late Roland
E. Mohr, and John B. Picchiottino, who have acted so
generously as my engineering mentors. I give special
thanks to John B. Picchiottino, whose suggestions for
this chapter in its first edition marked a long and muchappreciated history of helpfulness with biofeedback projects. Special thanks must also go to Mark S. Schwartz,
without whose enthusiasm the first edition’s chapter and
3. A Primer of Biofeedback Instrumentation
subsequent editions would doubtless have remained on
my list of things to do someday. Thanks must also go to
the late Peter G. Ossorio, founder of Descriptive Psychology, for his mentorship on how to clarify language
and definitions in emerging fields and how to use analogies in explaining technical subject matter.
References
Bolek, F. (2013). Digital sampling, bits, and psychophysiological data: A primer, with cautions. Applied Psychophysiology and Biofeedback, 38(4), 303–308.
Boucsein, W. (2012). Electrodermal activity (2nd ed.). New
York: Springer Science and Business Media.
Brown, B. (1974). New mind, new body: New directions for
the mind. New York: Harper & Row.
Cacioppo, J. T., Tassinary, L. G., & Berntson, G. G. (Eds.).
(2000). Handbook of psychophysiology (2nd ed.). New
York: Cambridge University Press.
Dawson, M. E., Schell, A. M., & Filion, L. (2000). The
electrodermal system. In J. T. Cacioppo, L. G. Tassinary,
& G. G. Berntson (Eds.), Handbook of psychophysiology
(3rd ed.). New York: Cambridge University Press.
Jennings, J. R., Tahmoush, A. J., & Redmond, D. D. (1980).
Non-invasive measurement of peripheral vascular activity. In I. Martin & P. H. Venables (Eds.), Techniques in
psychophysiology (pp. 70–131). New York: Wiley.
Leeming, M. N., Ray, C., & Howland, W. S. (1970). Low-
67
voltage, direct-current burns. Journal of the American
Medical Association, 214(9), 1681–1684.
Mathieu, P. A., & Sullivan, S. J. (1990). Frequency characteristics of signals and instrumentation: Implication for
EMG biofeedback studies. Biofeedback and Self-Regulation, 15(4), 335–352.
Neumann, E., & Blanton, R. (1970). The early history
of electrodermal research. Psychophysiology, 8(4), 463–
474.
Otis, J., Rasey, H., Vrochopoulos, S., Wincze, J., & Andrasik, F. (1995). Temperature acquisition as a function of
the computer-based biofeedback system utilized: An
exploratory analysis. Biofeedback and Self-Regulation,
20(2), 185–190.
Stern, R. M., Ray, W. J., & Quigley, K. S. (2001). Psychophysiological recording (2nd ed.). New York: Oxford University Press.
Toomim, M., & Toomim, H. (1975, February). Psychological dynamic correlates of the paradoxically invariant GSR.
Paper presented at the fifth annual convention of the
Biofeedback Research Society, Monterey, CA.
Venables, P. H., & Christie, M. J. (1973). Mechanisms,
instrumentation, recording techniques, and quantification of responses. In W. F. Prokasy & D. C. Raskin
(Eds.), Electrodermal activity in psychological research
(pp. 111–116). New York: Academic Press.
Venables, P. H., & Christie, M. J. (1980). Electrodermal
activity. In I. Martin & P. H. Venables (Eds.), Techniques
in psychophysiology (pp. 3–67). New York: Wiley.
Chapter 4
Advanced Topics in Surface Electromyography
Instrumentation and Applications
Jeffrey E. Bolek, Ronald L. Rosenthal, and Richard A. Sherman
Surface electromyography (SEMG), as a science,
has an uncanny ability to lure the unsuspecting
into believing that the modality is much simpler
than it really is. Unfortunately this leads to three
outcomes: First, those new to the field suddenly
find that what appeared to be a healthy application of SEMG with their patient in reality was
not; second, often on the heels of the point just
listed is the clinician’s understandable abandonment of the field of SEMG for other, more efficient
methods to use with patients; and third, for those
who stick with it, there is the tendency to rely on
canned “protocols” or “menu driven” applications.
These applications often have catchy titles such as
“pelvic floor dysfunction protocol.” Unfortunately
they are typically a one-size-fits-all arrangement,
and the clinician places great trust that the creator
of the program had enough clinical background
to consider the myriad permutations involved in
using SEMG with the patient group. This chapter is designed to shed some light on all of these
points. In the first section, Jeffrey E. Bolek reviews
some supposedly “basic” SEMG principles that are
often misunderstood because they are typically
given cryptic definitions, if they are defined at all.
He also suggests a novel way of viewing SEMG
data from multiple sites, a form of “quantified
EMG,” but for surface, not needle electromyography. In the second section, Ronald L. Rosenthal
then dives into the intricacies of neuromuscular
retraining by providing a road map for hands-on
applications in stroke rehabilitation, with special
emphasis on muscle synergies, pattern training,
and timing. Finally, in the third section, Richard
A. Sherman discusses techniques for optimizing
the results of rehabilitation biofeedback.
Interpretation of the SEMG Signal
When one studies the rich and convoluted interaction
between the many factors that influence the information
content of the EMG signal, it is reasonable to ask if there
is any hope of using the EMG signal in a constructive
fashion. . . . The answer is a confident “yes” for some . . .
and a guarded “maybe” for other applications.
—De Luca (1997, p. 141)
In 1793, Volta proved that dissimilar metals in
contact with an electrolyte (like those present in
body tissue) could generate an electric current. In
1849, DuBois-Reymond was the first to report the
detection of voluntarily elicited electrical signals
from human muscles (Basmajian & DeLuca, 1985).
The terms “electrical” and “current” are included
in these discoveries. All too often their meaning
in electromyography is obscured by theoretical
definitions at best, or mere mention in a glossary
at worst. This is unfortunate, because it tends to
cover an important part of electromyography with
a shroud of complexity that few have the engineering background to understand. The key to under68
69
4. Advanced Topics in Surface Electromyography
standing electromyography is a working knowledge
of the fundamentals of electricity, because it is the
foundation upon which electromyography is built.
My (J. E. B.) goal in this section is to present the
origin, travels, and end processing of the SEMG
signal in a technically comprehensible language,
followed by further elucidation on the SEMG signal as an indicator of functional performance.
There are two kinds of current, direct (DC)
and alternating (AC). In DC, the electrons travel
down a wire in one direction, and the direction
remains constant over time. In AC, the current
changes polarity constantly from positive to negative at the very high rate of 60 hertz (Hz) or 60
cycles per second. AC current produces a sine wave
as the current quickly changes from positive to
negative (Figure 4.1).
Likening this to water flowing in a pipe, the
“current” is the rate of flow of the water, and the
“voltage” is the “push,” the force that makes it
move. Current is the number of electrons flowing past a given point per second and is measured
in amperes. Every circuit (whether wire or motor
neuron) has some resistance, which is measured
in ohms. “Watts” is the measure of power (like a
100 w light bulb). Voltage (or the “push”) can be
measured by multiplying the amperage times the
resistance. Amperage (or flow of electrons) can be
measured by dividing volts by resistance and watts
(power) can be found by multiplying voltage times
current. SEMG is the measurement in microvolts
of an action potential propagating down a motor
neuron. A “motor neuron” is a nerve cell that has
a motor function. The amplitude dimension of
SEMG is measured in microvolts (mV); the frequency (because it is a sine wave) in hertz (Hz),
or cycles per second. If SEMG were a string on an
instrument, the amount of space the string travels
when plucked is “voltage”; the speed that it travels
while moving in that space is “frequency.” If one
considers a vocalist singing a song, SEMG amplitude is to volume as change in pitch is to SEMG
frequency. The vocalist can sing loudly or softly
(amplitude) while singing in a given pitch (frequency). Similarly, a muscle may fire using more
or less slow (20–90 Hz) or fast (90–500 Hz) twitch
fibers. Three events may account for the increase
in an SEMG reading: Additional motor units may
“kick in,” the motor units currently in play may
fire more quickly, or both. The frequency component allows one to see another dimension of the
SEMG signal, that is, what is behind the increase–
decrease in amplitude. A group of muscle fibers is
supplied by one motor neuron, with some neurons
supplying only a few muscle fibers (e.g., laryngeal)
and other serving many (e.g., gastrocnemius)
(Yolanda et al., 2007). This nerve cell plus its
long axon running down the motor nerve, along
with the terminal branches, constitute the “motor
unit.” The muscle fibers it innervates can range
Voltage
Period (T)
AMPLITUDE
Time
Wavelength
FIGURE 4.1. A pictorial representation of a sine wave.
70
II. INSTRUMENTATION
from 3 to 2000 (Buchthal & Schmalbruch, 1980).
SEMG measures the electrical activity in microvolts (millionths of a volt or mV, not to be confused
with millivolts, mV, thousandths of a volt) of all
activated motor units or a whole muscle located in
the vicinity of the skin via surface electrodes. The
duration of electrical potentials of motor units
ranges from a few to 14 milliseconds (ms); amplitude can range from 0.01 to 5 microvolts (Soderberg, 1992). An action potential(s) is given off by
the motor unit(s) activated during a muscle contraction. The usual resting electrical charge that is
negative on the axon interior and positive on the
exterior is reversed as the wave of action potential
passes at speeds of one-fourth of a .22 caliber rifle
shot or 320 feet/second (Cottman & McGaugh,
1980). These action potentials last about 1 ms and
quickly repolarize, causing movement along the
axon. An impulse descending the motor neuron
causes all the innervated muscle fibers to contract.
A muscle consists of hundreds of muscle fibers,
depending on size, grouped together in motor
units in which each motor unit is innervated by
a single motor nerve. When a signal is conducted
via the nerve fiber servicing a motor unit, all the
muscle fibers of that unit contract simultaneously.
Selective activation of these different motor units
is used as a means to control muscle contraction,
with the electrical activity of active motor units
being detected by the electrodes. All this motor
unit activity is called motor unit action potential
(MUAP). Raw SEMG produces a sine wave (Figure 4.2) as the signal flows from + to – with fluctuating peaks and troughs. Notice how irregular the
signal is in Figure 4.2 compared to that in Figure
4.1. This is because that in Figure 4.1 is one frequency, 60 Hz, and the raw SEMG is made up of
many frequencies. One cannot just take the average of the values, because that would represent a
snapshot at a point in time, not across time. DC
produces no wave; it is a constant value and simple
to measure. Myoelectric signals (e.g., SEMG) are
more challenging to interpret because they are
bipolar and produce a sine wave. The only way to
measure the value of AC (and SEMG) is by some
calculation of the sine wave, the root mean square
(RMS), which is the foundation on which SEMG
is transformed from raw data to an intelligible display but is rarely explained in detail. It is a method
of taking a signal that flows above and below the
zero gradient line on the monitor (i.e., is bipolar)
and making it monopolar (never dips below zero).
One has a constantly moving signal over time, so
the positive and negative values do not occur at
the same time. Because of the fluctuating nature
of the sine wave, one has to devise a method that
enables a direct comparison with a nonoscillating energy source (e.g., comparing AC to DC).
We want to know how big a nonoscillating energy
source (e.g., DC) needs to be to deliver the same
energy as the sine wave (as in an AC source, like
the current in SEMG) in a set amount of time. The
RMS produces a value that is an approximation of
the energy about halfway through the peak and
valley. The SEMG waveform resembles AC in that
it swings from negative to positive, back and forth.
The RMS voltage is a measure of the magnitude of
a set of numbers; it gives one a sense for the typical
size of the numbers, and displays this information
in an “amplitude envelope.” For example, 1000
data points may be sampled consecutively from
the raw data. RMS is performed by squaring the
data (which removes the negative signs), summing
the squares, determining the mean (central tendency), and taking the square root (which reverses
the effect of squaring). If one were to measure the
“average” voltage of the 120-volt AC coming into
the house, it would be zero. Why, then, does one
get a shock from 120-volt AC? The average voltage
is zero, but the average power is not zero, it is 120
FIGURE 4.2. Raw SEMG sine wave.
4. Advanced Topics in Surface Electromyography
volts RMS, and this is what the power meter is displaying. In the same way, one cannot just take the
“average” raw SEMG reading.
Why is the myoelectric signal measured in
microvolts? Earlier I stated that power is measured
in watts and voltage is merely the “push” that
moves electrons past a given point in time, measured in amperage. Would not power (in watts)
or current (in amps) be a better indicator? All
electrical energy has two main characteristics:
voltage and current. Recall that in SEMG, three
electrodes are used (see Peek, Chapter 3, this volume). The two active electrodes pick up the voltage difference and send it to the encoder. Recall
that to find current (in amps) we need to know the
resistance in the body, which is very difficult to do,
so SEMG uses the voltage difference as the next
best choice, although it is not a very “clean” signal.
This is why efforts are being made to develop new
means to analyze the SEMG signal (Abdallah &
Zahran, 2009).
An important (and confusing) dimension of
the SEMG signal is that SEMG can be displayed
as two domains: The amplitude and frequency
domains and frequency can be displayed as a
power spectral density (PSD) or median frequency
in time (MFT). The amplitude domain just discussed is a time-based measure of signal amplitude
(or effort) with time on the x-axis and amount of
effort (in mV) on the y-axis. This is the familiar
real-time display observed as a moving line on the
monitor. The PSD frequency domain is displayed
as frequency (think “pitch” in hertz) on the x-axis
and power (think “loudness”) on the y-axis (this
is PSD). Since the PSD does not include the time
dimension, one must specify the time interval
(called an “epoch,” usually .5 to .2 seconds) used in
the analysis (called an FFT, or fast Fourier transformation analysis). The MFT displays time on the
x-axis and Hz on the y-axis but not power. Neither
display is “better”; it depends on the kind of data
desired. Therefore, the SEMG signal can be mined
for different types of information depending on
the needs of the clinician. For example, as fatigue
accumulates, the frequency tends to shift downward as a result of a reduced conduction velocity of
the action potential, while the amplitude remains
the same or increases.
The sequence of events to the microvolts displayed on the monitor is as follows: A group of
muscle fibers is innervated by the axon of the
motor unit, creating action potentials that flow
through (and can be attenuated by) skin and adipose tissue before reaching the electrode. From
71
there the signal is sampled, filtered, quantified/rectified (typically by RMS) before finally producing a
digital signal in millivolts. “Sampling” is a means of
managing the digital data obtained from the raw
SEMG. Unlike analog signals (e.g., varying light
intensity), the digital signal processing (DSP) age
born in the 1960s ushered in the ability to produce bits of data that can be stored and manipulated in a computer. Unlike our base 10 number
system (from 10 fingers), computers use multiples
of eight (8 bits = 1 byte), so samples are typically
eight items or multiples thereof (e.g., 16, 32, 64 . . .
2048). Phonograph records are an example of analog data, in that a continuous stream of music is
recorded. Compact discs (or iPods) play music that
has been sampled off the continuous music stream
at 44,100 times per second, producing “digital
data.” Research has shown that the sampling rate
must be at least twice the maximum frequency to
be expected (for human hearing, 20,000 Hz). Five
hundred hertz is often considered the limit of the
usable SEMG range, which results in a sampling
rate of 1000 Hz, which is the upper limit on many
SEMG units. The range of frequencies capable of
being processed by the typical SEMG unit is 20 to
1000 Hz, with the majority of EMG activity being
below 200 Hz and very little found above 500 Hz.
About 80% of the signal resides between 30 and
80 Hz (Cram, 1991). Twenty hertz and below is
where movement artifact resides (De Luca, 1997).
If a high-pass filter is set for 20 Hz (allowing only
designated high frequencies to pass), movement
artifact will be eliminated, but so will some of
the spectrum of the internal oblique, which lives
in the 8- to 150-Hz range. Error is inevitable by
design in digital sampling, because the instantaneous snapshot of the analog data has to be
rounded off to the nearest available digital value
to make the sample. There is a limited number of
binary digits (or bits) that can be used to sample a
signal. The “bits” indicate the number of decimal
points any given sample can possess; the greater
the number of bits, the more detail the sample
contains. The “sampling rate” defines how many
samples are taken over X amount of time.
Filters are used to eliminate energy radiating
from the atmosphere (e.g., power lines) or energy
from the body itself (e.g., heart rate) from contaminating the signal. But this energy (in the
50- to 60-Hz range) also contains desirable parts
of the EMG signal (Winter, 1990). Furthermore,
every frequency has a fundamental frequency and
a “harmonic,” which is multiples of that frequency
(for 60, 120, 180, 240 Hz, etc.), so the filter will not
72
II. INSTRUMENTATION
filter these (Kasman, 1995). In addition, tissues
separating the muscle fibers from the electrodes
act as a low-pass (allowing low and blocking some
high hertz) filter, depending on tissue thickness
(De Luca, 1979), while the differential electrode
acts as a high-pass filter allowing high hertz to pass
and blocking low hertz (Barry, 1991).
From this discussion, we see that the value displayed on the monitor (e.g., 35.6 mV) represents
an approximation of the recruitment level of the
targeted muscle. The signal has been filtered by
the electrode, skin, and adipose tissue; sampled to
enable analog-to-digital conversion; and filtered
to cut out undesirable frequencies (along with
desired ones), with parameters that exist within a
bandwidth and are transposed from a raw to RMS
signal! After all the filtering, sampling, bandwidth
limitation, and approximation to create the RMS,
the 35.6-mV value on the screen no longer looks so
exact. The greater the number of decimal points
in a number, the greater the tendency for humans
to place unearned confidence in it. RMS is at best
a rough but useful indicator of what is going on
at the level of the muscle fibers. The frequency
domains have problems, too, in that at low contraction forces, the signal-to-noise ratio is very
low (i.e., more noise in the signal) (Baratta, Solomonow, Zhou, & Zhu, 1998).
Given the convoluted nature of the SEMG signal, Bolek (2006, 2012) created a means whereby
the relationship between the observed SEMG
activity and function could be quantified, termed
quantitative surface electromyography (QSEMG).
In QSEMG, the practitioner must begin with an
understanding of how the muscles involved in a
movement act synergistically, because training on
a single muscle site is not sufficient to make meaningful changes in function. Training goals are set
for up to 10 muscles (five bilaterally), and the combination of recruitment/relaxation is rewarded, for
example, by turning on–off a video. The key point
is that the patient does not seek to control a single
group of agonist–antagonist muscles but to produce a patterned response that produces a reward.
This is similar to the z-score used in neurofeedback in which multiple variables are monitored at
more than one brain site. The motor control score
(MCS) is a measure of performance over time
based on meeting or exceeding the recruitment
goal for the targeted muscle and as such is based
on an approximation, the RMS. For example, for
each 0.5-second sample of RMS, that value is compared to a set threshold with a two-tailed decision
tree: s ≥ x or s ≤ x, where s is the signal and x is the
threshold setting. Multiple sites are used, so longer
decision trees (formulated in Excel) are common.
The MCS is the total time demonstrating proper
muscle recruitment divided by the total time at
work. Success over X amount of time can be plotted as in Figure 4.3.
In motor reeducation, a premium is based on
the use of SEMG in facilitating functional gain.
Most clinicians target at most two muscles, but
attending to the myotatic unit (or constellation
of muscles) involved in a targeted skill leads to
faster and more enduring learning (Sella, 2000).
A “myotatic unit” is a functional unit or group of
agonist and antagonist muscles (four in the case
18
Seconds thresholds met
16
14
12
10
8
6
4
2
0
1
384
767 1150 1533 1916 2299 2682 3065 3448 3831 4214 4597 4980 5363 5746 6129 6512 6895 7278 7661 8044 8427 8810 9193 9576
X.H. 3/6/09
Time of sessions in 1/8 seconds
FIGURE 4.3. A functional display of patient progress.
73
4. Advanced Topics in Surface Electromyography
below) that function together as a unit or have
close functional relationships (Travell & Simons,
1993). For example, targeting the bilateral sternocleidomastoid (SCM) to be above threshold and
bilateral cervical paraspinals to be below creates
a program that facilitates mastering head control
post–motor vehicle accident (MVA), in that it is
sensitive to establishing a neutral head position
in all four planes (flexion, extension, right and
left lateral shifts; Bolek, 2006). Cram, Kasman,
and Holtz (1998) noted the cooperative function
between the SCMs and C4 paraspinals during
head rotation, with the SCMs active during rotation and the C4 paraspinals providing a stabilizing
function.
Consider Figure 4.4, which displays the RMS of
wrist extensor and wrist flexor over the course of
a 50-minute treatment session with threshold-contingent signaling based on wrist extensor values
being ≥ 16.6 and ≤ 51.8 and flexor values in the
range of ≥ 15.8 and ≤ 49.2 (RMS). The threshold
directions (recruit–relax) are set by the clinician
and are visual assessments of the performance of
the patient. In other words, this is incremental
muscle training to use the wrist extensor while
allowing for wrist stabilization with the flexor
bundle.
Figure 4.3 presents a composite picture of the
same data called a motor control score (Bolek,
2012). The 9,778 samples obtained from both sites
over the course of the session were placed into a
decision tree in Excel, where “go” = 0.125 second
if, and only if, the extensor and flexor values listed
earlier are met. In other words, for each sample of
0.125 second, if both criteria are met, a “credit” of
0.125 second is listed; if criteria are not met, it is
a “no go” and the credit is zero. The numbers on
the x-axis are time of the session beginning at the
left; on the y-axis, the total consecutive seconds
is added where both values are met. For example,
early in the session, there is a period of 15 consecutive seconds of both flexor and extensor meeting the criteria for reward, with a slow but definite
trailing off as the session progressed. Figure 4.3 is
a visual display of the functional progress made by
the patient. It is a coupling of the data from both
SEMG channels over time, designed to reflect the
effect of the SEMG values (the independent variable) on the functional progress (the dependent
variable) of the patient, and adds information
to the treatment program that would not be easily ascertained by only viewing the SEMG RMS.
Here we see nearly 16 consecutive seconds with
the extensors between 16.6 and 51.8 and flexors
between 15.8 and 49.2. It is a typical graph of
motor learning with a few minutes to “find” the
correct motor plan, followed by the best performance, then by a trailing off as fatigue sets in.
Consider a program in which the reward is contingent on five sites, with some of the target muscles set to be recruited and others to be relatively
relaxed. This patient, with a diagnosis of cerebral
palsy, tended to stand on her right leg with the
left leg externally rotated and slightly flexed. The
RMS of specific sites and the planned direction
of recruitment are shown in Figures 4.5–4.10. We
see a slight but nonsignificant increase in the left
gluteus maximus, a variable left adductor, a small
decrease in the left hamstring, and no change on
the left medial quadriceps but more variability
than on the left lateral quad, which, again shows
little change on SEMG RMS. Overall the SEMG
FIGURE 4.4. RMS of wrist extensor (lighter) and wrist flexor (darker).
74
II. INSTRUMENTATION
70
60
µV
50
40
30
20
10
0
1
206
411
616
821
1026 1231 1436 1641 1846 2051 2256 2461 2666 2871 3076 3281 3486 3691 3896 4101 4306 4511 4716 4921 5126
Time of session in 1/8 seconds
FIGURE 4.5. RMS of left gluteus maximus.
80
70
60
µV
50
40
30
20
10
0
1
206
411
616
821
1026 1231 1436 1641 1846 2051 2256 2461 2666 2871 3076 3281 3486 3691 3896 4101 4306 4511 4716 4921 5126
Time of session in 1/8 seconds
FIGURE 4.6. RMS of left abductor.
80
70
60
µV
50
40
30
20
10
0
1
206
411
616
821
1026 1231 1436 1641 1846 2051 2256 2461 2666 2871 3076 3281 3486 3691 3896 4101 4306 4511 4716 4921 5126
Time of session in 1/8 seconds
FIGURE 4.7. RMS of left hamstring.
75
4. Advanced Topics in Surface Electromyography
60
50
µV
40
30
20
10
0
1
206
411
616
821
1026 1231 1436 1641 1846 2051 2256 2461 2666 2871 3076 3281 3486 3691 3896 4101 4306 4511 4716 4921 5126
Time of session in 1/8 seconds
FIGURE 4.8. RMS of left medial quadriceps.
21.5
21
20.5
µV
20
19.5
19
18.5
18
17.5
17
16.5
1
209
417
625
833
1041 1249 1457 1665 1873 2081 2289 2497 2705 2913 3121 3329 3537 3745 3953 4161 4369 4577 4785 4993 5201
Time of session in 1/8 seconds
FIGURE 4.9. RMS of left lateral quadriceps.
Seconds thresholds met
90
80
70
60
50
40
30
20
10
0
1
206
411
616
821
1026 1231 1436 1641 1846 2051 2256 2461 2666 2871 3076 3281 3486 3691 3896 4101 4306 4511 4716 4921 5126
COMPOSITE MCS
Time of session in 1/8 seconds
FIGURE 4.10. Composite motor control score.
76
RMS results show either a slight change in the
desired direction or no change. The MCS reveals
what would have been seen by an observer present in the session. There is a gradual rise in the
consecutive seconds the patient met the threshold criteria, beginning with negligible time at the
start of the session to nearly 80 seconds toward the
end. The MCS is statistically richer in information, in that it pools all the data obtained from the
SEMG and operationally displays it on a graph. It
can be considered a form of QEMG that is limited in application to needle EMG in that it mines
the SEMG data to perform a quantitative analysis.
Every treatment session involves setting the bar for
expected performance (i.e., during a typical session
the clinician will be giving some kind of feedback
as to the performance of the patient, such as “good
job” or “don’t give up, keep trying!”). By looking
only at the SEMG RMS values, particularly when
many sites are involved, it is often difficult for the
patient to read what is on the monitor and for the
clinician to determine how to interpret the SEMG
readings and construct a treatment plan. QSEMG
can be a useful tool in giving immediate feedback
and quantifying this judgment call.
Neuromuscular Retraining:
General Considerations
Biofeedback systems have evolved considerably
over the past several years. Computerized systems
priced at under $10,000 typically record at least
four channels of EMG, and systems that can record
eight or more channels are available. Because
these systems can sample over 2000 times/second,
they are capable of recording the raw signal at
bandwidths from 20 to 1000 Hz. The software can
provide power spectra for multiple channels in real
time with median frequencies.
The software for these systems can program
virtual channels and provide feedback based on
thresholds for multiple channels. This gives the
clinician a powerful tool to help clients learn to
change motor recruitment patterns. Rather than
being constrained to train for increases or decreases
in one muscle at a time, feedback can be determined by changes in the recruitment patterns of
multiple muscles. This is important, because even
simple movements often involve an organized pattern of recruitment of many muscles. Impairments
in movement or posture are rarely attributable to
improper activation of one muscle; rather, there is
II. INSTRUMENTATION
a disruption in the timing and level of activity of
all of the muscles involved in the activity.
An early example of this approach was in
the work of Will Taylor (1990) on the scapular
rhythm. Taylor realized that normal movement
of the arm in abduction is dependent on rotation of the scapula as the arm is raised. This is
brought about by the simultaneous activation
of the upper and lower trapezius muscles. As the
upper and lower trapezius muscles fire, the scapula rotates laterally and the glenoid fossa moves
in an upward direction. Without this action, the
humerus would bump into the acromion and the
range of motion would be restricted. For many
patients with shoulder problems, the lower trapezius become inhibited and full range of motion in
shoulder abduction becomes painful and difficult.
Biofeedback training to increase lower trapezius
recruitment in abduction is often very beneficial
for these patients. An effective way to accomplish
this training is to construct a virtual “proportional
channel,” defined as the amplitude of the lower
trapezius divided by the sum of the amplitudes
of the upper and lower trapezius. This channel
ranges from 0 to 1 and is .5 when the two signals
are equal. The setup involves recording from the
upper and lower trapezius and the medial deltoids,
and instructions could include trying to keep the
shoulder down as the arm is raised. Initially, the
threshold for the proportional channel can be
set quite low, in the range of .1 to .2, and auditory feedback is provided when the medial deltoid is above its threshold, the lower trapezius is
above its threshold, and the proportional channel
is above its threshold. If the balance of the upper
and lower trapezius shifts too much toward the
upper trapezius, the proportional channel will fall
below threshold and turn off the auditory reward.
As training progresses, all of the thresholds can be
raised until the amplitudes approach full normal
and the balance of the upper and lower trapezius is
nearly equal, with the proportional channel consistently at or above .45.
Muscle Pattern Training
Scapular training can also be conducted for problems in shoulder flexion. In this case, the target
muscle is the serratus anterior, below the axilla, or
armpit. The serratus anterior rotates the scapula
in an anterior direction, permitting full range of
shoulder flexion that would be impossible without movement of the scapula. Biofeedback train-
4. Advanced Topics in Surface Electromyography
77
ing should emphasize the joint recruitment of the
anterior deltoids and the serratus anterior as the
arm is moved forward and up.
Another example of muscle pattern training
involves the knee. The quadriceps, which extends
the knee and is also important in supporting the
body weight while walking, has a lateral component (the vastus lateral, or VL) and a medial component (the vastus medialis oblique, or VMO).
When the two muscles generate equivalent forces,
the diagonal vectors cancel out and the patella
remains in a central position in the trochlea. In
some cases, the patella can be pulled out of the
trochlea, which is called “patellar subluxation.”
Invariably, the dislocation is in the lateral direction and a crucial aspect of conservative management of patellar subluxation is strengthening of
the VMO. Biofeedback has often been used as
an adjunctive modality, and the protocol should
include simultaneous recording of the VMO and
VL so that VL recruitment does not simultaneously increase while the VMO is uptrained. Proportional feedback is an excellent option, because
it provides feedback that is sensitive to the balance
of the two muscles. The training should include
seated knee extension, an open-chain task, as well
as work in standing and weight bearing, a closedchain task.
A final example of muscle pattern training
with biofeedback is in hand rehabilitation. SEMG
training with hand function can be difficult and
demanding. In contrast with the lower extremity, the forearm and hand contain a large number
of muscles, many of them small and overlapping.
Functional use of the hand also requires the coor-
dination of muscles that control the wrist and fingers. Biofeedback training to improve hand functioning that uses only one or two channels runs
the risk of training inappropriate muscle activity
due to the possibility of volume conduction and
overflow from adjacent muscles.
Fortunately, with judicious placements, it is possible to get reasonably well-isolated signals from
the major extensors and flexors of the wrist and
fingers (Figures 4.11 and 4.12). Distal placements
on the dorsal and ventral aspects of the forearm,
2–4 cm from the wrist crease, can still detect substantial activity of the finger extensors and flexors.
These distal placements are relatively insensitive
to activation of the wrist muscles. More proximal
placements near the elbow respond to the wrist
extensors and flexors. These proximal placements
respond to finger extension and to finger flexion as
well, but if the distal sites are low, then one can be
relatively confident that the signals are from the
wrist muscles.
Finger extension and flexion both depend
upon the simultaneous recruitment of the wrist
extensors and flexors to stabilize the wrist joint,
permitting the fingers to open or close. In the
power grip, the wrist muscles must maintain the
wrist in a position of moderate extension. Without
this stabilization, the finger flexors would flex the
wrist and the fingers would extend by a mechanical action called “tenodesis.” Similarly, in finger
extension, the wrist extensors and flexors must
maintain the wrist in a neutral or slightly flexed
position. Biofeedback training for the power grip
should emphasize the simultaneous recruitment
of the finger flexors along with the wrist extensors and flexors. There is considerable variability
in the balance of the wrist extensors to the wrist
FIGURE 4.11. Ventral placements to record surface
EMG activity from the wrist flexors and finger flexors.
FIGURE 4.12. Dorsal placements to record surface
EMG activity from the wrist extensors and finger
extensors.
78
flexors, but in most cases it is beneficial to increase
the level of extensor recruitment.
In addition to the extrinsic muscles in the forearm, it is also possible to conduct biofeedback
training with some of the intrinsic muscles of the
hand (Lertprapamongkol & Suksathien, 2007).
There are three sets of muscles in the palm of the
hand, the dorsal interossei, the palmar interossei,
and the lumbricales. While SEMG electrodes cannot distinguish among these three groups, use of
SEMG can be enhanced with the addition of sites
on the dorsal surface of the hand, between the
metacarpal bones. This placement is responsive
to activation of the lumbricales, and these muscles
are critical for full extension of the fingers. An
effective protocol is to combine two channels of
intrinsic activity (one from between the middle
and ring fingers and a second one between the
index finger and thumb) with the extrinsic finger
flexors and extensors. Training for finger extension would reward the simultaneous recruitment of
both intrinsic sites along with the extrinsic extensors. Care should be taken to observe the patient
carefully because the intrinsic sites will respond to
activation of the palmar interossei muscles should
the patient go into the power grip.
Biofeedback for Stroke Rehabilitation
Biofeedback has a long history in the treatment
of motor dysfunction after stroke (Basmajian,
Kukulka, Narayan, & Takebe, 1975). Initial reports
revealed improvement in ankle dorsiflexion with
biofeedback training and in a number of studies in
the 1980s, and benefits in lower extremity functioning (and to a lesser degree in upper extremity
functioning) with biofeedback were demonstrated
in the 1990s. However, in recent years, there have
been almost no studies of EMG biofeedback with
stroke patients, and it appears that many clinics
specializing in neurorehabilitation do not routinely offer biofeedback to stroke patients. It is difficult to account for this trend, and many factors
are most likely involved. One important factor is
a lack of training in SEMG biofeedback for most
physical and occupational therapists. As a result,
therapists may not appreciate the potential benefits of adding SEMG training to their therapeutic
protocols.
In order to implement a program using SEMG
retraining to enhance the motor recovery of a
stroke patient, it is necessary to understand some
basic principles of motor control, as well as the
deficits brought about by a stroke. Strokes typi-
II. INSTRUMENTATION
cally cause “hemiplegia,” a unilateral paralysis of
the side of the body contralateral to the location
of the stroke. The deficits may be relatively mild or
they can be so profound as to lead to a complete
inability to use the affected limbs in any functional activities. A simplistic notion of hemiplegia
would be to think of it as generalized weakness,
analogous to lowering the volume on a radio, but
without distorting the basic patterns of movement.
Regrettably, this is far from the case. Deficits in
motor control after a stroke include spasticity,
hypertonicity, ataxia, and several other conditions reflecting damage to central motor control
systems.
The complex deficits found in motor function
subsequent to a stroke reflect the hierarchical
nature of motor control. Movement is organized at
multiple levels within the central nervous system.
There are primitive, reflexive patterns of wholelimb movement that are organized at the spinal
level. In the midbrain, we see the emergence of
righting reactions involved in equilibrium and
balance. Finally, in the cortex, we find the development of more sophisticated control systems
that are able to organize movements into complex
functional activities. The higher motor systems
exert a strong inhibitory influence on the lower
systems. Strokes can disrupt this inhibitory activity, leading to ineffective motor recruitment patterns and the emergence of more primitive wholelimb synergies. From a biofeedback perspective,
several factors need to be addressed. These include
hypertonicity, hyperactive stretch reflexes leading to spasticity, alterations in temporal patterns
of motor recruitment, and the presence of wholelimb synergies.
Hypertonicity and Spasticity
Muscle tone is regulated at multiple levels to balance the needs of stability and mobility. Increased
muscle tone leads to more stability, but at the cost
of less fluidity of movement. Conversely, when
muscle tone is too low, mobility is enhanced, but
there may be significant postural instability. In
intact individuals, muscle tone is regulated, so that
there is enough activity to stabilize the trunk and
limbs but not so much as to interfere with fluid
movements. In stroke patients, the disruption of
descending motor pathways may lead to hypertonicity and spasticity. It is more common to see it
in flexors, although increased tone from extensors
can also occur. Common sites affected include the
biceps, wrist flexors, and pectoralis major. Muscles
79
4. Advanced Topics in Surface Electromyography
involved in neck rotation and flexion may also be
a factor.
Biofeedback training for hypertonicity and spasticity is straightforward and consists of inhibitory
training of the involved muscles. However, it is not
clear that most stroke patients with excess motor
tone or spasticity will respond favorably to such
training. Initial efforts to downtrain the hyperactive
stretch reflex of the biceps using biofeedback were
largely unsuccessful, despite the use of thousands of
feedback trials in some cases (Wolf, personal communication, 2014). However, Wolf and Segal (1990)
reported modest success in downtraining the spastic stretch reflex of the biceps in elbow extension.
Thus, the findings are not consistent and there may
be some combinations of patients and protocols that
are effective in reducing spasticity.
A common observation when working with
stroke patients is that motor recruitment will
increase dramatically when resistance is provided.
Isometric training can be effective in increasing
motor recruitment of the triceps or wrist extensors, and the absence of movement eliminates the
antagonistic corecruitment caused by disinhibition of the stretch reflexes. However, if the training does not include a progression to concentric
movements without resistance, then it will be
unlikely to lead to any functional improvement.
The question of whether biofeedback training
can be effective in reducing tone and spasticity of
stroke patients may become moot because of the
growing use of botox injections in stroke rehabilitation. Botox can be used in a graded manner to reduce or totally block the recruitment of
spastic flexors. The injections are effective for up
to 6 months and must be readministered or the
dysfunctional tone will recur. Biofeedback training can be combined with botox injections, and it
can be advantageous to train for increased extensor recruitment while the flexors are “chemically
quiet.” It will be interesting to see whether research
supports the efficacy of a combined approach.
Synergies
Synergies are a normal part of the process of recovery from a stroke. “Synergistic movements” are
stereotyped patterns of movement across multiple
joints that are activated as a unit. These patterns
are commonly observed as the acute stroke patient
moves from a flaccid state into a more spastic
state. There are both flexor synergies and extensor
synergies, and they have been well described in a
number of sources (e.g., Brunnstrom, 1970). Flexor
synergy is usually dominant in the upper extremity, while extensor synergy is stronger in the lower
extremity.
Effective biofeedback programs for stoke rehabilitation must take into account the presence of
synergies. For instance, a common problem during
ambulation after a stroke is poor heel strike, and
many stroke patients need an ankle–foot orthosis
(AFO) to compensate for poor control of ankle
dorsiflexion and eversion. If dorsiflexion training
is conducted while seated, the stroke patient may
show fairly high levels of anterior tibialis recruitment if allowed to flex and externally rotate the
hip. In order to be functional, dorsiflexion and
eversion must be sustained throughout the swing
phase, until the foot hits the ground. An effective
biofeedback program must address this need. Dorsiflexion training should progress from a seated
position, which allows the flexor synergy, to standing and stepping with a proper heel strike. An
effective intermediate protocol gradually trains a
combination of anterior tibialis recruitment with
activation of the quadriceps to extend the knee.
This combination is out of synergy and is quite difficult for most stroke patients. Mastery of this task
marks a major progression in the recovery from a
stroke and is a good prognostic indicator for continued improvement in the gait pattern.
Timing Changes In Motor Recruitment
In addition to the previously mentioned changes
in motor recruitment patterns, stroke also plays
havoc with the exquisite timing of muscular activation and relaxation in most skilled movements.
Intact individuals are able immediately to recruit
muscles as needed, and motor activity falls back
to baseline levels rapidly when the movement no
longer requires the activation of a given muscle.
Patterns of EMG activity in rapid alternating
movements are often too fast for basic biofeedback
displays, and it is not uncommon to ask a patient
to slow down when recording baseline patterns
during alternation between wrist extension and
wrist flexion if a nice display is desired.
In contrast with the rapid activation and inhibition of muscles in normal movement, stroke
patients show a delay in muscle recruitment during
movement and a lag in the return to baseline levels after the completion of a given movement. This
is particularly troublesome for rhythmic activities,
such as walking, which require alternating patterns
of motor recruitment as the legs move from the
stance phase to the swing phase and back. Stroke
80
patients often strain to accomplish difficult motor
tasks, increasing tone and making it difficult to
inhibit motor activity that is no longer required.
Biofeedback training that focuses solely on static
patterns of motor recruitment is less likely to be
as effective as programs that emphasize the ability
to activate and inhibit appropriate motor activity quickly. Auditory feedback during ambulatory
training can be programmed to help patients learn
to terminate activity more rapidly as needed.
Sensory Changes and Their Role
in Motor Dysfunction
In intact individuals, the central nervous system
constructs a spatial map of the body derived primarily from the stretch receptors in the muscles.
This map enables an individual to be aware of how
the various body parts relate to each other without
looking at them. This sensory ability was given
the label “proprioception” by noted physiologist
Charles Sherrington in 1906. Sherrington distinguished proprioception from the sensory inputs
such as vision and audition, which were defined
as exteroceptive inputs, as well as from visceral
sensations, which were defined as interoceptive
inputs. Proprioception is basically a positional
sense and it is closely related to the sensations of
movement (kinesthesia) and balance. Proprioceptive and kinesthetic abilities allow us to eat our
breakfast safely while reading the newspaper. They
also permit visually impaired individuals to move
about freely.
Proprioception is processed in the cerebellum and in the parietal lobes. Strokes can have a
major impact on proprioception, and many stroke
patients report little or no intrinsic awareness of
their affected limbs when questioned. Proprioceptive deficits may be related to the problem of
neglect, in which stroke patients may have limited
awareness of objects in the field of vision contralateral to their lesion. They may also neglect body
parts on the affected side, washing only half of
their body and putting shoes or socks on only one
foot.
The presence of significant proprioceptive
deficits will have a major impact on the results
of a biofeedback training program with a stroke
patient. Patients need to feel the movements
trained with EMG feedback. Without an intrinsic
awareness of what they have accomplished with
feedback, patients are very unlikely to be able to
repeat the movement on their own. In my own
(R. L. R.) work with stroke patients, I have seen
II. INSTRUMENTATION
a young woman successfully open her hand while
doing finger extension training, only to look down
and gasp at what she had accomplished after seeing that her fingers were indeed straight.
The experimental findings of Edward Taub
(1977) with deafferented monkeys are relevant to
this discussion. Taub reported that monkeys would
fail to use an arm after surgical transaction of the
sensory roots at the level of the spinal cord. However, if the nonoperated arm was immobilized with
restraints, the monkeys would soon start to use the
deafferented limb. There are a number of differences between a stroke victim and a monkey with
a surgically deafferented limb, but Taub’s findings
support the notion that primates are capable of
functional use of a limb in the absence of proprioceptive feedback.
Perhaps the best recommendation that can be
made is to try to facilitate the recovery of proprioception. The affected limbs can be moved
passively, and the patient can alternate between
watching the movement and trying to sense the
movement with the eyes closed. Patients can be
shown how to range and stretch their affected
limbs by themselves, and the simultaneous feedback from the intact hand and the impaired limb
may serve to improve the sense of movement and
position.
Techniques for Optimizing the Results
of Rehabilitation Biofeedback
Effective incorporation of biofeedback into the
rehabilitation process requires an in-depth knowledge of biofeedback techniques and devices, as
well as numerous associated techniques. Therapists need to understand the underlying pathology
causing the problem, how the muscles and other
physiological systems relate to the problem, limitations of the recording technology, and how to
set the psychophysiological recorder so that it will
provide the required information.
Very often therapists need to compare signals
from two sets of “paired” muscles performing the
same function, such as when recording from (1)
the left and right paraspinals of the low back during a low back evaluation, (2) the left and right
vastus medialis and lateralis (which control kneecap stability) during a subluxation of the patella
evaluation, (3) the left and right upper trapezius
(in the shoulders) during a headache evaluation,
and (4) the major muscles of the residual limb during a phantom limb pain evaluation. It is crucial
4. Advanced Topics in Surface Electromyography
to remember that a difference of even a few centimeters in location of the sensors over the muscle
(including angle with respect to the direction the
muscle runs and distance between the active sensors) or a small difference in impedance between
the sets of sensors can result in huge differences in
the microvolt levels from the two sets of muscles.
Ignoring these factors also leads to very low test–
retest reliability when relating changes in muscle
tension to changes in pain over a period of weeks,
so one needs to be sure the sensors are placed in
virtually exactly the same spot and have similar
impedance at each recording.
Even with the best sensor placement and skin
preparation, one is not likely to obtain similar
readings from most sets of laterally paired muscles,
because the muscles are not likely to be the same
size. For example, the mass of paraspinal muscles
can normally differ up to 20%, so the signals
may differ up to 20%. For this reason, teaching a
patient to match the levels of tension in order to
decrease pain due to sustained muscle tension can
be ineffective unless the differences are greater
than about 20%. The patterns of SEMG activity recorded from different muscles (e.g., abdomen and bicep) appear different as the muscles
go through a cycle of tensing and relaxing. When
paired muscles (e.g., the left and right paraspinals
of the low back) are recorded, their patterns of
activity should be virtually the same regardless
of minor differences in muscle tension. Thus, it
is far more important to train the patterns to be
the same rather than attempt to train the muscles’
tension levels to differ less than 20%.
Patterns of muscle activity between two paired
muscles frequently appear very different, because
the muscles are not actually performing the same
activities at the same time and to the same extent.
Nearby muscles whose tension is recorded along
with the muscles of interest due to cross talk, even
when sensors are placed close together along the
length of the muscle of interest, may be doing different accessory activities during the recording.
Thus, it is important to control the motion of
interest to the fullest extent possible and to observe
very carefully how the patient is moving to guide
the motion toward being as bilaterally symmetrical
as practical. For example, when comparing muscle
patterns during an evaluation of subluxation of the
patella, the motion of both legs has to be carefully
controlled, as does change in the angle of the feet
during the motion, so that the patterns observed
on the monitor will be reflective of symmetrical
motions. This is usually accomplished by guiding
81
the motion with the assistance of a moving foot
rest on the chair in which the patient sits.
When working with a set of muscles for the first
time, it is frequently not obvious which muscles
are contributing to the abnormal situation, such
as incorrectly sustained muscle tension that occurs
with jaw musculoskeletal pain or incorrect patterns of motion that occur with subluxation of
the patella. Just what a normal pattern of activity
should look like is also not easy to determine when
the practitioner is not familiar with the area being
evaluated. Typical kinesiology books (e.g., Lippert,
2000) are profusely illustrated and go into great
detail about which muscles contribute to which
motions, as well as the order in which the muscles
tense during different phases of the motion. For
example, when evaluating pain and incontinence
in the pelvic floor, it is crucial to record from the
lower abdomen, because incorrect tension in these
large muscles can easily overwhelm whatever the
pelvic floor muscles are doing to maintain continence. Authors such as Kasman, Cram, and Wolf
(1998) have published books showing many of the
typical normal and abnormal patterns of muscle
tension that are likely to be seen during an evaluation. Sometimes SEMG recording technology
simply does not adequately record the signals from
muscles that may well be contributing to a pain
problem without mixing the signal with those
from larger, overlaying, uninvolved muscles running in different directions and having different
functions from the muscle of interest (Wolf, 1980).
Setting the psychophysiological recorder correctly for recording and feeding back muscle activity is crucial. An area frequently overlooked is
setting the bandwidth (the frequencies recorded
by the amplifier) wide enough to encompass all
of the muscle’s relevant activity. If the bandwidth
is set too narrow (e.g., 100–200 Hz), the muscle’s
power spectrum (the amount of power at each frequency) may change so much as the muscle’s tension changes from relaxed to very tense that much
of the power may be outside the recording window
during some parts of the tensing continuum. This
leads to the false conclusion that the muscle is far
less tense (by as much as 30%) than it actually is
at that level of tension, because the power is never
detected (Sherman, 2004). Many patients are
incorrectly taught that their muscles are relaxed
when they feel that they are tense.
Numerous studies have shown that patients
with chronic pain are less able to align how tense
they think the muscles in painful areas are with
actual levels of tension than either (1) people
82
who are pain free or (2) patients with pain who
have muscles that are pain free (Flor, Schugens,
& Birbaumer, 1992). This inability to calibrate
actual levels of tension with sensations correctly is
especially important among patients with musclerelated jaw area pain and low back pain, because
these patients tend to permit their muscles to
remain tense for so long that pain can result.
Because a key part of muscle rehabilitation is to
help patients correctly align sensations of tension
with actual levels of muscle tension, it is vital not
to confuse the situation with recordings that do
not reflect actual levels of tension.
The biofeedback display has to be set correctly
to facilitate learning. For example, if the device is
set to “autogain” (in which the software controls
the gain for each channel recorded), with each
channel of muscle tension responding independently, differences in tension are obscured when
the machine keeps switching the gain for each
channel to keep the signal about the same size
on the display. Thus, channels may have very
different levels of amplification, each of which
changes independently during a single motion.
Two muscles that appear to have the same amount
of tension because the signals are the same size on
the monitor may have very different levels of tension. Relative patterns of tension among several
muscles contributing to a motion are obscured if
the gain for each muscle’s display changes differently throughout the course of the motion. The
therapist will probably notice the rapid changes
in gain because the numbers on the vertical line
indicating range of microvolts keep changing, but
the patient is not likely to be able to track these
rapid changes and incorporate them into learning
to compare the signals.
The sweep speed for each channel not only
needs to be the same, so that the patterns of activity appear similar, but it also has to be appropriate
to the situation being recorded. For example, if the
patient is being asked to bend and rise as part of
a low back evaluation or bend and straighten the
legs as part of a subluxation evaluation, the sweep
speed needs to be set so that several cycles of activity can be seen clearly on the screen. This permits
patient and therapist to observe differences in patterns immediately as the patient is coached to correct the motion and tension patterns.
Great care needs to be taken to adjust the
amount of signal averaging or “smoothing,” so that
important data are not lost. Essentially, the longer
the integration time constant, the less momentary
changes in the signal are reflected in the display.
II. INSTRUMENTATION
When a signal is averaged too much (too long a
time constant), brief spasms can be obscured or
take so long to show on the monitor that the spasm
is long gone before it is observed. This prevents
the patient from pairing the sensations accompanying the spasm with the signal displayed on the
monitor. For example, when treating cramping
phantom limb pain, the patient must observe the
spasms causing the pain just as the spasms occur.
Otherwise, the patient rarely learns to prevent
the spasms, which means the treatment is far less
effective or fails (Sherman, 2004).
Too much smoothing or averaging of a signal
essentially produces a delay in the feedback. When
doing muscle retraining, any delay in feedback
increases the number of times the task needs to be
repeated before it is learned, therefore decreasing
the effectiveness of the training.
A proper assessment is crucial to knowing
what information to feed back to the patient. For
example, only about 7% of amputees with phantom pain are helped when treatments are not
aligned with the underlying muscle and blood
flow problems, while about 80% are helped when
treatments match the underlying problem (Sherman, 2004). Also, a crucial part of any rehabilitation assessment is teasing out the role of stress in
maintaining and intensifying muscle tension. For
example, without conducting a psychophysiological stress profile, it may be impossible to relate the
contribution of stress to changes in muscle tension to changes in low back pain. It is common
for patients’ generalized musculoskeletal stress
responses to result in muscles all over their bodies being chronically tense and fatigued. Learning
to recognize specific patterns of stress responses
related to onset of pain may be impossible if a
patient’s overall muscle tension obscures specific
reactions. A course of general relaxation training may be required before specific training can
be effective. Heart rate variability feedback may
be especially relevant when chronic sympathetic
arousal underlies pain-related problems, such as
irritable bowel syndrome and non-cardiac-based
chest pain.
An important part of the evaluation is knowing
what to look for, beyond the psychophysiological
assessment. For example, the contributions of posture to tension headache and stress to muscle tension-related jaw area pain should not be ignored
(Middaugh, Kee, & Nicholson, 1994). Practitioners
need to know which common psychological and
physical tests might be useful, along with psychophysiological assessments, in designing an optimal
4. Advanced Topics in Surface Electromyography
course of treatment. For example, incorporation of
cognitive restructuring may be of value in helping patients with irritable bowel syndrome who
reveal features of conversion disorder (Blanchard,
Greene, Scharff, & Schwatz-McMorris, 1993).
Trigger points need to be evaluated, because many
muscle-related pain problems appear to be intensified by trigger points (Dommerholdt, Bron, &
Fransson, 2006). The practitioner performing the
evaluation need not be an expert in treating trigger points. Rather, the professional needs to be
proficient enough to recognize them, so that an
appropriate referral can be made. For example,
piriformis syndrome, in which spasms in the hip’s
piriformis muscle put pressure on the sciatic nerve,
is a common cause of low back pain. Unless the
piriformis is checked for problems, including trigger points, during a low back pain evaluation, the
underlying cause will be missed.
Biofeedback training in rehabilitation is frequently performed in a manner very different
from that used with other applications. Especially for muscular rehabilitation, considerable
use is made of traditional operant conditioning
techniques, such as shaping and multitrial learning (Brucker, Maiatico, & Pauza, 1992; Brucker,
& Bulaeva, 1996). Development of unconscious
control through repetition of specific small tasks
taken in steps is emphasized. Biofeedback devices
are typically used, with their thresholds set so that
it is easy to reach an immediate goal. When that
level is reached, the threshold is reset so that the
next goal can be strived for and met. Thus, muscle
tension is gradually shaped to the correct pattern
of activity. The authors believe that operant conditioning is more effective than guided practice
and similar techniques for muscular reeducation.
Stated more directly, practitioners using biofeedback in rehabilitation need to understand the
principles of operant conditioning in order to be
optimally effective in working with their patients.
Further information about using psychophysiological techniques for assessing and treating
patients whose primary symptom is chronic pain
may be found in Sherman (2004).
References
Abdallah, M., & Zahran, B. (2009). Analysis: Signal-step
analysis of surface myoelectric signal. European Journal
of Scientific Research, 26, 298–304.
Baratta, R. V., Solomonow, M., Zhou, B. H., & Zhu, M.
(1998). Methods to reduce variability of EMG power
83
spectrum estimates. Journal of Electromyography and
Kinesiology, 8, 279–285.
Barry, D. T. (1991). AAEM minimonography #36: Basic
concepts of electricity and electronics in clinical electromyography. Muscle and Nerve, 14, 937–946.
Basmajian, J., & De Luca, C. (1985). Muscles alive. Baltimore, MD: Williams & Wilkins.
Basmajian, J. V., Kukulka, C. G., Narayan, M. G., &
Takebe, K. (1975). Biofeedback treatment of foot-drop
after stroke compared with standard rehabilitation
technique: Effects on voluntary control and strength.
Archives of Physical Medicine and Rehabilitation, 56(6),
231–236.
Blanchard, E. B., Greene, B., Scharff, L., & SchwarzMcMorris, S. P. (1993). Relaxation training as a treatment for irritable bowel syndrome. Biofeedback and SelfRegulation, 18, 125–32.
Bolek, J. (2006). Use of multiple-site performance-contingent sEMG reward programming in pediatric rehabilitation: A retrospective review. Applied Psychophysiology
and Biofeedback, 31, 263–272.
Bolek, J. E. (2012). QSEMG: Quantitative Surface Electromyography: Applications in neuromotor rehabilitation.
Biofeedback: A Clinical Journal, 40(2), 47–56.
Brucker, B. S., & Bulaeva, N. (1996). Biofeedback effect on
electromyography responses in patients with spinal cord
injury. Archives of Physical Medicine and Rehabilitation,
77, 133–137.
Brucker, B. S., Maiatico, M. A., & Pauza, C. (1992). Long
term recovery in spinal cord injury with the use of biofeedback techniques. Archives of Physical Medicine and
Rehabilitation, 73(10), 960.
Brunnstrom, S. (1970). Movement therapy in hemiplegia.
New York: Harper & Row.
Buchthal, F., & Schmalbruch, H. (1980). Motor unit of
mammalian muscle. Physiological Reviews, 60, 90–142.
Cottman, C. W., & McGaugh, J. L. (1980). Behavioral neuroscience. New York: Academic Press.
Cram, J. (1991). Clinical EMG for surface recordings: Vol. 2.
Nevada City, CA: Clinical Resources.
Cram, J., Kasman, G., & Holtz, J. (1998). Introduction to
surface electromyography. Alexandria, VA: Aspen.
DeLuca, C. J. (1997). The use of surface electromyography
in biomechanics. Journal of Applied Biomechanics, 13,
135–163.
DeLuca, C. J. (1979). Physiology and mathematics of myoelectric signals. IEEE Transactions on Biomedical Engineering, 26, 313–325.
Dommerholdt, J., Bron, C., & Fransson, J. (2006). Myofascial trigger points: An evidence-informed review. Journal of Manual and Manipulative Therapy, 14(4), 203–221.
Flor, H., Schugens, M., & Birbaumer, N. (1992). Discrimination of muscle tension in chronic pain patients and
healthy controls. Biofeedback and Self-Regulation, 17,
165–177.
Kasman, G. (1995). Surface EMG and biofeedback in physical and behavioral medicine: Applications in chronic pain
management. Seattle, WA: Virginia Mason Medical
Center.
84
Kasman, G., Cram, J., & Wolf, S. (1998). Clinical applications in surface electromyography: Chronic musculoskeletal
pain. Gaithersburg, MD: Aspen.
Lertprapamongkol, W., & Suksathien, R. (2007). Electromyographic activities during isometric contraction of
interphalangeal joint extensors of the finger. Journal of
the Medical Association of Thailand, 90(8), 1657–1664.
Lippert, L. (2000). Clinical kinesiology for physical therapy
assistants (3rd ed.). Philadelphia: Davis.
Middaugh, S., Kee, W., & Nicholson, J. (1994). Muscle
overuse and posture as factors in the development and
maintenance of chronic musculoskeletal pain. In R.
Grzesiak & D. Cicconie (Eds.), Psychological vulnerability to chronic pain. New York: Springer.
Sella, G. E. (2000). Guidelines for neuromuscular re-education with SEMG/ biofeedback. Martins Ferry, OH:
GENMED.
Sherman, R. (2004). Pain assessment and intervention from a
psychophysiological perspective. Wheat Ridge, CO: Association for Applied Psychophysiology.
Sherrington, C. (1906). The integrative action of the nervous
system. New Haven, CT: Yale University Press.
Soderberg, G. L. (1992). Selected topics in surface electromyography for use in the occupational setting: Expert perspectives. Washington, DC: U.S. Department of Health and
II. INSTRUMENTATION
Human Services, National Institute for Occupational
Safety and Health.
Taub, E. (1977). Movement in nonhuman primates deprived
of somatosensory feedback. Exercise and Sports Science
Reviews, 4, 335–374.
Taylor, W. (1990). Dynamic EMG biofeedback in assessment and treatment using a neuromuscular re-education
model. In J. Cram (Ed.), Clinical EMG for surface recordings: Vol. 2. Nevada City, CA: Clinical Resources.
Travell, J., & Simons, D. (1993). Myofascial pain and dysfunction: The trigger point manual. Baltimore, MD: Williams & Wilkins.
Winter, D. A. (1990). Biomechanics and motor control of
human movement. New York: Wiley Interscience.
Wolf, S. (1980). Video comparing cadaver views of muscle
location with SEMG sensor placement. Recorded at
Emory University in Atlanta about 1980. Retrieved from
www.aapb.org.
Wolf, S., & Segal, R. (1990). Conditioning of the spinal
stretch reflex: Implications for rehabilitation. Physical
Therapy, 70(10), 652–656.
Yolanda, D., Ackah, H., Mandel, S., Manon-Espaillat, R.,
Abaza, M., & Sataloff, R. (2007). Laryngeal electromyography. Otolaryngological Clinics of North America, 40,
1003–1023.
Chapter 5
Cardiorespiratory Measurement and Assessment
in Applied Psychophysiology
Richard N. Gevirtz, Mark S. Schwartz, and Paul M. Lehrer
In this chapter, we provide an overview of cardiorespiratory physiology and the most common
measurement techniques used with some of their
metrics.
that separates the chest cavity from the abdominal cavity. It forms a flexible, moving floor for the
lungs. When the diaphragm is at rest, its shape is a
double dome, and it extends upward into the chest
under the lungs.
To start inhalation, the diaphragm contracts,
flattens downward, and descends. This allows the
lungs to fill. It displaces the abdominal contents,
expanding the belly. The natural return of the diaphragm to its resting state occurs with exhalation.
Other muscles involved in breathing include the
intercostal muscles (which act on the rib cage) and
the scalene muscles (which raise the chest by lifting the first and second ribs). In some cases, the
muscles of the abdominal area contract to push
the abdominal contents upward and push upward
on the diaphragm.
In the heart–lung system, special large molecules of a substance called hemoglobin carry
the fresh oxygen to every part of the body. Oxygen crosses the respiratory tubules called pulmonary alveoli. CO2 plays an important role in how
the hemoglobin releases the oxygen. As blood pH
changes, based on breathing changes, the hemoglobin molecule releases its oxygen cargo. If too
little CO2 is present, the oxygen is overbound to
the hemoglobin and not available to fuel body
organ tissue.
To understand this process more clearly, consider the analogy of a milk truck trying to deliver
Anatomy and Physiology
The Respiratory System
The respiratory system is among the most complex organ systems in the body. Descriptions of
the anatomy and especially the physiology of this
system are provided in medical textbooks (e.g.,
Guyton & Hall, 1995) and in abbreviated forms
in Fried and Grimaldi (1993) and Naifeh (1994).
A practitioner of applied psychophysiology needs
to be able to provide a basic explanation for a client; we therefore offer a rationale for respiratory
function that may be useful for such purposes.
Oxygen is taken in through the trachea or windpipe and pumped through a system of increasingly
smaller tubes, which have the characteristic of
letting some gases through to the blood and in
turn taking waste gases (mostly carbon dioxide,
or CO2) back to be exhaled. This process, called
“gas exchange,” takes place in he lungs. The lungs
themselves have no intrinsic muscles for breathing; instead, the diaphragm is the major muscle
for breathing. It is a sheet-like muscle stretching
from the backbone to the front of the rib cage
85
86
individual bottles of milk to local stores. The
oxygen is represented by the milk; the hemoglobin is the milk truck, and the store is the body
tissue needing fuel to function properly. The dairy
(the heart–lung system) loads the truck up with
an excess of milk bottles, and the truck sets off on
its appointed rounds. Once the truck arrives at the
store, the cargo door must be opened wide enough
to make an adequate delivery. Since CO2 controls
the release of the oxygen from the hemoglobin, it
would be seen as regulating the width of the door,
so that in the scheduled time enough milk can be
dropped off at the store. Not enough CO2 means
an inadequate delivery and shortages. In physiology, this oxygen dissociation function is known as
Bohr’s law. It says that the oxygen can be “overbound” to the hemoglobin, creating hypoxia, or
lack of oxygen, which can produce symptoms such
as lightheadedness, heart pounding, cold hands,
nervous emotional states, or even mental “fog.”
Hyperventilation (HV) means that the lungs are
releasing too much CO2, because breath rate and/
or tidal volume (the amount of air that is breathed
out) exceed the level needed for the current conditions. This is often referred to as “overbreathing,”
hyperpnea, or hypocapnia. The shift in blood flow
to the outer shell of the brain, where most complex
thought is processed, can be dramatic. (See www.
improve-mental-health.com/hyperventilation.html.)
HV can be obvious or subtle. The image of a nervous person breathing rapidly in his or her upper
chest and needing a paper bag is well known. Few
people realize that sighing or breath holding and
gasping can produce some of the same effects.
When someone overbreathes frequently, as
might happen in a very stressful period, the person
can create a condition in which his or her body
adjusts to the low level of CO2 and maintains it
with continuing sighs, yawns, and the like. When
this individual slows down his or her breathing,
the respiratory center in the brain may try to reestablish the previous CO2 levels by using the sighs,
yawns, or shallow breaths to blow off excess CO2.
Although this is usually not extremely dangerous, it does create a chronic condition with many
potential symptoms, such as dizziness, shortness
of breath, palpitations, and so forth. This chronic
overbreathing is sometimes called hyperventilation syndrome (HVS). A medical device called
a “capnometer” is sometimes used to measure the
amount of CO2 in the expired (exhaled) breath.
Normal levels are 40 millimeters of mercury (mm
Hg) or 40 torr. A level under 30 mm Hg or 30 torr
is usually considered low.
II. INSTRUMENTATION
It should now be clear that breathing patterns
are a powerful force within the body, and that
learning to alter these patterns can be a powerful
intervention for change.
Assessing HV
Methods for assessing HV include observations,
interviews, self-report questionnaires, HV provocation, blood assays, a transcutaneous instrument that roughly estimates CO2, and a noninvasive instrument that measures the percentage of
exhaled CO2. The last instrument is an infrared
gas analyzer called a “capnometer,” which produces a “capnograph.” Physiological monitoring
can be performed during rest and during office
stress challenges, while the patient is supine,
seated, or standing.
During HV provocation, one compares the
similarity of the symptoms during the provocation
with the presenting complaints. Voluntary HV
during provocation results in considerable individual variability of symptoms and patterns (Clark &
Hemsley, 1982; Fried, Fox, & Carlton, 1990; Fried
& Grimaldi, 1993), for adverse effects, dangers,
and contraindications for HV provocation).
Relying on patients to report all the symptoms
is insufficient. They often do not recall or report
many of the symptoms unless these are provoked
in the office. Observing unprovoked breathing
patterns in the office is also insufficient.
The Nijmegen Questionnaire (van Dixhoorn &
Duivenvoorden, 1985) is a self-report measure with
reasonable psychometric properties. It consists of a
short number of symptoms that have been found
to be associated with HVS.
Criteria for Diagnosing HV
Criteria for diagnosing HV vary, but typically
involve measures of CO2. Low CO2, less than
an average of 38 torr (less than 5% partial pressure of expired CO2: petCO2, or the percentage of
end-tidal CO2) at sea level is usually considered
a normal level. This criterion is independent of
symptoms. However, symptoms can emerge with
higher and lower levels of petCO2 (Fried, 1987;
Fried & Grimaldi, 1993). Practitioners typically
focus on symptoms rather than CO2 level. One
view (Fensterheim, 1994) is that clinicians want
to avoid missing a diagnose HV when it is present.
Fensterheim also agrees with Bass and Gardner
(1985) and Gardner (1994), who believe that no
symptom or clinical definition of HVS is widely
5. Cardiorespiratory Measurement and Assessment
accepted. Thus, clinicians can be less precise than
researchers, who wish to avoid including a person
without HV in a group with HV, and hence must
use a strict physiological criterion.
Some practitioners and researchers use a criterion of a respiration rate at rest equal to or more
than a specified number of breaths per minute (b/
min), such as 16. There is no universally accepted
normal breathing rate. A resting breathing rate
of 8–14 b/min, reported by Holloway (1994), is a
commonly accepted criterion. However, there are
reports of 16–17 b/min (nearly 3 b/min greater;
Tobin, Mador, Guenther, Lodato, & Sackner,
1988) for younger and older persons monitored
without their awareness. Other reports suggest
that men and people in their early 20s show
slightly lower rates than women or older people, as
reported in Fried and Grimaldi (1993). Fried and
Grimaldi suggest a goal of “no more than 9 to 12
b/min” (p. 246) for a person at rest. He assumes
normal tidal volume and partial pressure of CO2
(pCO2). People with organic diseases that affect
respiration typically show faster respiration rates,
from about 18–28 b/min (Fried, 1987; Fried &
Grimaldi, 1993).
Therefore, using a breathing rate above 16
is a crude and insufficient criterion for HV, and
one cannot rely on this. Ley (1993), Fried and
Grimaldi (1993), Bonn, Readhead, and Timmons
(1984), Folgering and Colla (1978), Howell (1990),
and Timmons and Ley (1994) provide good discussions of this topic. For example, Ley (1993) states
that even sound operational definitions are often
insufficient unless they include information about
the person’s medical conditions (pulmonary or
heart disease, etc.) at the time.
HV Provocation Test
The “HV provocation test” (HVPT) involves
directed, intentional, and very rapid and usually
deep breathing. Instructions often include something like filling the lungs with each inhalation
and exhaling as completely as possible. The purpose is to reproduce the patient’s symptoms and
complaints as a diagnostic aid. Limited empirical
examination of this test exists and thus it has it
critics (Lindsay, Saqi, & Bass, 1991), especially
related to the reliability and validity of the test,
but this technique remains in common clinical
practice.
Methods reported usually involve a specified
breathing rate of at least 20 b/min, and sometimes
as many as 60 b/min. Patients do this for a speci-
87
fied time of at least 60 seconds and usually for 2–3
minutes. There are guidelines, but no standard
protocol exists (Timmons & Ley, 1994). Howell
(1990) suggests the 20 Deep Breaths Test. Folgering and Colla (1978) use “one minute of deep
breathing”(p. 509). Bonn, Readhead, and Timmons (1984) have added breathing with the upper
chest and 60 b/min.
Some researchers and many practitioners rely
only on the rate-and-time method. This can be
very unpleasant for some patients. Therefore, practitioners sometimes stop when many of the symptoms appear. Some use a pacing device such as a
fast metronome or audiotape (Salkovskis & Clark,
1990). Others also specify a percentage (e.g., 50%)
drop in pCO2 as an alternative criterion (Craske
& Barlow, 1990; Gardner, 1994; Nixon & Freeman, 1988: Salkovskis & Clark, 1990). One can
specify a specific level of alveolar pCO2. Criteria
vary—for example, less than 19 mm Hg (19 torr;
Nixon & Freeman, 1988) to below 38 torr (Fried
& Grimaldi, 1993). Gardner (1994) uses the criterion of “below about . . . 30 mm Hg at rest or
during or after exercise, or remains low 5 minutes
after voluntary over-breathing” (p. 1093). Opinions of normal pCO2 range from 38 to 40 mm Hg.
There are individual differences in the level below
which symptoms appear. This also always depends
on other factors, such as altitude where collected.
Measures of petCO2 are usually sea-level measures.
The measurement drops as altitude increases
because of decreased air pressure at higher altitudes; petCO2 is a percentage of CO2 relative to air
pressure in the surrounding environment. petCO2
is less at higher altitudes and in high-elevation cities such as Denver.
Some consider a patient’s awareness of the
similarity of naturally occurring symptoms to
HVPT-induced symptoms as “the most important
element in the diagnosis of HVS” (Garssen, De
Ruiter, & van Dyck,, 1992, pp. 149–150; see also
Lewis & Howell, 1986). However, studies of this
criterion are rare. In the few studies available, “the
response to the HVPT does not predict the occurrence of HV during panic attacks” (Garssen et al.,
1992, p. 150). This review summarized studies that
concluded there was no difference between the
recognized symptoms typical for HV during the
HVPT and to those occurring during a stressful
time-pressured task without decreased petCO2.
One should read Ley’s (1993) comments about
these studies and the Garssen et al. (1992) review.
Ley is more favorable about the HVPT. However,
he points out that the absence of an HVPT effect
88
does not mean that a patient does not have a HVrelated problem.
Garssen et al. (1992) note that although some
patients recognize symptoms during an HVPT,
CO2 does not always drop during ambulatory monitoring of panic. This conclusion is drawn from the
study by Hibbert and Pilsbury (1989), who used a
transcutaneous estimate of the partial pressure of
CO2 (ptcCO2). Furthermore, this study concluded
that some patients do not recognize the provoked
symptoms. These include patients who show large
drops in ptcCO2 during panic. One conclusion is
that panic does not always result in HV. The transcutaneous method is slow in showing changes that
usually occur for each breath. Thus some practitioners and researchers who have tried it have not
found it useful, especially for office assessments.
One view is that HV is a necessary factor for
most of the somatic symptoms associated with
HVS, but is not sufficient for panic symptoms.
For example, cognitive factors are also necessary, according to Garssen et al. (1992), and thus
cognitive therapy is necessary for treatment. Ley
(1992) agrees with this for a subset of patients.
Others question the role of hyperventilation as a
consistent risk factor (Hornsveld & Garssen, 1996,
1997). HVPT does not always evoke the symptoms
in question for a specific person, which leads to
other concerns. For example, chest pain is not
easy to reproduce. Other views are critical about
the role of HV for panic and the value of HVPT
as the only or best criterion. For example, stressful
mental tasks with only small decreases in alveolar
CO2 can also evoke HVS symptoms (Garssen et
al., 1992). However, the decreases in petCO2 were
1.2 mm Hg. Furthermore, a mental task of thinking about various stressful topics led to HV and
decreased pCO2 among more subjects (33/54, or
61%) than did the HVPT (7/54, or 13%; Nixon &
Freeman, 1988).
Some practitioners (Fried & Grimaldi, 1993)
are far more cautious about HVPT. Fried considers it hazardous and recommends against this
procedure. However, he provides no guidelines or
exceptions about when it may be acceptable. He is
cautious partly because he is a nonphysician. He is
very concerned about inducing changes in blood
acid–base balance, coronary and cerebral vasoconstriction, and ischemic hypoxia. This is worrisome to him, as it should be to everyone, because
a patient could have an undiscovered and undiagnosed organic disease that places him or her
at risk. For example, the patient with diagnosed
II. INSTRUMENTATION
“functional chest symptoms” might turn out to
have an organic cardiac disease for which induced
biochemical and cardiovascular changes increase
the risks of cardiac dysfunction. However, he adds
that “in fairness to my colleagues . . . a number of
them use the procedure and have reported no consequent ill effects in their clients” (p. 42).
Compernolle, Hoogduin, and Joele (1979) caution against using voluntary HV with patients with
chronic anemia or vascular diseases. They refer to
loss of consciousness and fatal accidents when HV
is followed by breath holding during underwater
swimming and diving competition (Hill, 1973;
Craig, 1976). Neurological impairment and fatal
accidents have also occurred following HV in children with sickle-cell anemia (Allen, Imbus, Powars, & Haywood, 1976). However, aside from the
examples cited, Compernolle et al. (1979) noted
that “nothing in the literature to substantiate the
fear that provoking hyperventilation may be dangerous. There are no reports of accidents resulting
from two to five minutes of hyperventilation followed by breathing into a bag” (p. 616).
Prudent practitioners are extra cautious with
people at risk for syncope. Causes for syncope
include certain cardiovascular, metabolic, or neurological disorders, such as seizures. Remember that
hypocapnia-induced vasoconstriction from HV
reduces cerebral blood flow (Berkow & Fletcher,
1992). Thus, caution is necessary for many older
adult patients and those with compromised cerebral blood flow. The clinician should also be aware
that many organic medical conditions can cause
HV (Gardner, 1994). Practitioners must be very
careful not to provoke unintended and potentially
risky symptoms. For example, even psychologically distressed patients who hyperventilate as a
result of habit and anxiety can also have diabetes.
Induced HV may affect or at least interact with
blood glucose (Guthrie, Moeller, & Guthrie, 1983;
C. Lum, 1994; L. C. Lum, 1975). If blood glucose
is very low because of poorly controlled diabetes,
induced HV can add to the acidosis and intensify
symptoms.
Nonmedical practitioners and researchers who
believe they need to use the HVPT or intend to
do so for other clinical or research reasons must
get medical clearance. This is crucial, particularly
in patients with a history of respiratory, cardiovascular, or some neurological diseases. For example,
a history of or evidence for coronary artery disease
or unexplained chest pain clearly indicates caution.
89
5. Cardiorespiratory Measurement and Assessment
A related caution stems from the potential for
some patients to self-initiate breath holding or the
Valsalva maneuver” to abort symptoms, including
those from HVS and panic. Quiet breath holding combined with irregular breathing is a pattern associated with just as many problems as the
obvious HV pattern (Holloway, 1994). There is
one report (Sartory & Olajide, 1988) of the use
of the Valsalva maneuver as a potential treatment
for panic symptoms. Physicians suggest using the
Valsalva to abort paroxysmal atrial tachycardia for
selected patients. Such information may encourage patients to try these without proper consultation. For example, Rapee (1985) reported a case in
which a woman began aborting her panic symptoms by holding her breath without instructions
to do so. In most people, these attempts are not
dangerous, but one must be cautious and complete
in patient education instructions.
In some weight lifters, “hyperventilation before
lifting causes hypocapnia, cerebral vasoconstriction, and peripheral vasodilation” (Berkow &
Fletcher, l992, p. 432). The lifting involves the
Valsalva maneuver, which affects blood return
to the heart, reducing cardiac output and altering CO2 levels. Potentially, systemic vasodilation
and decreased blood pressure may occur, increasing the risk of syncope for some people engaged in
similar activities.
Although, for most people, the risks do not
exist, there are enough reasons to be very cautious
and obtain approval from a qualified physician.
Prudent practitioners, especially nonmedical ones,
may decide to avoid the procedure unless it is absolutely necessary to make a diagnosis or convince a
patient of the diagnosis.
Instrumentation‑Based Breathing Feedback
There are several instrumentation-based breathing feedback systems available to enhance relaxed
breathing. All are relatively simple and in use by
practitioners. There is no research showing any
differential outcomes among them. Practitioners
use those systems that are available and with
which they feel most comfortable.
Nasal Airflow Temperatures
A thermistor (an electrical resistor that measures
temperature) taped below a nostril detects the
changes in temperature of air inhaled (which is
cooler) and exhaled air (warmer). With a sensitively set computer-based visual display, therapist
and patient can clearly see the rapid changes in
temperature. One sees hills and valleys in the
curve on the screen, as the temperature falls during inhalation and rises during exhalation. One
goal is to make the hills and valleys about the
same size and duration. The patient watches the
curve and, with this feedback, regulates the size
and timing of breaths to create a regular rhythm of
hills and valleys in the curve.
The clinician should set the display screen
width to a time reflecting a few inhalations and
exhalations. The temperature should be centered
in the middle of the screen, and the range of temperature (sensitivity) should be set to create hills
and valleys that are easy to see. The temperatures
should not extend beyond the limits of the screen.
The range depends partly on the patient. Therapists who are just starting to use this technique are
advised to try it on themselves first, starting with a
display in which the range is at or less than about
5°F, and tailor it for the individual.
This is a simple technique requiring only one
temperature feedback unit and a simple computerbased software display. However, it does not give
any other information, such as that about muscles
used and CO2. Therefore, practitioners often use it
in conjunction with other feedback instruments.
Finally, in order to avoid transmission of infections
from the nose, therapists must be very careful to
disinfect the thermistor between sessions for all
patients.
Strain Gauges
Many practitioners wrap stretchable devices around
the patient’s abdomen, chest, or both. These allow
monitoring and feedback about abnormal breathing patterns, such as irregularity, breath holding,
and apneas. Therapists also use this to help teach
new breathing patterns, as it is better than observation alone (Timmons & Ley, 1994). The device
has sensors that convey the degree of expansion.
These are connected to a computer-based feedback system that permits viewing the signals on
a computer monitor. The purpose of this type of
system is to allow one to see the expansions in
each body area, as well as the difference between
the abdominal area and the chest area during each
breath. The specific numbers are unimportant and
may be different for each person. The numbers
depend on the tightness of the bands and the sen-
90
sitivity setting. The feedback signal provides hills
and valleys that depend on the size of the breath
and body area used. This is similar to the nasal airflow temperature measure; however, the advantage
over the latter measure is that it gives information
about abdominal versus chest breathing. Like the
nasal airflow temperature measure, this method
provides no information about other muscles or
CO2 and has certain technical limitations, such as
movement artifact. Recalibration after movements
and position changes may be necessary (Timmons
& Ley, 1994). The reader is directed to Fried and
Grimaldi (1993, p. 40) and Timmons and Ley
(1994, pp. 281–282) for more discussion and references for using strain gauges.
Many of the newer biofeedback systems now
have sophisticated breathing pattern guides that
have yet to be studied systematically. These allow
the practitioner to set the pattern to vary inspiration–expiration ratio, respiration rate, and pauses.
EMG from Accessory Breathing Muscles
Practitioners also use EMG feedback from accessory breathing muscles. The sternomastoids, upper
back muscles (including the rhomboids and levator scapulae), and upper chest muscles (including
the pectoralis and/or scalene muscles) are among
the preferred choices. The selection of muscles
depends on therapist preference and practical considerations. The purpose is to show whether there
are EMG increases from these muscles during each
inhalation, to indicate the degree of any increases,
and to give feedback to help the patient reduce
these increases.
The Capnometer and Oximeter Method
The capnometer and oximeter methods allow measurement and feedback of ETCO2 and good estimates of arterial blood oxygen saturation (paO2).
One inserts into either nostril a narrow roughly
0.25-inch plastic tube, which is taped to the skin
near the upper lip (Fried, 1987; Fried & Grimaldi,
1993; R. Fried, personal communication, March
13, 1994). Others have used a standard nasal cannula. The tube’s outer width is about 4 millimeters.
This method allows continuous sampling of endtidal breath conducted to an infrared gas analyzer.
The signal then goes to a computer that feeds back
the wave of rising and falling petCO2 on a video
monitor. It provides a hard copy output of the pattern and gives statistics for specified periods. The
II. INSTRUMENTATION
therapist can place a goal wave for size and rate of
the breaths. One needs a system that recognizes
and displays analyzed blood gas data. These plastic
tubes are readily available and disposable after use.
Spirometry
For therapists without access to a capnometer,
Lum (1991, cited in Timmons & Ley, 1994) recommends “spirometry.” This provides a measure
of the amount of air moved. Nonphysicians and
nonphysiologists should consider the smaller, less
expensive, and relatively accurate Wright spirometer (Timmons & Ley, 1994). This allows an estimate of overbreathing by providing an estimate
of minute volume. Normal resting minute volume
is about 6 liters (Naifeh, 1994). An example of a
criterion for overbreathing is 30 liters per minute
(Lum, 1991, cited in Timmons & Ley, 1994).
Arterial Blood Oxyhemoglobin
Saturation: Oximetry
Fried and Grimaldi (1993) use and recommend a
measure of arterial blood oxyhemoglobin saturation, in which an oximeter attached to the patient’s
index finger. One also could use an ear lobe, a
practice commonly employed by sleep laboratories
for overnight oximetry for assessing decreased oxygen saturations in patients with obstructive sleep
apnea. The output from the oximeter connects to
a physiological monitoring system The oximeter
shows percentage of saturated hemoglobin, and
the biofeedback screen displays the saturation of
arterial blood oxygen over each breath cycle.
This display shows variations of oxygen. It gives
an index of the oxygen delivery to tissues during
monitoring of petCO2. For example, normal pCO2
and elevated oxygen perfusion in the blood (SaO2)
reflect oxygen perfusion expected with deep diaphragmatic breathing. Some experts wonder about
the accuracy of this index. Contrast this to the
reduced oxygen in the tissues associated with
hypocapnia.
Skin Temperatures: Hand and Head Apex
Skin temperature is an indirect measure of breathing, according to Fried and Grimaldi (1993). He
attaches a temperature sensor to the fifth digit
of the nondominant hand while monitoring skin
temperature. Other practitioners may use different digits. By itself, skin temperature is not a good
5. Cardiorespiratory Measurement and Assessment
index; however, it is a common index of relaxation
and is often monitored during breathing therapy
procedures.
Plethysmography: Pulse Rate and Sinus Rhythm
Another indirect measure is obtained by pulse
information (Stern, Ray, & Quigley, 2001). A
biofeedback interface allows for measurement of
variation of beat-to-beat pulse. This index of vagal
tone and respiratory sinus arrhythmia (RSA) during
the breath cycles provides an indirect index of cardiopulmonary status.
Cardiopulmonary techniques such as RSA biofeedback are now becoming popular methods for
breathing therapy.
The Cardiovascular System:
Heart Rate Variability
The cardiac system, like most biological systems, demonstrates constant variation when in a
healthy (or homeostatically balanced) state. It has
long been postulated that heart rates that show a
highly complex but organized pattern of variability (characterized as “chaotic”) are healthier than
very steady rates or simpler patterns of variability.
In recent years we have more understanding of the
nature of variation in the human heart rate (or
more technically in interbeat interval [IBI]). Several papers (Giardino, Lehrer, & Feldman, 2000;
Norris, Gollan, Berntson, & Cacioppo, 2010) present a detailed discussion of the meaning of oscillations in biological and psychological systems.
Several technical reviews of variability in heart
rate also have been published (Berntson et al.,
1997; Nunan, Sandercock, & Brodie, 2010; Task
Force of the European Society of Cardiology and
the North American Society of Pacing and Electrophysiology, 1996). Measures of heart rate variability are derived from cardiac interbeat interval,
measured most accurately from the R-spike in the
electrocardiogram (ECG) but also estimated from
interbeat intervals measured from pulses in the
finger or the ear (Heilman, Handelman, Lewis,
& Porges, 2008). Measures of heart rate variability include “time domain” measures such as standard deviation of normal heartbeats (SDNN), root
mean square of sucessive differences (RMSSD),
and percentage of adjacent heartbeats differing by
more than 50 milliseconds (PNN50). Other “frequency domain” measures are derived from spectral analysis of interbeat intervals.
91
The various rhythms are often reported in hertz
(Hz; cycles per second). The oscillators of interest are between 0.003 and 0.4 Hz. High-frequency
(HF) rhythms are between 0.15 and 0.4 Hz. With
a cardiotachometer, one can see the HF rhythms,
but other slower oscillators are more difficult to
see. To observe them, a spectral analysis is often
used. A low-frequency (LF) rhythm occurs within
the range of .08–.14 Hz, usually at around six per
minute (0.1 Hz, with a period of 10 seconds). On
a cardiotachometer, this would be difficult to
see, but on the spectral it is clearly visible. This
oscillator correlates with measures of a reflex
that plays an important role in the regulation of
blood pressure (BP). Small pressure sensors in the
major arteries (“baroreceptors”) send information
back to the sinus node of the heart to maintain
homeostasis in the BP system. When BP rises, the
baroreceptors stimulate the vagus brake to slow
down the heart so as to reduce pressure. Similarly,
with BP decreases, the baroreceptors send a signal
to the sympathetic cardioaccelerator to speed up
the heart and increase BP. A delay in the baroreflex system of approximately 5 seconds causes the
10-second (0.1 Hz) waves in heart rate. Baroreflex
gain is currently of interest to cardiologists because
it may be an early detector of cardiac disease. It is
quantified as microseconds per millimeter of Hg
(the change in IBI [in milliseconds] that co-occurs
with changes in BP [in mm Hg].
A third oscillator, even more difficult to see in
the cardiotachometer record, is the very-low-frequency (VLF) rhythm, defined as oscillations from
.003 to .08 Hz. It is thought to be driven by a slow
rhythm mediated by the sympathetic nervous system, possibly related to thermoregulation or gastrointestinal regulation. Vaschillo, Lehrer, Rishe,
and Konstantinov (2002) published evidence suggesting that this wave reflects baroreflex effects on
smooth muscle tone in the blood vessels (vascular
tone). Oscillations in vascular tone and BP tend
to show particularly large frequency peaks in the
VLF range, centering at approximately 0.05 Hz
(i.e., three times/minute, or having a period of 20
seconds). This oscillation thus suggests a delay in
the vascular tone limb of the baroreflex system of
about 10 seconds, perhaps due to plasticity of the
blood vessels. This rhythm also is found in heart
rate.
Heart rate variability (HRV) has been used to
measure both autonomic balance and regulatory
capacity of the individual (Berntson, Norman,
Hawkley, & Cacioppo, 2008). Decreased HRV has
92
been associated with increased cardiac mortality
and morbidity (Buccelletti et al., 2009; Thayer,
Yamamoto, & Brosschot, 2010) and mortality in
kidney disease (Ranpuria, Hall, Chan, & Unruh,
2008), diabetes (Manzella & Paolisso, 2005), muscular dystrophy (Politano, Palladino, Nigro, Scutifero, & Cozza, 2008), schizophrenia (Koponen et
al., 2008), and following surgery (Laitio, Jalonen,
Kuusela, & Scheinin, 2007). It is also suppressed
in stress, anger suppression, anxiety, and depressed
affect (Horsten et al., 1999; Marques, Silverman,
& Sternberg, 2010; Shinba et al., 2008), as well
as in reaction to infection (Fairchild, Srinivasan, Moorman, Gaykema, & Goehler, 2011) and
inflammation (Alvarez et al., 2007; Huston &
Tracey, 2011; Jan et al., 2009; Thayer, 2009). Cardiac variability is suppressed in clinical depression,
and compounded suppression of HRV in the frequent comorbidity of cardiac disease and depression is associated with compounded morbidity
and mortality (Carney & Freedland, 2009). Suppressed HRV is, indeed, related to all cause mortality (Lauer, 2009).
HRV also can be used to assess the balance
between the sympathetic and parasympathetic
branches of the autonomic nervous system (ANS;
Berntson et al., 1997; Task Force of the European
Society of Cardiology and the North American
Society of Pacing and Electrophysiology, 1996).
Higher frequencies of HRV (0.15–0.4 Hz) are associated with breathing and are known as “respiratory sinus arrhythmia”; these are under parasympathetic control. Modulation of blood pressure
through changes in heart rate appears to be influenced by both sympathetic and parasympathetic
systems and is reflected in LF heart rate variability (0.05–0.15 Hz; Kuznetsova & Son'kin, 2008;
Palma-Rigo et al., 2010). Vascular tone control,
mediated primarily by alpha sympathetic activity,
is reflected in HRV in the VLF range (0.005–0.05
Hz; Hayoz et al., 1995; Stauss, Rarick, Deklotz, &
Sheriff, 2009). HRV appears to be indicative of
autonomic control of the lung (Yasuma & Hayano, 2004), the gut (Cain, Jarrett, Burr, Hertig, &
Heitkemper, 2007; Chen, Lin, Orr, Yang, & Kuo,
2004; Elsenbruch & Orr, 2001; Huang, Yang, Lai,
& Kuo, 2010; Kaneko, Sakakibara, Mitsuma, &
Morise, 1995; Suzuki et al., 2009), and perhaps
facial muscles (Lindh, Wiklund, Sandman, &
Hakansson, 1997; Stifter, Fox, & Porges, 1989), as
well as the heart.
The sympathetic and parasympathetic systems
interact in a complex synergistic relationship that
is sometimes reciprocal, sometimes additive, and
II. INSTRUMENTATION
sometimes subtractive. Particular rhythms characteristic of each of them operate through these pathways. Although the fight–flight reflex described in
earlier literature on psychological stress is usually
associated with sympathetic activation and parasympathetic suppression, it has long been known
that sympathetic activity increases reactivity in
the parasympathetic system (Gellhorn, 1957).
This parasympathetic activity often rebounds
to high levels after a high level of psychological
stress decreases such as those found in nocturnal
parasympathetic gastric (Nada et al., 2001) and
asthma symptoms (Ballard, 1999; Gelder, Hubble,
& Hetzel, 1988), where parasympathetic arousal
is associated with increased symptomatology. Elevated parasympathetic arousal also occurs during
states of relaxation (Lehrer et al., 1997) and can
be associated with feelings of calmness and wellbeing. Another aspect of the relationship between
the two branches of the ANS of particular interest
to biofeedback practitioners is called accentuated
antagonism. Vagal tone predominates over sympathetic tone at rest. Under normal physiological conditions, abrupt parasympathetic stimulation inhibits tonic sympathetic activation and its
effects at rest during exercise (Olshansky, Sabbah,
Hauptman, & Colucci, 2008).
Frequency Domain Measures
By using an ongoing spectral display that is sensitive to minute-by-minute changes in the ANS,
we can obtain a rich picture of potential pathways
for “mind–body” interaction. Together with other
physiological measures, such as skin conductance,
temperature, or muscle activity, the HRV measures
can help us build a mediational model for disorders
such as noncardiac chest pain, and so forth. For
example, the spectral display provides an online
assessment of sympathetic–parasympathetic balance (the ratio between HF and LF or VLF activity; Berntson et al., 1997) and a rough index of
baroreflex gain (Bernardi et al., 1994).
Porges (2007) has postulated that the vagus
nerve in humans evolved into two pathways: One,
more primitive and older in evolutionary terms,
originates in the dorsal motor nucleus (DMNX),
while the other, more recently evolved in higher
mammals, originates in the nucleus ambiguous
(NA). The DMNX system is best characterized as
a primitive cardiac braking system, the best known
example of which is the “diving reflex.” This system may be excited when the organism experiences cold water in the face or chest and greatly
93
5. Cardiorespiratory Measurement and Assessment
reduces cardiac speed and output. The NA system
is of greater interest within applied psychophysiology because it seems to be involved in heart rate
pacing in nonthreatening social situations. This
is observed as a rhythmic braking and speeding of the heart rate (or IBI) in respiratory sinus
arrhythmia. With inhalation, the vagal braking is
removed; thus, heart rate speeds up. Upon exhalation, the vagal brake is reapplied and slows the rate
down again. Porges and his colleagues further have
reported some evidence that vagus nerve activity, which mediates HF HRV, also reflects social
relatedness, both in humans and across species
(Grippo, Lamb, Carter, & Porges, 2007).
Instrumentation Issues
Measurement of HRV starts with measurement
of heart rate or IBI. The most reliable means of
obtaining an IBI is through the use of an ECG
with sampling rates of at least 512 Hz. The interval between R-waves (the electrical representation
of ventricular contraction) in milliseconds (ms) is
estimated and recorded. This data are then processed as described earlier for use in either measurement or feedback. Often the beat-by-beat
record is displayed with a strain gauge respiration
trace. It is important to remember that the respiration trace is in arbitrary units.
Many commercially available systems derive IBI
from a pulse photoplethysmograph (PPG). While
this method would not be adequate for research,
it seems to work well in clinical biofeedback. One
caution: Make sure the pulse amplitude trace is
clear and distinct, or all the analyses will be corrupted. ECG electrodes can often be used from the
wrists or ankle to wrist, while the PPG sensors are
used on the fingers or ear lobe. Most manufacturers include documentation on electrode placement
and skin preparation.
Glossary
Accentuated antagonism. A term that refers to the fact
that “ . . . Vagal ‘tone’ predominates over sympathetic
tone at rest. Under normal physiological conditions,
abrupt parasympathetic stimulation will inhibit tonic
sympathetic activation and its effects at rest and during exercise. This response is known as ‘accentuated
antagonism’ ”(Olshansky et al., 2008, p. 863).
Acidodis. A condition stemming from a buildup of acid
or depletion of the alkaline reserve (bicarbonate content) in blood and body tissues. There is an increased
concentration of hydrogen ions (i.e., decreased
pH). “Hypercapnic acidosis,” also called “respiratory acidosis,” results from excessive retention of
C02.“Compensated respiratory acidosis” occurs when
the kidneys compensate and raise the low pH toward
normal. There also are other causes. Compare with
alkalosis (see below).
Alkalosis. A condition stemming from a buildup of base
or alkali or from a loss of acid without comparable
loss of base in body fluids. There is a decreased hydrogen ion concentration (i.e., increased pH). “Respiratory alkalosis” results from excess loss of C02 from
the body. “Compensated respiratory alkalosis” occurs
when the blood pH returns toward normal by acid
retention or kidney mechanisms that excrete base
(bicarbonate). There are also other causes. Compare
with acidosis (see above).
Base. In chemistry, the nonacid part of a salt. It pro-
duces hydroxide ions in liquids such as blood.
Bicarbonate. A type of salt (HCO3-). “Blood bicarbon-
ate” is an index of the alkaline reserve level.
Buffer and bicarbonate buffering system. In biochemis-
try, a buffer is any chemical system preventing change
in concentration of another chemical substance such
as hydrogen ion concentration (pH). The kidneys
release bicarbonate as part of the bicarbonate buffering system of the body.
-capnia. A suffix referring to C02. “Hypocapnia” is
low or below-normal C02. “Hypercapnia” is high or
above-normal C02.
Dyspnea. Labored or difficult breathing. “Functional
dyspnea” is dyspnea not related to exercise and without an organic cause. “Sighing intermittent dyspnea”
is very deep sighing respirations without a significant
change in rate, without wheezing. It has functional or
emotional causes rather than organic causes.
Hyperpnea. Breathing large volumes of air in each
breath. Compare with tachypnea (see below).
Hyperventilation (HV). Hyperventilation may be defined
as more tidal volume (or total air flow) than is needed
for metabolic demands. Alveolar carbon dioxide falls
below normal.
Hyperventilation syndrome (HVS). A condition where
prolonged hyperventilation leads to systemic compensation and the loss of the alkaline buffering
system. There is some controversy surrounding this
concept.
Hypocapnia. See “-Capnia.”
Hypoventilation. A condition occurring when there is
too little air entering the pulmonary alveoli.
Hypoxia. A deficiency of oxygen in tissues. This can
occur despite sufficient blood in the tissues. It can
occur if not enough oxygen enters the blood, as with
decreased barometric pressures at high altitudes. It
can also result from decreased oxyhemoglobin in the
94
blood, which is partly a function of the pH of the
blood and is affected by fluctuations of C02 and other
gases. There are other causes.
Isocapnic Overventilation Test (IOT). A technique that
artificially prevents lowered percentage of end-tidal
C02 (PetC02; see below) during hyperventilation by
using a mixture of air enriched with C02.
Paresthesia. An abnormal sensation, such as burning
or prickling.
PetCO2. Percentage end-tidal C02.
Ph. The concentration level and ratio of alkalinity to
acidity. A pH of 7.35 to 7.45 is neutral for blood. A
pH above 7.45 means more alkalinity, and one below
7.35 means more acidity. The symbol pH refers to the
hydrogen ion concentration or activity of a solution,
such as blood.
Photoplethysmograph (PPG). An optically obtained ple-
thysmogram, a volumetric measurement of an organ.
A PPG is often obtained by using a pulse oximeter
which illuminates the skin and measures changes in
light absorption.
Pulmonary alveoli (or vesicles). Tiny sacs at the ends of
the bronchial tree through which gas is exchanged
with the pulmonary capillaries.
Respiratory sinus arrythmia (RSA). Respiratory sinus
arrhythmia (RSA) is a naturally occurring variation
in heart rate that occurs during a breathing cycle.
Heart rate increases during inspiration and decreases
during expiration. Heart rate is normally controlled
by centers in the medulla oblongata. One of these
centers, the nucleus ambiguus, increases parasympathetic nervous system input to the heart via the
vagus nerve.
Tachypnea. Rapid cycles of inhaling and exhaling.
Compare with hyperpnea (see above).
Torr. A unit of pressure equal to 1 millimeter of mercury
(1 mm Hg). It is used in the measurement of ETC02.
Valsalva Maneuver. The valsalva maneuver is a forced
exhalation with the glottis closed. It substantially
increases intra-thoracic pressure and disrupts venous
blood returning to the heart.
Volumes
Minute volume (MV). (1) “Quantity of gas (air)
expelled from the lungs per minute” (Dorland's Illustrated Medical Dictionary, 1988, p. 1847); (2) “volume
of air expelled from the lungs per minute” (Dox, Melloni, & Eisner, 1979, p. 521); (3) “sum of tidal volumes
breathed per minute” (Ley, 1988, p. 253); (4) “volume
of air inspired per minute” (Kaufman & Schneiderman, 1986, p. 112).
Tidal volume (Vy). Amount of gas inspired and
expired (i.e., ventilation) during one respiratory cycle
of a normal breath.
II. INSTRUMENTATION
References
Allen, J. P., Imbus, C. E., Powars, D. R., & Haywood, L. J.
(1976). Neurological impairment induced by hyperventilation in children with sickle cell anemia. Pediatrics,
58(1), 124–126.
Alvarez, S. M., Katsamanis Karavidas, M., Coyle, S. M.,
Lu, S.-E., Macor, M., Oikawa, L. O., et al. (2007). Lowdose steroid alters in vivo endotoxin-induced systemic
inflammation but does not influence autonomic dysfunction. Journal of Endotoxin Research, 13(6), 358–368.
Ballard, R. D. (1999). Sleep, respiratory physiology, and
nocturnal asthma. Chronobiology International, 16(5),
565–580.
Bass, C., & Gardner, W. N. (1985). Respiratory and psychiatric abnormalities in chronic symptomatic hyperventilation. British Medical Journal, 290(6479), 1387–1390.
Berkow, R., & Fletcher, A. (Eds.). (1992). The Merck manual of diagnosis and therapy. Rahway, NJ: Merck.
Bernardi, L., Leuzzi, S., Radaelli, A., Passino, C., Johnston,
J. A., & Sleight, P. (1994). Low-frequency spontaneous
fluctuations of R-R interval and blood pressure in conscious humans: A baroreceptor or central phenomenon?
Clinical Science (London), 87(6), 649–654.
Berntson, G. G., Bigger, J. T., Jr., Eckberg, D. L., Grossman,
P., Kaufmann, P. G., Malik, M., et al. (1997). Heart rate
variability: Origins, methods, and interpretive caveats.
Psychophysiology, 34(6), 623–648.
Berntson, G. G., Norman, G. J., Hawkley, L. C., &
Cacioppo, J. T. (2008). Cardiac autonomic balance versus cardiac regulatory capacity. Psychophysiology, 45(4),
643–652.
Bonn, J. A., Readhead, C. P., & Timmons, B. H. (1984).
Enhanced adaptive behavioural response in agoraphobic patients pretreated with breathing retraining. Lancet, 2(8404), 665–669.
Buccelletti, E., Gilardi, E., Scaini, E., Galiuto, L., Persiani,
R., Biondi, A., et al. (2009). Heart rate variability and
myocardial infarction: Systematic literature review and
metanalysis. European Review for Medical and Pharmacological Sciences, 13(4), 299–307.
Cain, K. C., Jarrett, M. E., Burr, R. L., Hertig, V. L., & Heitkemper, M. M. (2007). Heart rate variability is related to
pain severity and predominant bowel pattern in women
with irritable bowel syndrome. Neurogastroenterology
and Motility, 19(2), 110–118.
Carney, R. M., & Freedland, K. E. (2009). Depression and
heart rate variability in patients with coronary heart
disease. Cleveland Clinic Journal of Medicine, 76(Suppl.
2), S13–S17.
Chen, C. L., Lin, H. H., Orr, W. C., Yang, C. C. H., & Kuo,
T. B. J. (2004). Transfer function analysis of heart rate
variability in response to water intake: Correlation with
gastric myoelectrical activity. Journal of Applied Physiology, 96(6), 2226–2230.
Clark, D. M., & Hemsley, D. R. (1982). The effects of
hyperventilation; individual variability and its relation
to personality. Journal of Behavior Therapy and Experimental Psychiatry, 13(1), 41–47.
5. Cardiorespiratory Measurement and Assessment
Compernolle, T., Hoogduin, K., & Joele, L. (1979). Diagnosis and treatment of the hyperventilation syndrome.
Psychosomatics, 20(9), 612–625.
Craig, A. B., Jr. (1976). Summary of 58 cases of loss of consciousness during underwater swimming and diving.
Medicine and Science in Sports, 8(3), 171–175.
Craske, M. G., & Barlow, D. H. (1990). Nocturnal panic:
response to hyperventilation and carbon dioxide challenges. Journal of Abnormal Psychology, 99(3), 302–307.
Dorland's illustrated medical dictionary (27th ed.). (1988).
Philadelphia: Saunders.
Dox, I., Melloni, B., & Eisner, G. M. (1979). Melloni's illustrated medical dictionary. Baltimore, MD: Williams &
Wilkins.
Elsenbruch, S., & Orr, W. C. (2001). Diarrhea- and constipation-predominant IBS patients differ in postprandial
autonomic and cortisol responses. American Journal of
Gastroenterology, 96(2), 460–466.
Fairchild, K. D., Srinivasan, V., Moorman, J. R., Gaykema,
R. P. A., & Goehler, L. E. (2011). Pathogen-induced
heart rate changes associated with cholinergic nervous
system activation. American Journal of Physiology: Regulatory, Integrative and Comparative Physiology, 300(2),
R330–R339.
Fensterheim, H. (1994). Hyperventilation and psychopathology: A clinical perspective. In B. H. Timmons &
R. Ley (Eds.), Behavioral and psycgological approaches to
breathing disorders. New York: Plenum Press.
Folgering, H., & Colla, P. (1978). Some anomalies in the
control of paCO2 in patients with a hyperventilation
syndrome. Bulliten Europeen de physiopathologie respitatoire, 14, 503–512.
Fried, R. (1987). The hyperventilation syndrome: Research
and clinical treatment. Baltimore, MD: Johns Hopkins
University Press.
Fried, R., Fox, M., & Carlton, R. (1990). Effect of diaphragmatic respiration with end-tidal CO2 biofeedback on
respiration, EEG, and seizure frequency in idiopathic
epilepsy. Annals of the New York Academy of Sciences,
602, 67–96.
Fried, R., & Grimaldi, J. (1993). The psychology and physiology of breathing. New York: Springer.
Gardner, W. N. (1994). Diagnosis and organic causes of
symptomatic hyperventilation. In B. H. Timmons &
R. Ley (Eds.), Behavioral and psychological approaches to
breathing disorders. New York: Plenum Press.
Garssen, B., De Ruiter, C., & van Dyck, R. (1992). Breathing retraining: A rationale placebo. Clinical Psychology
Review, 12, 141–153.
Gelder, C. M., Hubble, D., & Hetzel, M. R. (1988). Parasympathetic nervous system in nocturnal asthma. British
Medical Journal, 297(6645), 424.
Gellhorn, E. (1957). Autonomic imbalance and the hypothalamus: Implications for physiology, medicine, psychology,
and neuropsychiatry. London: Oxford University Press.
Giardino, N. D., Lehrer, P. M., & Feldman, J. M. (2000).
The role of oscillations in self-regulation: Their contribution to homeostasis. In D. T. Kenny, J. G. Carlson, F.
T. Guigan, & J. L. Sheppard (Eds.), Stress and health:
95
Research and clinical applications (pp. 27–51). Amsterdam: Harwood Academic.
Grippo, A. J., Lamb, D. G., Carter, C. S., & Porges, S. W.
(2007). Social isolation disrupts autonomic regulation of
the heart and influences negative affective behaviors.
Biological Psychiatry, 62(10), 1162–1170.
Guthrie, D., Moeller, T., & Guthrie, R. (1983). Biofeedback
and its application to the stabilization and control of
diabetes mielitus American Journal of Clinical Biofeedback, 6, 82–87.
Guyton, A., & Hall, J. (1995). Textbook of medical physiology
(9th ed.). Philadelphia: Saunders.
Hayoz, D., Bernardi, L., Noll, G., Weber, R., Porret, C. A.,
Passino, C., et al. (1995). Flow-diameter phase shift. A
potential indicator of conduit artery function. Hypertension, 26(1), 20–25.
Heilman, K. J., Handelman, M., Lewis, G., & Porges, S. W.
(2008). Accuracy of the StressEraserReg. in the detection of cardiac rhythms. Applied Psychophysiology and
Biofeedback, 33(2), 83–89.
Hibbert, G., & Pilsbury, D. (1989). Hyperventilation: Is it a
cause of panic attacks? British Journal of Psychiatry, 155,
805–809.
Hill, P. M. (1973). Hyperventilation, breath holding and
alveolar oxygen tensions at the breaking point. Respiration Physiology, 19(2), 201–209.
Holloway, E. A. (1994). The role of the physiotherapist in
the treatment of hyperventilation. In B. H. Timmons &
R. Ley (Eds.), Behavioral and psychological approaches to
breathing disorders. New York: Plenum Press.
Hornsveld, H., & Garssen, B. (1996). The low specificity of
the Hyperventilation Provocation Test. Journal of Psychosomatic Research, 41(5), 435–449.
Hornsveld, H., & Garssen, B. (1997). Hyperventilation
syndrome: An elegant but scientifically untenable concept. Netherlands Journal of Medicine, 50(1), 13–20.
Horsten, M., Ericson, M., Perski, A., Wamala, S. P.,
Schenck-Gustaffson, K., & Orth-Gomer, K. (1999).
Psychosocial factors and heart rate variability in health
women. Psychosomatic Medicine, 61(1), 49–57.
Howell, J. B. (1990). Behavioural breathlessness. Thorax,
45(4), 287–292.
Huang, Y.-M., Yang, C. C. H., Lai, C. J., & Kuo, T. B. J.
(2010). Involvement of sympathetic function in the
sleep-related change of gastric myoelectrical activity in
rats. Journal of Sleep Research, 19(1, Pt. 2), 192–200.
Huston, J. M., & Tracey, K. J. (2011). The pulse of inflammation: Heart rate variability, the cholinergic antiinflammatory pathway and implications for therapy.
Journal of Internal Medicine, 269(1), 45–53.
Jan, B. U., Coyle, S. M., Oikawa, L. O., Lu, S.-E., Calvano,
S. E., Lehrer, P. M., et al. (2009). Influence of acute epinephrine infusion on endotoxin-induced parameters of
heart rate variability: A randomized controlled trial.
Annals of Surgery, 249(5), 750–756.
Kaneko, H., Sakakibara, M., Mitsuma, T., & Morise, K.
(1995). Possibility of postprandial electrogastrography
for evaluating vagal/nonvagal cholinergic activity in
humans, through simultaneous analysis of postpran-
96
dial heart rate variability and serum immunoreactive
hormone levels. American Journal of Gastroenterology,
90(4), 603–609.
Kaufman, M. P., & Schneiderman, N. (1986). Physiological bases of respiratory psychophysiology. In M. G. H.
Coles, E. Donchin, & S. W. Porges (Eds.), Psychophysiology: Systems, processes, and applications. New York: Guilford Press.
Koponen, H., Alaraisanen, A., Saari, K., Pelkonen, O., Huikuri, H., Raatikainen, M. J. P., et al. (2008). Schizophrenia and sudden cardiac death: A review. Nordic Journal
of Psychiatry, 62(5), 342–345.
Kuznetsova, O., & Son’kin, V. (2008). Autonomic control
of the respiratory-hemodynamic system in 8- to 11-yearold children with different baroreflex sensitivities.
Human Physiology, 34(5), 633–642.
Laitio, T., Jalonen, J., Kuusela, T., & Scheinin, H. (2007).
The role of heart rate variability in risk stratification
for adverse postoperative cardiac events. Anesthesia and
Analgesia, 105(6), 1548–1560.
Lauer, M. S. (2009). Autonomic function and prognosis.
Cleveland Clinic Journal of Medicine, 76(Suppl. 2), S18–
S22.
Lehrer, P. M., Hochron, S. M., Mayne, T., Isenberg, S.,
Lasoski, A. M., Carlson, V., et al. (1997). Relationship
between changes in EMG and respiratory sinus arrhythmia in a study of relaxation therapy for asthma. Applied
Psychophysiology and Biofeedback, 22(3), 183–191.
Lewis, R. A., & Howell, J. B. L. (1986). Definition of the
hyperventilation syndrome. Bulletin of European Clinical
Respiratory Physiology, 22, 201–205.
Ley, R. (1988). Panic attacks during relaxation-induced
anxiety: A hyperventilation interpretation. Journal of
Behavior Therapy and Experimental Psychiatry, 19(4),
253–259.
Ley, R. (1992). The many faces of Pan: Psychological and
physiological differences among three types of panic
attacks. Behaviour Research and Therapy, 30(4), 347–357.
Ley, R. (1993). Breathing retraining in the treatment of
hyperventilatory complaints and panic disorder: A reply
to Garssen, de Ruiter, and van Dyck. Clinical Psychology
Review, 13(5), 393–408.
Lindh, V., Wiklund, U., Sandman, P. O., & Hakansson, S.
(1997). Assessment of acute pain in preterm infants by
evaluation of facial expression and frequency domain
analysis of heart rate variability. Early Human Development, 48(1-2), 131–142.
Lindsay, S., Saqi, S., & Bass, C. (1991). The test–retest reliability of the hyperventilation provocation test. Journal
of Psychosomatic Research, 35(2–3), 155–162.
Lum, C. (1994). Hyperventilation syndromes: Physiological considerations in clinical managment. In B.
H. Timmons & R. Ley (Eds.), Behavioral and physiological approaches to breathing disorders. New York: Plenum
Press.
Lum, L. C. (1975). Hyperventilation: The tip and the iceberg. Journal of Psychosomatic Research, 19(5–6), 375–
383.
II. INSTRUMENTATION
Manzella, D., & Paolisso, G. (2005). Cardiac autonomic
activity and Type II diabetes mellitus. Clinical Science,
108(2), 93–99.
Marques, A. H., Silverman, M. N., & Sternberg, E. M.
(2010). Evaluation of stress systems by applying noninvasive methodologies: Measurements of neuroimmune
biomarkers in the sweat, heart rate variability and salivary cortisol. Neuroimmunomodulation, 17(3), 205–208.
Nada, T., Nomura, M., Iga, A., Kawaguchi, R., Ochi, Y.,
Saito, K., et al. (2001). Autonomic nervous function in
patients with peptic ulcer studied by spectral analysis
of heart rate variability. Journal of Medicine, 32(5-6),
333–347.
Naifeh, K. H. (1994). Basic anatomy and physiology of the
respiratory system and the autonomic nervous system.
In B. H. Timmons & R. Ley (Ed.), Behavioral and psychological approaches to breathing disorders. New York:
Plenum Press.
Nixon, P. G., & Freeman, L. J. (1988). The ‘think test': A
further technique to elicit hyperventilation. Journal of
the Royal Society of Medicine, 81(5), 277–279.
Norris, C. J., Gollan, J., Berntson, G. G., & Cacioppo, J.
T. (2010). The current status of research on the structure of evaluative space. Biological Psychology, 84(3),
422–436.
Nunan, D., Sandercock, G. R. H., & Brodie, D. A. (2010).
A quantitative systematic review of normal values for
short-term heart rate variability in healthy adults. Pacing
and Clinical Electrophysiology, 33(11), 1407–1417.
Olshansky, B., Sabbah, H. N., Hauptman, P. J., & Colucci,
W. S. (2008). Parasympathetic nervous system and heart
failure: Pathophysiology and potential implications for
therapy. Circulation, 118(8), 863–871.
Palma-Rigo, K., Baudrie, V., Laude, D., Petrel, C., Clauser,
E., & Elghozi, J.-L. (2010). Cardiovascular rhythms and
cardiac baroreflex sensitivity in AT1A receptor gainof-function mutant mice. Chronobiology International,
27(1), 128–137.
Politano, L., Palladino, A., Nigro, G., Scutifero, M., &
Cozza, V. (2008). Usefulness of heart rate variability as
a predictor of sudden cardiac death in muscular dystrophies. Acta Myologica, 27, 114–122.
Porges, S. W. (2007). A phylogenetic journey through the
vague and ambiguous Xth cranial nerve: A commentary
on contemporary heart rate variability research. Biological Psychology, 74(2), 301–307.
Ranpuria, R., Hall, M., Chan, C. T., & Unruh, M. (2008).
Heart rate variability (HRV) in kidney failure: Measurement and consequences of reduced HRV. Nephrology
Dialysis Transplantation, 23(2), 444–449.
Rapee, R. M. (1985). A case of panic disorder treated with
breathing retraining. Journal of Behavior Therapy and
Experimental Psychiatry, 16(1), 63–65.
Salkovskis, P. M., & Clark, D. M. (1990). Affective
responses to hyperventilation: A test of the cognitive
model of panic. Behaviour Research and Therapy, 28(1),
51–61.
Sartory, G., & Olajide, D. (1988). Vagal innervation tech-
5. Cardiorespiratory Measurement and Assessment
niques in the treatment of panic disorder. Behaviour
Research and Therapy, 26(5), 431–434.
Shinba, T., Kariya, N., Matsui, Y., Ozawa, N., Matsuda, Y.,
& Yamamoto, K.-I. (2008). Decrease in heart rate variability response to task is related to anxiety and depressiveness in normal subjects. Psychiatry and Clinical Neurosciences, 62(5), 603–609.
Stauss, H. M., Rarick, K. R., Deklotz, R. J., & Sheriff, D.
D. (2009). Frequency response characteristics of whole
body autoregulation of blood flow in rats. American
Journal of Physiology: Heart and Circulatory Physiology,
296(5), H1607–H1616.
Stern, R. M., Ray, W. J., & Quigley, K. S. (2001). Psychophysiological recording (2nd ed.). Oxford, UK: Oxford
University Press.
Stifter, C. A., Fox, N. A., & Porges, S. W. (1989). Facial
expressivity and vagal tone in 5- and 10-month-old
infants. Infant Behavior and Development, 12(2), 127–137.
Suzuki, H., Watanabe, S., Hamaguchi, T., Mine, H., Terui,
T., Kanazawa, M., et al. (2009). Brain activation associated with changes in heart rate, heart rate variability,
and plasma catecholamines during rectal distention.
Psychosomatic Medicine, 71(6), 619–626.
Task Force of the European Society of Cardiology & the
North American Society of Pacing and Electrophysiol-
97
ogy. (1996). Heart rate variability. Standards of measurement, physiological interpretation, and clinical use.
European Heart Journal, 17(3), 354–381.
Thayer, J. F. (2009). Vagal tone and the inflammatory
reflex. Cleveland Clinic Journal of Medicine, 76(Suppl.
2), S23–S26.
Thayer, J. F., Yamamoto, S. S., & Brosschot, J. F. (2010).
The relationship of autonomic imbalance, heart rate
variability and cardiovascular disease risk factors. International Journal of Cardiology, 141(2), 122–131.
Timmons, B. H., & Ley, R. (Eds.). (1994). Behavioral and
psychological approaches to breathing disorders. New York:
Plenum Press.
van Dixhoorn, J., & Duivenvoorden, H. J. (1985). Efficacy
of Nijmegen Questionnaire in recognition of the hyperventilation syndrome. Journal of Psychosomatic Research,
29(2), 199–206.
Vaschillo, E., Lehrer, P., Rishe, N., & Konstantinov, M.
(2002). Heart rate variability biofeedback as a method
for assessing baroreflex function: A preliminary study of
resonance in the cardiovascular system. Applied Psychophysiology and Biofeedback, 27(1), 1–27.
Yasuma, F., & Hayano, J.-I. (2004). Respiratory sinus
arrhythmia: Why does the heartbeat synchronize with
respiratory rhythm? Chest, 125(2), 683–690.
Chapter 6
Electroencephalographic Measures
and Biofeedback
A Primer
Nicola Neumann, Ute Strehl, Niels Birbaumer,
and Boris Kotchoubey
Understanding electroencephalographic (EEG)
instrumentation requires a basic knowledge about
the recorded parameter. This chapter begins with a
description of the main EEG frequency bands, the
event-related potentials (ERPs), and their behavioral significance. EEG instrumentation proper
is then elucidated, with a particular emphasis on
biofeedback practitioners’ interests and questions.
potentials heighten this threshold and decrease
the probability of a spike.
The EEG is largely produced by the EPSP in the
pyramidal cells of the upper layers of the cerebral
cortex, with some contribution of granular and
glia cell activity and, possibly, of IPSP (for reviews,
see Creutzfeldt, 1974; Lopes da Silva, 1991; Speckmann & Elger, 1999; see Figure 6.1). Spike potentials that are so important for the function of the
nervous system do not have any substantial contribution to the EEG, because rather than summate,
they cancel each other.
The EEG is characterized by three groups of
different (though physiologically interrelated)
phenomena: (1) oscillations, or EEG rhythms; (2)
evoked potentials (EPs) and ERPs; and (3) slow
potential shifts. Almost all these phenomena are
generated by the cortex (exceptions are a few EP
effects, such as the early auditory EP, which are not
of much importance from the biofeedback point of
view). However, subcortical structures (particularly
the thalamus) also affect to a different extent special
characteristics of these cortical phenomena.
Neurophysiological Basis of the EEG and Its
Behavioral Correlates
The EEG results from the summation of electrical
potentials of neural cells. There are three groups
of such potentials: (1) axonal spike potentials generated according to the “all-or-none” principle,
which serve as the main mean of long-distance
information transmission in the nervous system; (2) gradual excitatory postsynaptic potentials
(EPSP), which are produced by the postsynaptic
membrane on dendrites and the soma of a neuron and may, when accumulated, finally lead to
generating a spike; and (3) gradual inhibitory postsynaptic potentials (IPSP), also produced in the
postsynaptic membrane when specific receptors
are stimulated by an inhibitory transmitter such
as gamma-aminobutyric acid. In contrast to EPSP,
which lower the firing threshold of a neuron, thus
increasing the probability of a spike, inhibitory
Rhythmic EEG Components
The EEG rhythms are defined as regularly recurring waveforms of similar shape and duration. The
following EEG rhythms are distinguished by their
frequency, shape, and functional meaning.
98
6. EEG Measures and Biofeedback
99
FIGURE 6.1. Postsynaptic sum potentials (PSPs) as the potential source of the EEG. Afferent fibers (a, b, c) are
shown with connections to an excitatory nerve cell (EC), to a large pyramidal cell (P), and to an inhibitory
nerve cell (IC). The course of the potential is depicted in the area of the apical dendrites (1), the basal dendrites
(2), the axon hillock (3), and the axon (4). Action potentials (APs) become more frequent with the presentation of a stimulus. At the apical dendrites, the summation of excitatory PSP leads to a negative DC shift, which
often does not pass filtering in the routine EEG. From Zschocke (1995). Copyright 1995 by Springer-Verlag
Berlin/Heidelberg. Reprinted by permission.
Alpha activity (8–13 hertz [Hz]) is characterized by a large amplitude and sinusoidal shape.
It is mostly present over posterior cortical areas
and related to a wake resting state. Amplitudes of
alpha waves vary considerably between individuals
and wax and wane over time but are mostly around
50 µV (maximal about 100 µV) in adults. Stimulation (particularly visual) or mental effort lead to
its attenuation or suppression, a phenomenon that
is referred to as alpha blocking. This blocking, as a
sign of cortical activation, should not be confused
with “alpha dropout,” observed at the beginning
drowsiness when alpha waves become more and
more discontinuous, giving way to a low voltage
pattern in sleep Stage I. The alpha rhythm belongs
to the class of synchronized EEG activity that was
originally attributed to thalamocortical circuits
(Andersen & Andersson, 1968). However, more
recent findings (e.g., Lopes da Silva, Van Lierop,
Schrijer, & Storm van Leeuwen, 1973; Lopes da
Silva, Vos, Mooibroeck, & Van Rotterdam, 1980,
later reviewed by Steriade, 1999, 2006), indicated
that the spread of the alpha rhythm is primarily
subserved by a system of intracortical connections,
whereas visual thalamic nuclei have only a moderate effect.
100
Whereas the alpha rhythm prevails over the
occipital and parietal cortex, its analogue over
the motor cortex (best pronounced at C3 and C4
cites, 10-20 System, see below) is referred to as mu
rhythm, also called Rolandic mu rhythm or sensorimotor rhythm (Sterman, 1973, 1996; Wolpaw,
McFarland, Neat, & Forneris, 1991). Its shape is
more arc-like in contrast to the sine-like shape of
alpha waves, and its frequency is sometimes higher
than that in alpha waves (i.e., up to 15 Hz). Its
individual variability is even greater than that
of the alpha rhythm: It is clearly visible in only
3–14% of subjects (Niedermeyer, 1999b, p. 156),
although a frequency analysis reveals it in almost
all individuals (Schoppenhorst, Brauer, Freund, &
Kubicki, 1980). Just as alpha oscillations reflect the
resting state of the posterior sensory areas, the mu
rhythm reflects the rest state of the motor areas.
A phenomenon functionally very close to the mu
rhythm, most apparent during Stages II and III of
sleep, are sleep spindles.
At the beginning of the biofeedback research,
feedback of the alpha rhythm was very popular,
because (1) the phenomenon is well pronounced
and easy to measure and (2) it was believed that
biofeedback-supported alpha increase can lead to
the so-called “alpha state” regarded as a specific
state of consciousness. Its popularity decreased in
the following decades, particularly after Plotkin
(1976, 1977) showed that alpha self-regulation is
usually attained with peripheral rather than central strategies (e.g., using defocusation of the gaze).
Now it is mainly used as a control condition for
other kinds of EEG biofeedback or in combination
with theta as “alpha–theta feedback.” This protocol rewards both alpha and theta activity in order
to reduce stress, anxiety, and substance dependence (for details, see Part VII, this volume).
Self-regulation of the mu rhythm is, in contrast,
actively used both in treatment for several neurological disorders (epilepsy, attention-deficit/hyperactivity disorder [ADHD]) and in brain–computer
interfaces (BCIs) that capitalize the fact that mu
power changes with motor imagery.
Beta activity (13–30 Hz) represents the desynchronized state of the EEG mainly recorded during
active wakefulness, but also in rapid eye movement
(REM) sleep (usually accompanying dreams). It
consists of a mixture of different frequencies of
low amplitude (mostly 5–20 µV). Beta activity is
recorded mainly over frontal and central regions.
Frontal beta is common and may be very fast (~30
Hz). Central beta is mixed with the Rolandic mu
rhythm (discussed earlier) and can be blocked by
II. INSTRUMENTATION
motor activity or tactile stimulation (Niedermeyer,
1999b).
Self-regulation of beta activity is mainly
employed in combination with theta (see below),
in which patients learn to control the ratio
between the two (for details, see Chapter 30, this
volume).
Gamma activity (30–80 Hz) is usually of low
amplitude, thus requiring more specific analytic
methods to separate it. The activity appears to
have specific cognitive and behavioral correlates
(e.g., Basar-Eroglu, Strüber, Schürmann, Stadler,
& Basar, 1996; Pulvermüller, Birbaumer, Lutzenberger, & Mohr, 1997; Pulvermüller, Keil, & Elbert,
1999; Tallon-Baudry & Bertrand, 1999). The relation of gamma activity to conscious awareness has
been widely disputed. While researchers in the
90s attempted to prove a specific link from conscious information processing to gamma activity,
later researchers criticized this point and indicated
a rather nonspecific relationship (e.g., Balasz et
al., 2006; Steriade, 2006; Vanderwolf, 2000). It
is rather gamma synchronization across cortical
areas, than merely an increase of gamma activity
that may possibly be regarded as a necessary (but
not sufficient) precondition of consciousness (e.g.,
Kaiser, Birbaumer, & Lutzenberger, 2001, 2002;
Kaiser, Ripper, Birbaumer, & Lutzenberger, 2003;
Meador, Ray, Echauz, Loring, & Vachtsevanos,
2002; Vanderwolf, 2000).
From a methodological standpoint, evoked and
induced gamma responses to stimulation are distinguished (Galambos, 1992). On the one hand,
like any EEG frequency, gamma oscillations can
respond to a stimulus by increasing their amplitude. This has a shape of gamma oscillatory bursts
whose latency fluctuates from trial to trial and is
referred to as an “induced gamma response.” In
contrast, an “evoked gamma response” is timelocked to external stimuli such as EPs, which we
discuss below.
Cortical oscillations faster than gamma are designated ripples (80–200 Hz) or fast ripples (200–
500 Hz). Because their amplitude is very low, they
currently have no practical importance, though
they probably indicate some aspects of information processing not manifested in other EEG
measures (Eschenko, Mölle, Born, & Sara, 2006;
Mölle, Yeschenko, Marshall, Sara, & Born, 2006).
The delta activity band comprises frequencies
from 0.5 to 4 Hz. Thalomocortical relay neurons
are involved in the generation of these high-voltage, low-frequency waves (Steriade, 2006). Diffuse delta activity is associated with deep sleep in
101
6. EEG Measures and Biofeedback
healthy humans, as well as pathological conditions
such as coma. Local delta waves are recorded over
massive brain lesions.
Theta activity (4–8 Hz) is rarely dominant but
almost always present in the adult EEG. In early
childhood (between 12 and 36 months) it is the
basic rhythm of wakefulness over the posterior
cortex. During the third year it moves into the
alpha range, but intermingled mild-to-moderate
theta activity at posterior scalp locations is still
seen in young adults until the age of 30 (Niedermeyer, 1999a). There are probably two main
types of theta activity (Basar, Schürmann, &
Sakowitz, 2001; Klimesch, 1999; Schacter, 1977)
A more diffuse theta is associated with drowsiness and is clearly discerned during transition
from wakefulness to sleep, while a more frontally
located theta rhythm is associated with stress and
problem solving (Asada, Fukuda, Tsunoda, Yamaguchi, & Tonoike, 1999; Jensen & Tesche, 2002;
Klimesch, 1999; Mizuki, Tanaka, Isozaki, Nishijima, & Inanaga, 1980). Theta activity related to
active problem solving also dominates the activity in the hippocampus in mammals, but there is
controversy over whether this hippocampal activity is related to frontal theta in humans (Steriade,
Gloor, Llinás, Silva, & Mesulam, 1990). Some
experimental results suggest that the master structure controlling at least one type of theta activity is the septohippocampal cholinergic system,
driven from the brainstem reticular core (Steriade,
1999).
seconds prior to movement onset; Pfurtscheller &
Aranibar, 1977). After termination of the movement, the mu rhythm recovers (event-related synchronization [ERS]), indicating the immobilization
or “idling” (Kuhlmann, 1978) of the pyramidal
motor system. Pfurtscheller (1989) suggested that
a variety of rhythms within the alpha and beta
bands are attenuated when cortical structures
become activated either during internal or external events.
Evoked and Event‑Related EEG Changes
Changes in the activity of neuronal populations
time-locked to a specific event, such as a sensory
stimulus, are traditionally studied with ERPs,
which are deflections in the EEG that have a fixed
time delay to the stimulus, while the ongoing EEG
activity (i.e., the oscillations not time-locked to
the event) is regarded as additive noise. To detect
ERPs, averaging techniques are used (see Figure
6.2); that is, many (this number, N, can vary from
dozens to thousands depending on the task) EEG
segments time-locked to a particular stimulus are
summated and the result is divided by N. In this
way, all noise oscillations (i.e., those that are not
phase-locked to the stimulus) are suppressed, while
average
EEG Synchronization and Desynchronization
1
Generally, the normal adult waking EEG can be
classified into two main patterns. The synchronized EEG pattern is characterized by rhythmic,
high-amplitude, low-frequency activity, while the
desynchronized pattern comprises lower voltage
and irregular higher frequency waves. Alpha, mu,
theta, and delta rhythms all characterize different
kinds and degrees of synchronization, whereas the
beta rhythm is typical for desynchronized states.
A synchronized alpha rhythm is measured in a
relaxed, eyes-closed state, while desynchronization
is recorded during visual stimulation and visual
attention. In contrast to the alpha rhythm, the mu
rhythm does not block with eye opening but with
a real, imagined, or intended movement of the
contralateral limb (event-related desynchronization
[ERD]; Pfurtscheller & Berghold, 1989). Interestingly, the desynchronization of the mu rhythm
has an anticipatory nature (i.e., it starts about 1–2
3
4
2
5
6
7
8
9
0
15
30
45
60
75
90
10
FIGURE 6.2. Illustration of the process of averaging.
The lines 1 to 10 present simulated data on 10 different individual trials. As can be seen, the waveforms
vary widely. A very small peak shortly before 30 ms
on the time axis is present in each trial but could
hardly be seen without the arrow pointing at it. The
bold line on the top is the result of averaging the 10
trials. The peak before 30 ms, although almost invisible in individual trials, now becomes the most prominent, sharp, and salient component of the waveform.
102
phase-locked ERP oscillations are highlighted
(Picton & Hillyard, 1988).
An averaged ERP comprises a series of large,
biphasic waves, in total lasting 500–1000 ms.
Although the nomenclature of these waves is far
from being logically clear and unified, the ERP
components are usually described with a letter
P or N, designating polarity (positive or negative, respectively), followed by either the serial
number (components P1, N1, P2, N2, P3), or the
typical peak latency in milliseconds (N270, P300,
N400). The earliest portion of an ERP, usually
called evoked potential (EP), is modality specific
and comprises the waves, which are immediately
related to the stimulus and whose amplitudes and
latencies are primarily determined by the physical
qualities of the stimulus (e.g., intensity and duration). This early portion can last for some 50 ms
(somatosensory EP), 100–110 ms (auditory EP),
or 140–150 ms (visual EP), depending on how fast
the excitation in a particular sensory modality is
transmitted from the receptors to the sensory cortical areas. From a practical point of view, EPs are
mainly used for differential diagnosis of sensory
disorders, such as deafness and blindness, as well
as for the diagnosis of clinical death.
With increasing latency, ERP components are
less and less influenced by simple physical stimulus
qualities, and more and more by the task performed
by a subject. It should be noted that the boundary
between the EP and the “ERP proper” is not exact,
and that the primary cortical responses to stimuli,
such as P1 (with a latency of about 50 and 100 ms,
for auditory and visual modality, respectively) and
N1 (with a latency of about 100 and 150 ms, for
auditory and visual modality, respectively), are
described sometimes as “EP components,” and
other times as “early ERP components.”
The following positive component P2 with a
latency of 180–240 ms is usually the last obligatory
(“exogenous”) response in the chain of oscillations
elicited by any stimulus. Other components, some
of them starting as early as between 50 and 200
ms, as well as all the later deflections, are referred
to as “endogenous” components, because they may
or may not be present, depending on the activity
of the subject. In contrast to EP components indicating information transfer to the cortex, endogenous ERP components manifest information
processing in the cortex and are therefore used to
assess changes in this processing in neurological
and psychiatric patients (e.g., Kotchoubey, 2006;
Kotchoubey, Lang, Bostanov, & Birbaumer, 2002).
II. INSTRUMENTATION
The best known example is the parietal positive
deflection P3 (Donchin & Coles, 1988; Verleger,
1988) related to most complex processing operations, perhaps even a correlate of conscious awareness of stimulus (review Kotchoubey, 2005, but
see Shevrin, 2001). A nice demonstration of the
purely endogenous nature of such components
includes experiments in which rare irregular omissions were introduced in a sequence of otherwise
regularly presented stimuli, and participants were
instructed to attend to the absent stimuli (Ito,
Kitagawa, & Kimura, 1997; Picton & Hillyard,
1988). The components N1 and P2 were elicited
by all stimuli, but only P3 was observed 400–500
ms after omissions, although no physical stimulus
at all was presented!
The model assuming that an ERP can be represented by a signal added to uncorrelated noise
is a simplification. Induced changes in the EEG
rhythmic activity such as ERS and ERD, which
are related but not phase-locked to stimuli, make
the situation more complex. Thus, if EEG rhythms
were additive noise, the quality of the ERP signal
would be proportional to √N (N = number of averaged trials), but in the real situation, it is always
somewhat worse than this. However, the signal-tonoise ratio model can be regarded as an approximation that, though theoretically wrong, is satisfactory from the practical point of view.
Slow Cortical Potentials
Slow cortical potentials (SCPs) are EEG changes
lasting from 300 ms to several seconds. They are
frequently observed as time-locked to stimuli or
subjects’ movements and thus regarded as slow
components of ERPs. The most important are
two slow negative deflections: the bereitschaftspotential (BP; Kornhuber & Deecke, 1965), a negativity starting 500–1500 ms before a voluntary
movement and reflecting movement preparation
in the supplementary motor area; and the contingent negative variation (CNV; Walter, Cooper,
Aldridge, McCallum, & Winter, 1964), recorded
between a warning stimulus (S1) and the following imperative stimulus (S2) in an S1–S2 task in
which participants should respond to S2. Whereas
the BP mainly reflects the activity of the supplementary motor area related to preparation and
timing of a motor action, the CNV manifests
more complex processes of expectancy and general
preparation to an important event (which may but
does not necessarily include a movement).
103
6. EEG Measures and Biofeedback
It must be said, however, that SCPs do not really
belong to ERPs, because the same slow shifts we
record as linked to stimuli and movements also
occur continuously in all functional states from
active wakefulness to deep sleep, regardless of
whether we are stimulated or not. They underlie all other, faster and more regular oscillations
from delta to gamma (listed earlier) and reflect a
mechanism of threshold regulation for local excitatory mobilization. Negative SCP shifts, such as
the BP or the CNV, indicate local preparation of
cortical networks to anticipated activity, whereas
positive potential shifts indicate either the actual
activity or the following disfacilitation. A consistent relationship between cortical negativity and
reaction time, signal detection, and short-term
memory performance has been found in many
studies in humans and monkeys (e.g., Bauer, 1984;
Birbaumer, Elbert, Lutzenberger, Rockstroh, &
Schwarz, 1981; Lutzenberger, Roberts, & Birbaumer, 1993; Rockstroh, Elbert, Lutzenberger, &
Birbaumer, 1982; Siegel, Sterman, & Ross, 1979).
Tasks requiring attention are performed significantly better when presented during spontaneous
or self-induced cortical negativity (summarized
in Birbaumer, 1997; Birbaumer, Elbert, Canavan,
& Rockstroh, 1990; Rockstroh, Elbert, Canavan,
Lutzenberger, & Birbaumer, 1989).
The possibility of SCP biofeedback was simultaneously demonstrated by Siegel et al. (1979) in
animals and by Lutzenberger, Elbert, Rockstroh,
and Birbaumer (1979) and Elbert, Birbaumer, Lutzenberger, and Rockstroh (1979) in humans. Since
then it has been applied for treatment of epilepsy
(see Chapter 37, this volume) and ADHD, reported
in several controlled studies (e.g., Kotchoubey
et al., 2001; Rockstroh et al., 1993; Strehl et al.,
2006).
Instrumentation and Recording
Although procedures for recording EEG activity
have improved greatly over the past decades with
the incorporation of computer-controlled and
digital amplifiers, there is still a need to consider
carefully how EEG is recorded. Possible artifacts
occur at every step of the recording procedure
from the electrodes to the recording system, and
interfering electrical potentials can be easily mistaken for the proper EEG signal. We describe specific problems in the corresponding sections of
the chapter.
Electrodes
Inside the brain tissue, electric charges are transported by ions (i.e., electrically charged atoms).
These charges have to be transmitted to the
recording system. For this transmission, a conductive electrolyte paste is required. The electrodes
are directly in contact with the electrolyte, and
ions move across the boundary. Initially after
fixing the electrodes, electric charges move and
generate an unstable signal. When a balance is
reached, a charged double layer is formed between
the electrolyte and the metal surface of the electrode. It corresponds to a direct current (DC) voltage source with the electric potential difference,
referred to as “electrode potential.”
It is important that the electrode potential
is stable after a few minutes, which is warranted
by clean nonpolarizable electrodes. The lack of
polarization is essential when slow EEG components such as P3 or SCP should be fed back. The
electrode potential should be the same for all electrodes. Therefore, although electrodes may generally be made out of different metals, all electrodes
in a recording system must be of the same material. Disturbances of the potential gradient may
be caused by alterations in temperature, sweating,
or mechanical displacement of the electrodes (see
below).
Electrode Placement
The system of locating electrodes is referred to as
the International 10-20 Recording System (Jasper,
1958), and originally comprised 19 electrodes (see
Figure 6.3). The term “10-20” derives from the
fact that electrodes are placed at sites in a 10 or
20% distance from four anatomical landmarks.
There are two landmarks at the front (nasion, the
bridge of the nose) and the back (inion, the bump
at the back of the head), and two landmarks on
the right and left sides of the head (preauricular
points, depressions in front of the ears above the
cheekbones). Between these landmarks, electrode
positions are determined by measuring distances
10 and 20%, respectively, of the total distance
between the landmarks. The standard numbering
system in the 10-20 system places odd-numbered
electrodes on the left and even-numbered electrodes on the right, with the letter designating
the anatomical area. Such electrode placement
can be replicated consistently over time, as well as
between laboratories. The American EEG Society
104
ii. instruMentAtion
fiGure 6.3. International 10–20 Electrode Placement System. From Jasper (1958). Copyright 1958 by Elsevier
Ireland Ltd. Reprinted by permission.
(1991) added electrode placement nomenclature
guidelines that designate specific locations and
identifications of 75 electrode positions. Which
and how many electrodes are used depends very
much on the research question or clinical considerations. For further information, see Reilly
(1999).
The Issue of Electrode Reference
Many EEG handbooks distinguish between monopolar and bipolar recordings. From the physical point of view, all EEG recordings are bipolar,
because EEG potentials are always recorded as the
voltage difference between two sites. However,
in practical use, we speak about bipolar recording
when we build electrode pairs (e.g., left hemispheric leads are recorded against the symmetrical electrodes on the right side, or anterior against
posterior electrodes), and we speak about monopolar recordings when using a single voltage point
(i.e., the reference) against which all the remaining electrode voltages are measured. Ideally, an
electrically silent position should be chosen for the
reference. In fact, there is no such place. The following solutions have been proposed:
1. A single cephalic reference site remote from
the brain (e.g., the nose or the chin). The problem of the latter is the presence of strong muscles
and the possibility of movement artifacts. The
nose is more appropriate but, unfortunately, some
patients do not like having an electrode on their
nose. Measured EEG voltage is always directly
related to the distance between the two measurement points. Therefore, nose reference creates a
small but substantial bias in that the amplitudes
recorded at posterior sites are larger than those at
anterior sites.
2. A cephalic reference point with a maximal
activity. The idea is the opposite (i.e., if it is impossible to find a point of silence, to place the reference in the epicenter of the quake). Thus, some
authors refer their recording to the vertex, but the
resulting voltages do not reflect the activity under
the corresponding electrodes but instead the difference between this activity and the maximal
activity at vertex.
3. A single noncephalic reference (e.g., the
clavicle) is attractive, because the farther from the
brain, the more neutral the condition. However,
noncephalic reference sites pick up more electromyographic and particularly electrocardiographic
activity than electrodes placed on the head.
4. Average references. The most common are
the linked ear lobe or mastoid reference. To avoid
a strong asymmetry, which would appear if we
use only one-side reference, the two symmetrical
sites were earlier electrically linked. This technique, however, is related to some electrotechnical problems due to unequal resistance. Now these
problems are avoided, because instead of wiring
together the two earlobe (or mastoid) electrodes,
105
6. EEG Measures and Biofeedback
their voltages are online averaged during recording. The disadvantage of the ear reference, and
especially the mastoid reference (the prominence
behind the ear), is their contamination by the
temporobasal activity of the brain. For this reason,
mastoid reference should be avoided whenever one
is interested in the topographical distribution of
the EEG voltages; neither mastoid nor earlobe
references should be used if one supposes that the
activity of temporal lobes might be of importance.
To avoid these disadvantages, a final option is
averaging, not across two symmetrical sites, but
across all recording sites: the so-called “common
average references” (CARs). In the ideal case,
when the entire head is covered by equally spaced
electrodes and the potential on the head is generated by point sources, the CAR should result in a
spatial voltage distribution with a mean of zero. In
the real case, the obtained values are differences
from the momentary mean over the scalp.
Other average references, quite useful in EEG
feedback, employ spatial filtering of the signal.
Most common of these are varieties of Laplacian
methods, in which the voltage at each electrode
is referred to the average of the surrounding electrodes. The recorded signal is therefore the gradient indicating how much the given site differs
from its environment. In a paradigmatic study,
McFarland, McCane, David, and Wolpaw (1997)
fed back the 8–12 Hz mu rhythm and compared a
conventional ear reference, a CAR and two Laplacian methods differing in the radius of the environment used for calculating the reference. The
CAR and the large Laplacian (6 cm circle of surrounding electrodes) showed a significantly better
signal-to-noise ratio than did the ear reference and
the smaller Laplacean method.
Finally, the development of the average reference technique has led to a discovery of methods
for localization of the sources of electrical activity in the brain. The best of these methods, called
LOw REsolution Electromagnetic TomogrAphy
(LORETA), can only be indicated here, since its
detailed description would require a separate chapter (see Pascual-Marqui, 2008; Pascual-Marqui,
Esslen, Kochi, & Lehmann, 2002). To date, there
have already been successful attempts to combine
LORETA with neurofeedback, selectively presenting to the patient the neural activity originated
from a particular area in the brain (Cannon, Congedo, Lubar, & Hutchen, 2009; Congedo, Lubar,
& Joffe, 2004).
Amplification and Filters
In modern amplifiers, only signal differences
between the two inputs of the amplifier are processed. By this means, the technical noise primarily caused by 60 Hz (in Europe: 50 Hz) main
power and arriving at both amplifier inputs is cancelled. The common mode rejection accomplishes
the same purpose, excluding signals swinging
in-phase. Thus, the common mode rejection ratio,
defined as the ratio between amplification of outof-phase signals and (residual) amplification of inphase signals, constitutes a quality characteristic
of amplifiers.
It is important that the amplifier is able to
record all desired EEG frequencies from slow
waves up to 100 Hz. Many commercially available
systems, however, cannot record slow potential
shifts and high-frequency rhythms. The limited
frequency range amplified adequately by a certain
type of amplifier is called its “bandwidth.” Often,
unwanted frequency ranges are additionally suppressed by filters. The high pass filter determines
the lower frequency limit; it lets pass higher frequencies and attenuates lower ones. It is called the
“time constant” and describes the time an alternating current (AC) signal needs to decay to twothirds of its initial amplitude (Figure 6.4). The low
pass filter determines the upper frequency limit:
Lower frequencies can pass and higher ones are
attenuated. Notch filters set at 60 Hz (in Europe:
50 Hz) should be carefully considered, as they are
not sharp and may remove frequencies of activity
below and above the unwanted frequency.
Artifacts
Artifacts are signals that do not originate in the
brain (i.e., are not a part of the EEG) but are
recorded together with the EEG signal and can
imitate some of the EEG phenomena described
earlier (oscillations, ERPs, or SCPs). One can distinguish between technical and biological EEG
artifacts. The former are caused by ambient technical devices, mostly by electromagnetic fields present in our technical environment and produced
by networks and the numerous electrical devices.
They are minimized or completely avoided when
the impedance between EEG electrodes and the
skin is low. Early EEG standards presumed impedance below 5 kOhm. This is a good value to be
attained, although modern EEG amplifier are electrically shielded to such an extent that recordings
free from technical artifacts can even be obtained
106
II. INSTRUMENTATION
FIGURE 6.4. High-pass and low-pass filtering and its influence on a rectangular calibration signal. The high-pass
filter is also referred to as time constant (tc) and determines the time an AC signal needs to decay two-thirds
of its initial amplitude. From Zschocke (1995). Copyright 1995 by Springer-Verlag Berlin/Heidelberg. Reprinted
by permission.
with impedance between 10 and 20 kOhm. An
electrode paste (usually containing sodium chloride as the electrolyte) is applied to the skin under
each electrode after fat and dead skin are removed
with alcohol, acetone, or an abrasive cream. A
similar procedure is used with electrode helmets
in which a blunted needle is used to move the hair
away from the electrode and a syringe-like device
is used to inject electrode paste between the electrode and the scalp. Electrodes and instruments
should be disinfected after each usage.
Although some decades ago all EEG examinations were performed in Faraday boxes, nowadays patients can be examined or trained at their
bedside and also at home, if necessary. However,
sometimes the surrounding noise from electrical
devices may be so strong that artifacts remain even
with very low electrode impedance. In such a case,
unnecessary devices should be switched off. Noise
can increase due to ground loops when different
devices are connected to different grounds. Therefore, one single ground and one common socket
should be used.
In contrast to technical artifacts, biological
artifacts result from the electrical activity of bodily
organs others than the brain, most importantly
(cranial) muscles and eyes. These artifacts can
easily be mistaken for EEG activity. Vertical eye
movements (Figure 6.5) are the most frequent artifacts occurring in EEG recording. The neurons in
the retina generate electrical potentials that constitute an electric dipole, with the inner (caudal)
side of the bulbus being negative. Eye movements
lead to a change of the dipole, thereby influencing potential changes recorded in the EEG. Blink
potentials usually occur in the absence of ocular
rotation. The eyelid, like a sliding electrode, picks
up a positive potential when moving over the positively charged cornea. Eye movements and blinks
can be measured by electrooculography (EOG),
recording potential differences from electrodes
placed around the eyes. Before an EEG pattern
can be interpreted or fed back, ocular influences
have to be removed from it. A common method
to account for ocular potentials picked up by the
EEG electrodes is to subtract a fraction of the
EOG. There are several EOG correction algorithms. Unfortunately, the best and most common algorithms (e.g., Gratton, Coles, & Donchin,
1983) are applied off line and cannot, therefore, be
used to correct EEG signals during feedback. For
neurofeedback purposes, therefore, there are more
107
6. EEG Measures and Biofeedback
FIGURE 6.5. Effects of eye movement on the EEG. Arrows indicate blinks. From Zschocke (1995). Copyright
1995 by Springer-Verlag Berlin/Heidelberg. Reprinted by permission.
approximate algorithms (e.g., Kotchoubey et al.,
1996) available that have demonstrated reliability.
Generally, the effect of biological artifacts on the
EEG is directly related to the distance between the
artifact source and the EEG recording site; therefore, eye movements particularly interfere the EEG
at frontal sites. Having a frequency of about 5 Hz,
they can imitate theta waves or, if stimulus-locked,
the P3 wave of the ERP.
Muscle activity and movements also lead to
artifacts in the EEG (Figure 6.6). When patients
are instructed to relax and not to move, artifacts
may still originate in muscles that are close to the
recording electrodes (e.g., frontalis, orbicularis
oculi, and temporalis muscles). Neck tensions lead
to artifacts at occipital sites. On the one hand,
muscle tension produces high-frequency oscillations, thus interfering with EEG beta and gamma
rhythms. Sometimes, therefore, the increment of
muscle tonus is interpreted as an increase in beta
or gamma power (Whitham et al., 2007). On the
other hand, swallowing, tongue movements, and
breathing cause low-frequency artifacts. To recognize and remove muscular artifacts, it is useful to
compare muscle-rich recording sites (e.g., temporal) with muscle-poor ones (e.g., vertex) (Lutzenberger, Preissl, Birbaumer, & Pulvermüller, 1997).
Another source of biological artifacts is changes
in skin resistance that may be caused by not only
sweating but also psychological factors such as
stress. In the EEG, changes of skin resistance
are reflected by slow potential shifts referred to
as “drifts.” Especially when slow brain activity is
recorded, drifts can be misinterpreted as SCPs.
Digitizing
After proper recording and amplification of the
bioelectric measures, the signals can be fed into
the analog to digital (A/D) converter of a digital
computer. The A/D converter measures the voltage at regular intervals (e.g., every 10 ms). The
frequency of measurements per second is referred
to as “sampling rate.” The original analog signal
can be reconstructed from the digital information
(i.e., numerical values) only if the original signal
does not contain any frequency above half of the
sampling frequency. Otherwise, the occurrence
108
II. INSTRUMENTATION
FIGURE 6.6. Muscle artifacts in the EEG. From Zschocke (1995). Copyright 1995 by Springer-Verlag Berlin/
Heidelberg. Reprinted by permission.
of fast frequencies would simulate oscillations on
the lower frequency range. This effect is called
aliasing. Therefore, an adequate low-pass filter is
mandatory for the input of the signal into a digital
computer (i.e., prior to A/D conversion).
EEG Biofeedback: How to Choose
the Right Program
The effectiveness of the chosen device has to be
evaluated on several dimensions:
1. Validity of the feedback parameter. The field
of application (see Part VI, this volume) should be
specific. The decision to change (e.g., the amount
of certain spectral power frequencies) has to be
substantiated theoretically and empirically, and
the empirical evidence should rely on controlled
studies. For example, if it appears useful to apply
SCP feedback, the appropriate electrodes should
be chosen, and the amplifier must have a long
(at least several seconds) time constant (Bauer,
Korunka, & Leodolter, 1989). Baseline data should
be representative, and the feedback signal should
be easily understood.
2. Reliability of the instrumentation. To reduce
artifacts (discussed earlier) the filtering capacities
have to be considered carefully. The susceptibility
to behaviorally or physiologically induced artifacts
should be checked in test sessions by the therapist.
Electrical interference from the instrumentation, as well as from the environment, may distort
the signal. The feedback signal should be as accurate and stable as possible, because the patient’s
learning depends on this information. This holds
especially for EEG feedback as compared with
feedback of peripheral parameters, since humans
normally cannot perceive their brain waves. To
ensure accuracy, the calibration of the instrument
has to be checked; for stability reasons, the power
source (in the case of batteries or accumulators)
has to be controlled.
3. Flexibility of the software. The software
should allow adjustment to individual characteristics. If the aim of therapy is to obtain changes
in frequency bands, it should be possible to define
the target (e.g., the “alpha spectrum” of a specific
patient).
When progress is lacking, the therapist must be
able to lower thresholds for positive feedback to
initiate shaping (or even prompting) procedures.
A selection of different screen surfaces improves
the motivation in some cases; in other cases it
might be important that the screen surface does
not interfere with the task.
Access to sophisticated data processing and
data storage can be an important feature if more
parameters are of interest. For instance, it might
be useful to analyze the consequences of feedback
of slow cortical potentials on the power spectra.
Glossary
A/D converter. Conversion of analog to digital signals.
Aliasing. A potential error in the conversion of ana-
logue to digital signals; variation in fast EEG activity
may mimic slow potential changes if the original signal contains frequencies above half of the sampling
frequency.
109
6. EEG Measures and Biofeedback
Alpha activity. EEG frequency band from 8 to 13 Hz.
Beta activity. EEG frequency band from 13 to 30 Hz.
Bipolar recording. EEG derivation using pairs of electri-
cally active electrodes.
Brain–computer interface (BCI). A system that directly
connects some brain function (e.g., the amplitude of
the alpha rhythm) to a computer, without an intermediate muscle-driven device such as keyboard or
mouse.
Rolandic mu rhythm. EEG rhythm of the sensorimotor
areas around 10 Hz.
Signal-to-noise ratio. Intensity of the signal of interest
compared to the background noise.
Slow cortical potential (SCP). Direct current shift of less
than 1 Hz.
Thalamus. Diencephalic brain structure; a major relay
center for both sensory and motor signals.
Theta activity. EEG frequency band from 4 to 7 Hz.
Common mode rejection ratio. The ratio between ampli-
fication of out-of-phase signals and (residual) amplification of in-phase signals.
Delta activity. EEG frequency band from 0.5 to 4 Hz.
Electroencephalography (EEG). Recording, amplifica-
tion, and analysis of the electrical activity of the
brain.
Electrooculography (EOG). Recording of electrical sig-
nals evoked by eye movement.
desynchronization
(ERD). Amplitude
attenuation or blocking of rhythmic components
within the alpha and beta bands time-locked (not
necessarily phase-locked) to an internal or external
event.
Event-related
Event-related potential (ERP). A series of deflections in
the EEG time- and phase-locked to an internally or
externally paced event.
Event-related synchronization (ERS). The opposite of
ERD; rhythmic, high amplitude, low-frequency brain
activity time-locked to an internal or external event.
Evoked potential (EP). Comprises the earlier, obligatory
and exogenous ERP components.
Excitatory postsynaptic potential (EPSP). Potential that
arises behind the junctional site between two nerve
cells, or between a nerve cell and an effector cell.
Gamma activity. EEG frequency band from 30 to 100
Hz.
Glia cell. Nonconducting cell that serves as support cell
in the nervous system and helps to protect neurons.
Inhibitory postsynaptic potential (ISPS). Potential that is
produced by an inhibitory transmitter.
International 10-20 recording system. System of locat-
ing electrodes at sites in a 10 or 20% distance from
four anatomical landmarks.
Laplacian method. EEG derivation method; the value at
each electrode location is calculated by subtracting
the value at that location from the values of a set of
surrounding electrodes.
Monopolar recording. EEG derivation with a single ref-
erence site.
Pyramidal cell. Vertically oriented cell in the cerebral
cortex with a pyramid-shaped soma.
References
American Electroencephalographic Society guidelines for
standard electrode position nomenclature. (1991). Journal of Clinical Neurophysiology, 8, 200–202.
Andersen, P., & Andersson, S. A. (1968). Physiological basis
of the alpha rhythm. New York: Meredith Corporation.
Asada, H., Fukuda, Y., Tsunoda, S., Yamaguchi, M., &
Tonoike, M. (1999). Frontal midline theta rhythm
reflects alternative activation of prefrontal cortex and
anterior cingulate cortex in humans. Neuroscience Letters, 274, 29–32.
Balasz, S., Stepan, C., Binder, H., von Gizycki, H., Avitable, M., Obersteiner, A., et al. (2006). Conjugate eye
movements and gamma power modulation of the EEG
in persistent vegetative state. Journal of the Neurological
Sciences, 246(1), 65–69.
Basar, E., Schürmann, M., & Sakowitz, O. (2001). The
selectively distributed theta system: Functions. International Journal of Psychophysiology, 39, 197–212.
Basar-Eroglu, C., Strüber, D., Schürmann, M., Stadler, M.,
& Basar, E. (1996). Gamma-band responses in the brain:
A short review of psychophysiological correlates and
functional significance. International Journal of Psychophysiology, 24, 101–112.
Bauer, H. (1984). Regulation of slow brain potentials affects
task performance. In T. Elbert (Ed.), Self-regulation of
brain and behaviour (pp. 216–226). Berlin: Springer.
Bauer, H., Korunka, C., & Leodolter, M. (1989). Technical requirements for high-quality scalp DC recordings.
Electroencephalography and Clinical Neurophysiology, 72,
545–547.
Birbaumer, N. (1997). Slow cortical potentials: Their origin, meaning, and clinical use. In G. J. M. van Boxtel
& K. B. E. Böcker (Eds.), Brain and behavior past, present
and future (pp. 25–39). Tilburg, The Netherlands: Tilburg University Press.
Birbaumer, N., Elbert, T., Canavan, A. G. M., & Rockstroh,
B. (1990). Slow potentials of the cerebral cortex and
behavior. Physiological Reviews, 70(1), 1–41.
Birbaumer, N., Elbert, T., Lutzenberger, W., Rockstroh, B.,
& Schwarz, J. (1981). EEG and slow cortical potentials
in anticipation of mental tasks with different hemispheric involvement. Biological Psychology, 13, 251–260.
Cannon, R., Congedo, M., Lubar, J., & Hutchen, T.
(2009). Differentiating a network of executive atten-
110
tion: LORETA neurofeedback in anterior cingulate and
dorsolateral prefrontal cortices. International Journal of
Neuroscience, 119(3), 404–441.
Congedo, M., Lubar, J. F., & Joffe, D. (2004). Low-resolution electromagnetic tomography neurofeedback. IEEE
Transactions on Neural Systems and Rehabilitation Engineering, 12(4), 387–397.
Creutzfeldt, O. D. (1974). The neuronal generation of the
EEG. In A. Remond (Ed.), Handbook of electroencephalography and clinical neurophysiology (Vol. 2B, pp. 1–55).
Amsterdam: Elsevier.
Donchin, E., & Coles, M. G. H. (1988). Is the P300 component a manifestation of context updating? Behavioral
and Brain Sciences, 11, 357–374.
Elbert, T., Birbaumer, N., Lutzenberger, W., & Rockstroh,
B. (1979). Biofeedback of slow cortical potentials: Selfregulation of central-autonomic patterns. In N. Birbaumer & H. Kimmel (Eds.), Biofeedback and self-regulation
(pp. 321–337). Hillsdale, NJ: Erlbaum.
Eschenko, O., Mölle, M., Born, J., & Sara, S. J. (2006).
Elevated sleep spindle density after learning or after
retrieval in rats. Journal of Neuroscience, 26(50), 12914–
12920.
Galambos, R. (1992). A comparison of certain gamma
band (40-Hz) brain rhythms in cat and man. In E. Basar
& T. H. Bullock (Eds.), Induced rhythms in the brain
(pp. 201–216). Boston: Birkhäuser.
Gratton, G., Coles, M. G. H., & Donchin, E. (1983). A
new method for off-line removal of ocular artifacts.
Electroencephalography and Clinical Neurophysiology, 55,
468–484.
Ito, J., Kitagawa, J., & Kimura, I. (1997). Event-related
potentials (ERPs) associated with omitted somatosensory stimuli. Electromyography and Clinical Neurophysiology, 37, 169–176.
Jasper, H. H. (1958). The ten–twenty electrode system of
the International Federation. Electroencephalography
and Clinical Neurophysiology, 10(2), 371–375.
Jensen, O., & Tesche, C. D. (2002). Frontal theta activity in humans increases with memory load in a working memory task. European Journal of Neuroscience, 15,
1395–1399.
Kaiser, J., Birbaumer, N., & Lutzenberger, W. (2001).
Event-related beta desynchronization indicates timing
of response selection in a delayed-response paradigm in
humans. Neuroscience Letters, 312, 149–152.
Kaiser, J., Birbaumer, N., & Lutzenberger, W. (2002). Magnetic oscillatory responses to lateralization changes of
natural and artificial sounds in humans. European Journal of Neuroscience, 15, 345–354.
Kaiser, J., Ripper, B., Birbaumer, N., & Lutzenberger, W.
(2003). Dynamics of gamma-band activity in human
magnetoencephalogram during auditory pattern working memory. NeuroImage, 20, 816–827.
Klimesch, W. (1999). EEG alpha and theta oscillations
reflect cognitive and memory performance: A review
and analysis. Brain Research Reviews, 29, 169–195.
Kornhuber, H., & Deecke, L. (1965). Hirnpotentialänderungen bei Willkürbewegungen und passiven Bewegun-
II. INSTRUMENTATION
gen des Menschen: Bereitschaftspotential und reafferente Potentiale. Pflügers Archiv für gesamte Physiologie,
284, 1–17.
Kotchoubey, B. (2005). Event-related potential measures
of consciousness: Two equations with three unknowns.
Progress in Brain Research, 150, 427–444.
Kotchoubey, B. (2006). Event-related potentials, cognition,
and behavior: A biological approach. Neuroscience and
Biobehavioral Reviews, 30, 42–65.
Kotchoubey, B., Lang, S., Bostanov, V., & Birbaumer, N.
(2002). Is there a mind?: Psychophysiology of unconscious patients. News in Physiological Sciences, 17, 38–42.
Kotchoubey, B., Schneider, D., Schleichert, H., Strehl, U.,
Uhlmann, C., Blankenhorn, V., et al. (1996). Self-regulation of slow cortical potentials in epilepsy: A retrial
with analysis of influencing factors. Epilepsy Research,
25, 269–276.
Kotchoubey, B., Strehl, U., Uhlmann, C., Holzapfel, S.,
König, M., Fröscher, W., et al. (2001). Modification
of slow cortical potentials in patients with refractory
epilepsy: A controlled outcome study. Epilepsia, 42(3),
406–416.
Kuhlmann, W. (1978). Functional topography of the
human mu rhythm. Electroencephalography and Clinical
Neurophysiology, 44, 83–93.
Lopes da Silva, F. (1991). Neural mechanisms underlying brain waves: From neural membranes to networks.
Electroencephalography and Clinical Neurophysiology, 79,
81–93.
Lopes da Silva, F., Van Lierop, T. H. M. T., Schrijer, C. F.
M., & Storm van Leeuwen, W. (1973). Organization of
thalamic and cortical alpha rhythm: Spectra and coherences. Electroencephalography and Clinical Neurophysiology, 35, 627–639.
Lopes da Silva, F., Vos, J. E., Mooibroeck, J., & Van Rotterdam, A. (1980). Relative contribution of intracortical and thalamo-cortical processes in the generation of
alpha rhythms, revealed by partial coherence analysis.
Electroencephalography and Clinical Neurophysiology, 50,
449–456.
Lutzenberger, W., Elbert, T., Rockstroh, B., & Birbaumer,
N. (1979). The effects of self-regulation of slow cortical
potentials on performance in a signal detection task.
International Journal of Neuroscience, 9, 175–183.
Lutzenberger, W., Preissl, H., Birbaumer, N., & Pulvermüller, F. (1997). High-frequency cortical responses: Do
they not exist if they are small? Electroencephalography
and Clinical Electrophysiology, 102, 64–66.
Lutzenberger, W., Roberts, L. E., & Birbaumer, N. (1993).
Memory performance and area-specific self-regulation of
slow cortical potentials: Dual-task interference. International Journal of Psychophysiology, 15, 217–226.
McFarland, D. J., McCane, L. M., David, S. V., & Wolpaw,
J. R. (1997). Spatial filter selection for EEG-based communication. Electroencephalography and Clinical Neurophysiology, 103, 386–394.
Meador, K. J., Ray, P. G., Echauz, J. R., Loring, D. W., &
Vachtsevanos, G. J. (2002). Gamma coherence and conscious perception. Neurology, 59, 847–854.
6. EEG Measures and Biofeedback
Mizuki, Y., Tanaka, O., Isozaki, H., Nishijima, H., &
Inanaga, K. (1980). Periodic appearance of theta rhythm
in the frontal midline during performance of a mental
task. Electroencephalography and Clinical Neurophysiology, 49, 345–351.
Molle, M., Yeschenko, O., Marshall, L., Sara, S. J., & Born,
J. (2006). Hippocampal sharp wave-ripples linked to
slow oscillations in rat slow-wave sleep. Journal of Neurophysiology, 96, 62–70.
Niedermeyer, E. (1999a). Maturation of the EEG: Development of waking and sleep patterns. In E. Niedermeyer
& F. Lopes da Silva (Eds.), Electroencephalography: Basic
principles, clinical applications, and related fields (4th ed.,
pp. 189–214). Baltimore, MD: Williams & Wilkins.
Niedermeyer, E. (1999b). The normal EEG of the waking
adult. In E. Niedermeyer & F. Lopes da Silva (Eds.),
Electroencephalography: Basic principles, clinical applications, and related fields (4th ed., pp. 149–173). Baltimore,
MD: Williams & Wilkins.
Pascual-Marqui, R. D. (2008). LORETA: Low resolution
brain electromagnetic tomography. Retrieved from
www.unizh.ch/keyinst/newloreta/loreta01.htm.
Pascual-Marqui, R. D., Esslen, M., Kochi, K., & Lehmann,
D. (2002). Functional imaging with low-resolution brain
electromagnetic tomography (LORETA): A review.
Methods and Findings in Experimental and Clinical Pharmacology, 24(Suppl. C), 91–95.
Pfurtscheller, G. (1989). Functional topography during sensorimotor activation studied with event-related desynchronization mapping. Journal of Clinical Neurophysiology, 6, 75–84.
Pfurtscheller, G., & Aranibar, A. (1977). Event-related
cortical desynchronization detected by power measurements of scalp EEG. Electroencephalography and Clinical
Neurophysiology, 42, 817–826.
Pfurtscheller, G., & Berghold, A. (1989). Patterns of cortical activation during planning of voluntary movement.
Electroencephalography and Clinical Neurophysiology, 72,
250–258.
Picton, T. W., & Hillyard, S. A. (1988). Endogenous eventrelated potentials. In T. W. Picton (Ed.), Human eventrelated potentials (Vol. 3, pp. 361–426). Amsterdam:
Elsevier.
Plotkin, W. B. (1976). On the self-regulation of the occipital alpha-rhythm: Control strategies, states of consciousness, and the role of physiological feedback. Journal of
Experimental Psychology: General, 195, 66–99.
Plotkin, W. B. (1977). On the social psychology of experimental states associated with EEG alpha biofeedback
training. In J. Beatty & H. Legewie (Eds.), Biofeedback
and behavior (pp. 121–133). New York/London: Plenum
Press.
Pulvermüller, F., Birbaumer, N., Lutzenberger, W., & Mohr,
B. (1997). High-frequency brain activity: Its possible role
in attention, perception and language processing. Progress in Neurobiology, 52, 427–445.
Pulvermüller, F., Keil, A., & Elbert, T. (1999). High-frequency brain activity: Perception or active memory.
Trends in Cognitive Sciences, 3, 250–252.
111
Reilly, E. L. (1999). EEG recording and operation of the
apparatus. In E. Niedermeyer & F. Lopes da Silva (Eds.),
Electroencephalography: Basic principles, clinical applications, and related fields (pp. 122–148). Baltimore, MD:
Williams & Wilkins.
Rockstroh, B., Elbert, T., Birbaumer, N., Wolf, P., Düchting-Röth, A., Reker, M., et al. (1993). Cortical selfregulation in patients with epilepsies. Epilepsy Research,
14, 63–72.
Rockstroh, B., Elbert, T., Canavan, A., Lutzenberger, W., &
Birbaumer, N. (1989). Slow cortical potentials and behavior. Baltimore, MD: Urban & Schwarzenberg.
Rockstroh, B., Elbert, T., Lutzenberger, W., & Birbaumer,
N. (1982). The effects of slow cortical potentials on
response speed. Psychophysiology, 19(2), 211–217.
Schacter, D. L. (1977). EEG theta waves and psychological
phenomena: A review and analysis. Biological Psychology, 5, 47–82.
Schoppenhorst, M., Brauer, F., Freund, G., & Kubicki, S.
(1980). The significance of coherence estimates in
determining cerebral alpha and mu activities. Electroencephalography and Clinical Neurophysiology, 48, 25–33.
Shevrin, H. (2001). Event-related markers of unconscious
processes. International Journal of Psychophysiology, 42,
209–218.
Siegel, J. M., Sterman, M. B., & Ross, S. (1979). Automatic
detection and operant reinforcement of slow potential
shifts. Physiology and Behavior, 23(2), 411–413.
Speckmann, E.-J., & Elger, C. E. (1999). Introduction to
the neurophysiological basis of the EEG and DC potentials. In E. Niedermeyer & F. Lopes da Silva (Eds.), Electroencephalography: Basic principles, clinical applications,
and related fields (4th ed., pp. 15–27). Baltimore, MD:
Williams & Wilkins.
Steriade, M. (1999). Cellular substrates of brain rhythms.
In E. Niedermeyer & F. Lopes da Silva (Eds.), Electroencephalography: Basic principles, clinical applications, and
related fields (4th ed., pp. 28–75). Baltimore, MD: Williams & Wilkins.
Steriade, M. (2006). Grouping of brain rythms in corticothalamic systems. Neuroscience, 137, 1087–1106.
Steriade, M., & Buzsáki, G. (1990). Parallel activation of
thalamic and cortical neurons by brainstem and basal
forebrain cholinergic systems. In M. Steriade & D. Biesold (Eds.), Brain cholinergic systems (pp. 3–64). Oxford,
UK: Oxford Medical Press.
Steriade, M., Gloor, P., Llinás, R. R., Lopes da Silva, F., &
Mesulam, M.-M. (1990). Basic mechanisms of cerebral
rhythmic activities. Electroencephalography and Clinical
Neurophysiology, 76, 481–508.
Sterman, M. B. (1973). Neurophysiologic and clinical studies of sensorimotor EEG biofeedback training: Some
effects on epilepsy. Seminars in Psychiatry, 5(4), 507–525.
Sterman, M. B. (1996). Physiological origins and functional
correlates of EEG rhythmic activities: Implication for
self-regulation. Biofeedback and Self-Regulation, 21(1),
3–33.
Strehl, U., Leins, U., Goth, G., Klinger, Ch., Hinterberger,
T, & Birbaumer, N. (2006). Self-regulation of slow cor-
112
tical potentials—a new treatment for children with
ADHD. Pediatrics, 118, e1530–1540.
Tallon-Baudry, C., & Bertrand, O. (1999). Oscillatory
gamma activity in humans and its role in object representation. Trends in Cognitive Sciences, 3, 151–162.
Vanderwolf, C. H. (2000). Are neocortical gamma waves
related to consciousness? Brain Research, 855, 217–224.
Verleger, R. (1988). Event-related potentials and cognition:
A critique of the context updating hypothesis and an
alternative interpretation of P3. Behavioral and Brain
Sciences, 11, 343–356.
Walter, W. G., Cooper, R., Aldridge, V. J., McCallum, W.
C., & Winter, A. L. (1964). Contingent negative varia-
II. INSTRUMENTATION
tion: An electric sign of sensorimotor association and
expectancy in the human brain. Nature, 203, 380–384.
Whitham, E. M., Pope, K. J., Fitzgibbon, S. P., Lewis, T.,
Clark, C. R., Loveless, S., et al. (2007). Scalp electrical recording during paralysis: Quantitative evidence
that EEG frequencies above 20 Hz are contaminated by
EMG. Clinical Neurophysiology, 118(8), 1877–1888.
Wolpaw, J. R., McFarland, D. J., Neat, G. W., & Forneris,
C. A. (1991). An EEG-based brain-computer interface
for cursor control. Electroencephalography and Clinical
Neurophysiology, 78(3), 252–259.
Zschocke, S. (1995). Klinische Elektroencephalographie. Berlin: Springer-Verlag.
Chapter 7
Quantitative Encephalography
and Electroencephalographic
Biofeedback/Neurofeedback
Robert W. Thatcher
Quantitative electroencephalography (QEEG)
is distinguished from visual examination of electroencephalographic (EEG)1 traces, referred to as
“nonquantitative EEG,” by the fact that the latter
is subjective and involves low sensitivity and low
interrater reliability (Cooper, Osselton, & Shaw,
1974; Woody, 1966, 1968), whereas the former
involves the use of computers and power spectral
analyses, and is more objective, with higher reliability and higher sensitivity (Hughes & John,
1999). The improved sensitivity and reliability
of QEEG was first recognized by Hans Berger
in 1934, when he performed a QEEG analysis
involving the power spectrum of the EEG with
a mechanical analog computer (Niedermeyer &
Lopes da Silva, 1995). QEEG at the present time
clearly surpasses conventional visual examination
of EEG traces, because of its high temporal and
spatial resolution in the millisecond time domain
and approximately 1 centimeter in the spatial
domain, which gives QEEG the ability to measure
network dynamics that are simply “invisible” to the
naked eye. Over the last 40+ years, the accuracy,
sensitivity, and resolution of QEEG have steadily
increased because of the efforts of hundreds of
dedicated scientists and clinicians who have produced approximately 120,000 QEEG studies cited
in the National Library of Medicine database. It is
recommended that the reader search the National
Library of Medicine database (www.ncbi.nlm.nih.
gov/sites/entrez?db=pubmed) using the key word
“EEG” and the few representative citations in this
chapter.2 Because of space limitations, no reviews
of this vast literature are attempted here; instead,
my purpose in this chapter is to briefly describe
some of the most recent advances in QEEG as they
relate to EEG biofeedback/neurofeedback.3
Neurological evaluation of space-occupying
lesions has been correlated with the locations and
frequency changes that have been observed in the
EEG traces and in QEEG analyses (e.g., lesions
of the visual cortex resulted in distortions of the
EEG generated from the occipital scalp locations,
or lesions of the frontal lobe resulted in distortions
of the EEG traces arising in frontal regions). However, early neurological and neuropsychological
studies indicated that function is not located in
any one part of the brain (Luria, 1973). Instead,
the brain is made up of complex and interconnected groupings of neurons that constitute
“functional systems,” such as the “digestive system” or the “respiratory system,” in which cooperative sequencing and interactions give rise to an
overall function at each moment of time (Luria,
1973). This widely accepted view of brain function as a complicated functional system, which
became dominant in the 1950s and 1960s, is still
the accepted view today. For example, since the
1980s, new technologies such as functional magnetic resonance imaging (fMRI), positron emis113
114
sion tomography (PET), single-photon emission
computed tomography (SPECT) and QEEG and
magnetoencephalography (MEG) have provided
ample evidence for distributed functional systems
involved in perception, memory, drives, emotions,
voluntary and involuntary movements, executive
functions, and various psychiatric and psychological dysfunctions. Modern PET, QEEG, MEG, and
fMRI studies are consistent with the historical
view of “functional systems” presented by Luria in
the 1950s (Luria 1973); that is, there is no absolute functional localization, because a functional
system of dynamically coupled subregions of the
brain is operating. For example, several fMRI and
MRI studies (e.g., diffusion tensor imaging [DTI])
have shown that the brain is organized by a relatively small subset of “modules” or “hubs” that represent clusters of neurons with high within-cluster
connectivity and sparse long-distance connectivity (Hagmann et al., 2008; Chen, He, Rosa-Neto,
Germann, & Evans, 2008; He et al., 2009). Modular organization is a common property of complex systems and “small-world” models in which
maximum efficiency is achieved when local clusters of neurons rely on a small set of long-distance
connections in order to minimize the “expense”
of wiring and shorten time delays between modules (Buzsaki, 2006; He et al., 2009). Also, recent
QEEG and MEG analyses have demonstrated
that important visually invisible processes such
as coherence, phase delays, phase locking and
phase shifting of different frequencies is critical
in cognitive functions and various clinical disorders (Buszaki, 2006; Sauseng & Klimesch, 2008;
Thatcher, North, & Biver, 2009a). Phase shift and
phase synchrony has been shown to be one of the
fundamental processes involved in the coordination of neural activity located in spatially distributed “modules” at each moment of time (Freeman
& Rogers, 2002; Freeman, Burke, & Homes, 2003;
Thatcher et al., 2009a, 2009b).
QEEG for Assessment and Neurofeedback
for Treatment: A Parent–Child Relationship
This use of the EEG changed dramatically in the
1960s, when computers were used to modify the
EEG through biofeedback, referred to today as
neurofeedback (NF). Studies by Fox and Rudell
(1968); Kamiya (1971), and Sterman (1973) were
a dramatic departure from the classical use of
conventional visual EEG and QEEG in that for
the first time clinicians could consider treating a
II. INSTRUMENTATION
disorder such as epilepsy or attention deficit disorders and other mental disorders by using operant
conditioning methods to modify the EEG itself.
Thus, QEEG and EEG biofeedback have a “parent–child” relationship in that EEG biofeedback
necessarily uses computers and is therefore a form
of QEEG that is focused on treatment based on
the science and knowledge of the physiological
meaning and genesis of the EEG itself. Ideally,
as knowledge about brain function and the accuracy and resolution of the EEG increases, then
EEG biofeedback should change in lockstep with
it to better link symptoms and complaints to the
brain and in this manner treat the patient based
on solid science. To the extent the EEG can be
linked to functional systems in the brain and to
specific mental disorders, EEG biofeedback could
“move” the brain toward a healthier state (i.e.,
“normalize” the brain; Thatcher, 1989, 1999).
Clearly, the clinical efficacy of EEG biofeedback
is reliant on knowledge about the genesis of the
EEG and specific functions of the human brain.
The parent–child relationship and interdependencies between QEEG and EEG biofeedback is active
today and represents a bond that, when broken,
results in reduced clinical efficacy and general
criticism of the field of EEG biofeedback. The traditional and logical relationship between QEEG
and NF is to use QEEG to assess, and to use NF
to treat, based on a linkage between the patient’s
symptoms and complaints and functional systems
in the brain. This parent–child linkage requires
clinical competence on the one hand and technical competence with computers and the EEG on
the other. Competence in both is essential, and
organizations such as the International Society
for Neurofeedback Research (ISNR), the Society
of Applied Neuroscience (SAN), the American
Board of Electroencephalography and Neurophysiology (ABEN), the EEG and Clinical Neuroscience Society (ECNS), the Biofeedback Certification Institute of America (BCIA), the Association
for Applied Psychophysiology and Biofeedback
(AAPB), and others are available to help educate
and test the requisite qualifications and competence to use EEG biofeedback. The parent–child
link is typically optimized by following three steps:
1. Perform a careful and thorough clinical inter-
view and assessment of the patient’s symptoms
and complaints (neuropsychological assessments are the most desirable).
2. Conduct a QEEG in order to link the patient’s
symptoms and complaints to functional sys-
7. QEEG and EEG Biofeedback/Neurofeedback
tems in the brain as evidenced in fMRI, PET
and QEEG/MEG.
3. Devise an EEG biofeedback protocol to address
the deregulations observed in the QEEG
assessment that best match the patient’s symptoms and complaints.
This approach reinforces the close bond
between parent (QEEG) and child (NF), and
allows objective evaluation of the efficacy of treatment in terms of both behavior and brain function.
Figure 7.1 illustrates a common, modern quantitative EEG analysis in which conventional EEG
traces are viewed and examined at the same time
that quantitative analyses are displayed so as to
facilitate and extend the analytical power of the
EEG. Seamless integration of QEEG and NF
involves two basic steps: (1) visual examination of
the EEG traces and (2) spectral analyses of the EEG
traces.4 Numerous studies have shown a relationship between the time and frequency domains of
an EEG time series and LORETA (Low Resolution
Electromagnetic Tomography) three-dimensional
source analyses, which provide 7 mm3 maximal
spatial resolution in real-time (Pascual-Marqui,
Michel, & Lehmann, 1994; Gomez & Thatcher,
2001; see Note 6). There is a verifiable correspon-
115
dence between the time series of the EEG and the
power spectrum and LORETA three-dimensional
source localization, for example, visual cortex
source localization of hemiretinal visual stimulation, temporal lobe source localization of auditory
simulation, theta source localization in the hippocampus in memory tasks, localization of theta in
the anterior cingulate gyrus in attention tasks, and
linkage between depression and rostral and dorsal
cingulate gyrus (see Note 4). The number of clinical QEEG studies cited in the National Library of
Medicine attests to the link between patient symptoms and functional systems in the brain, and protocols for treatment are commonly guided by this
scientific literature.
The Use of 19‑Channel Surface QEEG Z Scores
and EEG Biofeedback
As described by Thatcher and Lubar (2008), scientists at the University of California, Los Angeles
(UCLA) in the 1950s (Adey, Walter, & Hendrix,
1961) and later Matousek and Petersen (1973) were
the first to compute means and standard deviations in different age groups, and then z-scores to
compare an individual to a reference normative
database of means and standard deviations. The
FIGURE 7.1. Example of conventional digital EEG (left) and QEEG (right) on the same screen at the same
time. The conventional EEG includes examination and marking of EEG traces and events. The QEEG (right)
includes the fast Fourier transform (top right) and normative database Z scores (bottom right).
116
“Z statistic” is defined as the difference between
the value from an individual and the mean of
the normal reference population divided by the
standard deviation of the population. John and
colleagues (John, 1977; John et al., 1977; John,
Prichep, & Easton, 1987) expanded on the use of
the Z score and referenced normal databases for
clinical evaluation, including the use of multivariate measures such as the Mahalanobis distance
metric (John et al., 1987; John, Prichep, Fridman,
& Easton, 1988). For purposes of assessing deviation from normal, the values of z above and below
the mean, which include 95–99% of the area of
the Z-score distribution, is often used as a level
of confidence necessary to minimize type I and
type II errors. The standard-score equation is also
used to cross-validate a normative database, which
again emphasizes the importance of approximation to a Gaussian for any normative QEEG database (Thatcher, Walker, Bier, North, & Curtin,
2003; Thatcher, North, & Biver, 2005).
The standard concepts underlying the Z-score
statistic and QEEG evaluations were recently combined to give rise to real-time EEG Z-score biofeedback, also referred to as “live Z-score biofeedback”
II. INSTRUMENTATION
(Thatcher 1989, 1999, 2000a, 2000b; Thatcher &
Collura, 2006; Collura, Thatcher, Smith, Lambos,
& Stark, 2008). The use of real-time Z-score EEG
biofeedback is another method to advance the
integration of QEEG and NF. Figure 7.2 illustrates
the differences between raw score EEG biofeedback and real-time Z-score EEG biofeedback.
There are several advantages of real-time Z-score
biofeedback: (1) simplification by reducing different metrics (power, coherence, phase, asymmetry,
etc.) to a single common metric of the Z score; (2)
simplification by providing a threshold and direction of change (i.e., Z = 0) to move the EEG toward
a normal healthy reference population of subjects5;
and (3) improved linkage between patient’s complaints and symptoms and localization of functional systems in the brain. Figures 7.3, 7.4, and 7.5
show examples of how a symptom check list and
QEEG evaluation are linked to give rise to a neurofeedback protocol (see Appendix 7.1).
Modules or “hubs” are linked to the various
basic functional systems involved in cognition
and perception (Chen et al., 2008; Hagmann et
al., 2008; He et al., 2009). Recent neuroimaging
studies indicate that all of the various “modules”
FIGURE 7.2. Diagram of the difference between standard EEG biofeedback and Z-score EEG biofeedback. The
top system involves standard EEG biofeedback that relies on raw EEG measures such as power, coherence,
phase, amplitude asymmetries, power ratios, and an arbitrary and subjective threshold value. The bottom system is the same as the top but with a transform of the raw scores to Z scores and thus a simplification of diverse
metrics to a single metric of the Z score in which the threshold is mathematically defined as a movement toward
Z = 0. The magnitude of the Z score provides real-time feedback as to the distance between the patient’s EEG
and the EEG values in an age-matched sample of healthy, normal control subjects.
7. QEEG and EEG Biofeedback/Neurofeedback
117
FIGURE 7.3. Example of a computer-generated Symptom Check List in which the clinician first evaluates the
patient’s symptoms and complaints using clinical and neuropsychological tools, then enters a score of 0 to 10
based on the severity of the symptoms. Hypotheses formation as to the most likely scalp locations and brain
systems are based on the scientific literature that links symptoms and complaints to the locations of functional
specialization (see Appendix 7.1). From NeuroGuide 2.5.7.
FIGURE 7.4. Example of Brodmann areas as they relate to various general functions and “hubs” or “modules,” and
scalp electrode locations that “sense” electrical activity generated by various functional systems.
118
II. INSTRUMENTATION
FIGURE 7.5. Flow diagram of individualized protocol design based on linkage of patient's symptoms and complaints with surface QEEG Z scores and LORETA Z scores. The columns of the matrix are the 19 channels of
the 10/20 International electrode sites, and the rows are symptoms and QEEG EEG features. Hypotheses are
formed as to the most likely electrode site locations associated with a given symptom and complaint based on
the scientific literature. The hypotheses are then tested based on QEEG and LORETA Z scores. Weak systems
representing “loss of function” are identified when there is a match of QEEG Z scores, with the hypothesized
scalp locations based on symptoms. Compensatory locations are identified by a mismatch between hypothesized symptoms and complaints, and the locations of observed QEEG Z scores. The suggested neurofeedback
protocol that is then produced is based on the locations of the “weak” systems (see Appendix 7.1).
are dynamically linked and interactive, and that
subsets of neural groups in different modules
“bind” together for brief periods of time to mediate a given function (Sauseng & Klimesch, 2008;
Thatcher et al., 2008; Thatcher & Lubar, 2008).
An illustration of Brodmann areas and electrodes
as they relate to functional systems is shown in
Figure 7.4.
The link between a patient’s symptoms and
complaints to the localization of functional systems in the brain is based on the accumulated scientific and clinical literature from QEEG, MEG,
fMRI, PET, and SPECT studies conducted over the
last few decades, as well as the basic literature on
neurological and neuropsychological lesions. The
Russian neuropsychologist Alexander Luria (1973)
and the American neuropsychologist Hans-Lukas
Teuber (1968) were among the leading scientists
to make important links between symptoms and
complaints, and localization of functional systems in the brain. The integration of QEEG and
EEG biofeedback relies on such links as the initial
stage in the formation of NF protocols, as illustrated in Figures 7.3, 7.4, and 7.5. The first step is
to produce hypotheses about likely links between
a patient’s symptoms and complaints and the
location of functional systems based on the scientific literature prior to conducting a QEEG. The
second step is to confirm or disconfirm the link
by evaluating brain locations of deviations from
normal using QEEG and LORETA three-dimensional imaging, and the third step is to produce a
biofeedback protocol based on the match between
hypothesized locations and the QEEG and/or
LORETA evaluation (see Figure 7.5). Luria (1973)
emphasized that deregulation of neural populations is reflected by reduced homeostatic balance
in the brain, in which symptoms are represented
as “loss of function” that is often accompanied by
“compensatory” processes. One goal of the linkage of QEEG and NF is to identify and contrast
the weak or “loss of function” components in the
EEG with the compensatory processes in which
the weak systems are the initial target of the EEG
biofeedback protocol.
There are six steps that must be followed to use
the symptom check list and automatic Z-score protocol generator:
7. QEEG and EEG Biofeedback/Neurofeedback
1. Import the subject’s edited *.ng EEG file by
clicking File > Open.
2. Click Report > Symptom Check List Match.
3. Click Collection > Setup & Monitor > OK.
4. Click Collection > Neurofeedback > Surface
Neurofeedback.
5. Click Symptom Check List in the Surface EEG
Control Panel.
6. Select the Symptom(s) that best represent the
patient’s symptoms and assign a severity from
1 to 10 for each symptom. View the reduced
size and number of green circles as the QEEG
Z-score threshold is increased. Reduce the
Z-score cutoff and view an increase in size and
number of green circles. When a good fit to the
patient’s symptoms by the hypothesis test is
reached based on the clinical judgment of the
user, click OK to generate an automatic protocol (see Figure 7.6).
Figure 7.7 is an example of a 19-channel surface EEG biofeedback setup screen in Neuroguide,
where users can select a wide variety of measures
or metrics, all reduced to the single metric of the Z
score. This includes, power, coherence, phase differences, amplitude asymmetries, power ratios, and
the average reference and Laplacian montages.
Nineteen channels is a minimum number needed
in order to compute accurate average references
and the Laplacian montage, which is an estimate
of the current density vectors that course at right
angles through the skull.
119
Multiple frequencies and multiple metrics may
be selected in which a threshold must be reached
before a visual and/or auditory reward is given
(e.g., z < 2.0). The 19-channel Z-score approach
provides for seamless integration of QEEG assessment and 19-channel Z-score neurofeedback or
treatment. Because there are approximately 5,000
possible instantaneous Z scores, it is important to
limit and structure the biofeedback protocol in a
manner that best links to the patient’s symptoms
and complaints. The linkage of patient’s symptoms
and complaints as hypotheses that are confirmed
or disconfirmed by QEEG assessment is used to
develop a neurofeedback protocol. Blind and random selection of Z-score metrics runs the risk of
altering “compensatory” systems and not focusing
on the weak or “loss of function” systems that are
linked to the patient’s symptoms and complaints.
Control of the difficulty of the threshold is by
(1) lengthening the event interval and (2) lowering the Z-score threshold. To make neurofeedback easier, then, shorten the event interval and
raise the Z-score threshold. The event integration
interval is a time window that varies from 250 ms
to 1 s. In order to receive reinforcement, then,
100% of the time events within a window must
reach the Z-score criteria. By lengthening the time
window, one simultaneously reinforces reduced
variability. Thus, the time window provides a variability feedback method. Click Sound On for the
eyes-closed condition and/or use both visual and
auditory feedback with eyes open. Click Symp-
FIGURE 7.6. Seamless QEEG and neurofeedback: Approximately 50–60 minutes for a single session, in four steps
from clinical interview to QEEG to neuropathy.
120
II. INSTRUMENTATION
FIGURE 7.7. Example of 19-channel surface EEG Z-score biofeedback setup screen inside of NeuroGuide.
tom Check List if you have run a QEEG analysis and clicked Report > Create Symptom Check
List Match. If one decides not to use the Symptom
Check List, then manually select metrics, channels and frequencies and click OK to activate the
10/20 Reward display. If dual monitors are used,
then enable the monitor in the control panel >
Display Settings and depress the left mouse button over the 10/20 reward display and move it to
the second monitor. If criteria are met for all time
points in a window (e.g., Z < 2.0), then a reward
is the color green in the 10/20 locations selected
in the Z-score NF panel. The goal is to make the
10/20 head display show green as often as possible.
Start with an easy reward criteria, e.g., Z < 2.0),
then adjust the reward criteria to lower Z values
(e.g., Z < 1.0) in order to shape the client/patient
EEG features toward Z = 0.
Symptom Check List Hypotheses and QEEG
Z‑Score Tests
As explained, one must first import the patient’s
edited *.ng file and then click Report > Create
Symptom Check List Match before the Symptom
Check List is active (see Figure 7.8). Then identify
one or more of the 49 symptom(s) exhibited by the
patient/client and double-click the severity score
to activate the symptom(s) (see Figure 7.9). Enter
a severity score from 1 to 10. This creates a green
circle on the 10/20 scalp display and the size of the
circle increases as a function of the severity value.
As more symptoms are selected the Neuroguide
automatically weights the symptom locations and
scales the size of the green circles to represent
“hypotheses” of “weak systems” or “loss of function” systems (Luria, 1973).
The top-right 10/20 scalp display will change
depending on the symptom check list and the
match of the QEEG Z scores to the hypothesized
locations. The location of green circles in the
“Match” 10/20 display represents a match between
hypothesized scalp locations and observed QEEG
Z scores. The radius of a green circle is produced
by scaling with respect to the maximum average Z
score greater than the threshold for a given scalp
location. The larger the average Z score, the larger
the radius of the circle. After the user finds an
optimal link to hypothesized locations, then click
OK to automatically generate a Neurofeedback
Protocol. The automatic protocol is produced by
the cross-product of the symptom severity and the
average Z score at a given location or (S × Z)/N,
where S is the sum of the severity index from the
symptom check list for that location; Z is the aver-
7. QEEG and EEG Biofeedback/Neurofeedback
121
FIGURE 7.8. Example of Symptom Check List panel before making a symptom selection and assigning a symptom
severity score.
FIGURE 7.9. Example of the Symptom Check List after a symptom and severity are selected. A match of QEEG
Z-score deviations from normal to the hypothesized brain locations linked to the symptom is shown. The mismatch display shows brain regions that are not linked to the symptom and possibly are compensatory. The goal
is to avoid compensatory systems and target the “weak” brain systems linked to symptoms.
122
age Z scores in that location, where the absolute Z
or |Z| is greater than the threshold as determined
by the user (Default is |Z| = 2); and N is normalization by scaling to the maximum. The user can
veto or modify the automatically produced protocol by clicking clear or by clicking metrics, frequencies, auto-spectrum and cross-spectrum selections in the Surface Neurofeedback panel.
Click OK in the Symptom Check List Panel
to Return to the Surface Neurofeedback Panel
and View the Automatic Selections based on the
Symptom Check List as Hypotheses and QEEG
Z Scores as Tests of the Hypotheses. Click OK if
satisfied or modify by selecting or deselecting variables, or click Clear to start over. Then click OK in
the Surface EEG Neurofeedback panel to activate
the Neurofeedback Reinforcement Display.6
Begin Neurofeedback Using the Automatic
Symptom Match Protocol (20 Minutes)
After the Symptom Check List match and mismatch has been completed then start Neurofeedback by clicking OK. When the Z-Score Neurofeedback panel again appears, then check which
metrics have been selected by the automatic protocol process and edit, modify, or reject by clicking
“Reset.” When ready, then click OK to begin the
Neurofeedback. Move the 10/20 head display to a
second monitor and/or choose a sound feedback
in the Sound control. Select third- party vendor
II. INSTRUMENTATION
DVD/Flash and MIDI displays by clicking Display, then selecting CIS (Cybernetic Integration
Systems); Brainmaster (Multimedia control) or
3D Engine (Deymed DVD/Flash and MIDI system). The latter display systems must be purchased
directly from CIS, Brainmaster or Deymed. Neuroguide only provides access to already purchased
display products.
Figure 7.10 is the NeuroGuide 10/20 19-channel
display for Neurofeedback. Green circles (shown
here in gray) are the reinforcement when the Z
scores are less than the threshold. If cross-spectral coherence or phase, and so forth, is selected,
then only a single green circle will be present at
Cz. This is because it is not possible to provide an
unambiguous multiple-head location display when
multiple coherence or phase channels have been
selected. Therefore, all of the measures are combined into a single display (similar to what occurs
under DVD/Flash control). All of the selected protocol locations must meet the threshold criteria in
order to receive a reinforcement (single Cz display
or DVD/Flash).
Figure 7.10 shows an example of a simple 10/20
head display for feedback where the circles turn
green when threshold is met (e.g., Z < 2.0) and
provides feedback about the scalp locations that
are meeting threshold.
Figure 7.11 is an example of a progress-monitoring chart that is displayed for the clinician during the course of biofeedback. One strategy is to
develop a protocol based on the linkage to the
FIGURE 7.10. Example of 19-channel feedback display. The circles at a particular location turn green (but shown
here in gray) when threshold is reached (e.g., Z < 2.0).
7. QEEG and EEG Biofeedback/Neurofeedback
123
FIGURE 7.11. Example of progress charts that a clinician views during the course of neurofeedback. The idea is
to shape the patient’s brain toward the center of the normal healthy reference population, where Z = 0. Initially
the threshold is set so that the patient receives a high rate of reinforcement (e.g., Z < 2.0), then to lower the
threshold and make it more difficult (e.g., Z < 1.5), then, as the patient again receives a high rate of reinforcement, to again lower the threshold (e.g., Z < 1.0) so as to shape the brain dynamics using a standard operant
conditioning procedure.
patient’s symptoms and complaints, as discussed
previously, then to set the Z-score threshold, so
that it is easy for the subject to meet threshold and
therefore produce a high rate of successful “Hits”
or rewards. The second step is to lower the threshold and make the feedback more difficult (e.g., Z
< 1.5) and as the patient or client gains control
and receives a high rate of reinforcement to lower
the threshold again (e.g., Z < 1.0) in a “shaping”
process in which operant conditioning is used to
move the patient’s brain metrics toward the center
of the normal reference population, or Z = 0.
Neuroimaging NF or Real‑Time LORETA
Z‑Score Biofeedback
Improved accuracy in the linkage between a
patient’s symptoms and complaints and the localization of functional systems can be achieved by
the biofeedback of real-time, three-dimensional
locations or voxels in the brain. This method
has been successfully implemented with fMRI for
chronic pain, obsessive–compulsive disorders, and
anxiety disorders (Apkarian, 1999; Bray, Shimojo,
& O’Doherty, 2007; Caria et al., 2007; de Charms,
2008; de Charms et al., 2004; Weiskopf et al.,
2003; Yoo et al., 2006).
However, fMRI biofeedback, also referred to as
“neuroimaging therapy,” has several significant
limitations in comparison to LORETA 3-dimensional EEG biofeedback7: (1) a long time delay
between a change in localized brain activity and
the feedback signal (e.g., 20 seconds to minutes
for fMRI), whereas LORETA EEG biofeedback
signals involve millisecond delays; (2) fMRI
only provides indirect measures of neural activity, because blood flow changes are delayed and
secondary to changes in neural activity, whereas
EEG biofeedback is a direct measure of neural
electrical activity; and (3) expense—fMRI costs
$3 million for the MRI machine plus $30,000 per
month for liquid helium, whereas EEG biofeedback equipment and maintenance costs are less
than $10,000. The spatial resolution of LORETA
source localization is approximately 7 mm3, which
is comparable to the spatial resolution of fMRI.8
fMRI, however, offers the advantage of imaging of
noncortical structures such as the striatum, thalamus, cerebellum, and other brain regions, whereas
124
QEEG is limited to imaging of cortical dipoles
produced by pyramidal cells. Nonetheless, even
with this limitation, several studies have proven
that biofeedback using LORETA real-time, threedimensional sources is feasible and results in positive clinical outcomes (Cannon, Lubar, Thornton,
Wilson, & Congedo; 2005; Cannon et al., 2006,
2007; Cannon & Lubar, 2007; Cannon, Lubar,
Sokhadze, & Baldwin, 2008; Cannon, Congredo,
Lubar, & Hutchens, 2009; Lubar, Congredo, &
Askew, 2003). Figures 7.12 and 7.13 show examples of LORETA Z-score EEG biofeedback with
reinforcement toward Z = 0 as a method to reinforce stability and increased efficiency of information processing in brain networks linked to the
patient’s symptoms.
Future Directions of QEEG and NF
Dramatic improvements in sensitivity and spatial
resolution of neural sources linked to patient’s
symptoms and complaints have occurred in the 80
years since EEG was first discovered. The development of EEG biofeedback occurred almost in
parallel to the sudden growth of QEEG in the
1960s, and today there is ever more integration of
QEEG for clinical evaluation and EEG biofeedback for treatment. QEEG assessment and EEG
biofeedback treatment are inextricably bound,
in that advancements in sensitivity and spatial
resolution of QEEG as a clinical evaluation tool
are immediately translated to treatment using
biofeedback. There has been a veritable explosion of new discoveries in neuroscience related to
II. INSTRUMENTATION
FIGURE 7.12. Example of LORETA Z score biofeedback
in which reinforcement is toward Z = 0 or increased
stability in Brodmann areas linked to symptoms. A
change in the protocol occurred after the first session, then a steady movement toward greater stability in the symptom linked networks occurred in five
sessions. Unpublished clinical result reprinted with
permission from Wesley D. Center, PhD).
basic mechanisms of memory, attention, arousal
and cognition, and correlations to various neurological, psychiatric, and psychological disorders.
In parallel with these advancements, it is expected
that EEG biofeedback will keep pace by advancing
new methods of noninvasive treatment of a wide
variety of clinical disorders. Today high-speed
computers are available, in which 19 or more
channels of EEG can be measured, and biofeedback is applied in almost the same amount of time
FIGURE 7.13. Top: An example of reduced Z-score deviation from normal in the cingulate gyrus in after 10 sessions of Z-score neurofeedback. Bottom: An example of reduced Z scores in the right superior frontal gyrus
after 10 sessions of Z-score neurofeedback. Unpublished clinical result reprinted with permission of William A.
Lambos, PhD).
125
7. QEEG and EEG Biofeedback/Neurofeedback
as one- or four-channel biofeedback was applied in
the past. Thus, there are few practical limitations
in using high-density electrode arrays for EEG biofeedback in the future. As knowledge about basic
neurophysiological control measures emerge (e.g.,
thalamic-mediated phase shift and phase lock,
coherence, cross-frequency phase synchrony, and
LORETA spatial coordination of brain modules
underlying perception and cognition), the clinical
efficacy of EEG biofeedback will improve. Given
the expense of health care and the need for noninvasive treatment, the marriage of QEEG and EEG
biofeedback will continue to evolve.
Notes
1. The electroencephalogram is measured from the
scalp surface and is produced by the algebraic summation
of cortical synaptic potentials.
2. Since approximately 1975, it has been very difficult
even to publish studies that use only visual examination of
EEG traces. The estimate of 90,000 arises when one uses
the search term “EEG” and examines the abstracts to confirm that quantification of EEG was used. It is necessary
to use the search term “EEG” and not “QEEG,” because
the National Library of Medicine indexes articles based on
words in the titles and most QEEG studies do not use the
term “QEEG” in their titles.
3. While EEG biofeedback is sometimes referred to as
“neurofeedback,” the latter term is not specific, since many
treatments other than EEG may involve neurofeedback.
However, in this chapter, these terms are considered synonymous and are used interchangeably.
4. Spectral analysis includes space and time sequences
that are transformed, such as joint time–frequency analysis,
fast Fourier transform (FFT), and all other methods that
decompose EEG energies at different frequencies in space
and time.
5. Simultaneous suppression of excessive theta and
reinforcement of deficient beta is achieved by using a absolute Z-score threshold, which is a simplification compared
to standard raw score EEG biofeedback. For example, if the
threshold is set to an absolute value of Z < 2, then reduced
theta amplitude and elevated beta amplitude will both be
rewarded when the instantaneous EEG event exhibits a Z <
2 theta and beta power value.
6. Initially only absolute power, coherence, and phase
differences are used in the automatic Symptom Check List
protocol. This is because relative power is always a distortion and can be in conflict with absolute power when both
are used at the same time. Because absolute power is the
“mother” of relative power and unambiguously represents
the degree of local synchrony of EEG generators, a limitation to absolute power is best. Amplitude asymmetry is
not used because of a similar ambiguity (i.e., absolute power
differences cannot be resolved without reference to the
“mother” of amplitude asymmetry); therefore, amplitude
asymmetry is redundant. Also, coherence and phase are
limited to linked ears and are not available for the Laplacian and Average Reference montages, because these measures are not valid except with linked ears. A future release
will include phase shift and phase lock duration.
7. LORETA means “low-resolution electromagnetic
tomography” (Pascual-Marqui et al., 1994). Since the
inception of this method in 1994, there have been over 500
peer-reviewed publications (see www.uzh.ch/keyinst/newloreta/quoteloreta/papersthatquoteloreta05.htm for a partial
list of this literature).
8. The voxel resolution of LORETA is 7 mm3, which
is adequate spatial resolution, because the Brodmann
areas are much greater in volume than 7 mm3. Also, the
biological resolution of fMRI may be worse than that of
LORETA, because it depends on the vascular architecture
of the brain. For example, Ozcan, Baumgartner, Vucurevic,
Stoeter, and Treede (2005) showed that maximal fMRI spatial resolution is > 12 mm3.
APPENDIX 7.1. Symptom Check List
Symptoms
Problems with perception of letters
Slow reader
Dyslexia—letter reversal
Problems with spatial perception
Orientation in space problems
Receptive language problems
Insensitive to others’ emotional expressions
Blurred vision
Obsessive thoughts about self
Migraine headaches
Symptoms of fibromyalgia
Auditory sequencing problems
Short-term memory problems
Face recognition problems
Receptive language problems
Obsessive self-examination
Word-finding problems
Chronic pain
Poor skilled motor movements
Speech articulation problems
Balance problems
Decreased tactile or skin sensitivity
Problems recognizing objects by touch
Depression (sad and blue)
Problems with multitasking
126
Slowness of thought—easily confused
Poor judgment
Attention deficits—easily distractible
Hyperactive and/or agitated
Obsessive thoughts and/or hyperfocused
Compulsive behaviors and/or thoughts
Sequential planning problems
Executive function problems
Poor social skills
Oppositional defiant conduct
Problems concentrating
Mood swings
Impulsive behaviors
Low threshold for anger and loss of control
Self-esteem problems
Failure to initiate actions
Generalized anxiety
Insensitive to others’ feelings
Difficulty comprehending social cues
Anosognosia—denial of a problem
Dyscalcula—problems calculating
Delusional
Low motivation
References
Adey, W. R., Walter, D. O., & Hendrix, C. E. (1961). Computer techniques in correlation and spectral analyses
of cerebral slow waves during discriminative behavior.
Experimental Neurology, 3, 501–524.
Apkarian, A. V. (1999). Functional magnetic resonance
imaging of pain consciousness: Cortical networks of
pain critically depend on what is implied by “pain.” Current Review of Pain, 3, 308–315.
Berger, H. (1934). Uber das Electrenkephalogramm des
Menschen. Neunte Mitteilungj. Archiv für Psychiatrie
und Neverkrankheiten, 102, 538–557.
Bray, S., Shimojo, S., & O’Doherty, J. P. (2007). Direct
instrumental conditioning of neural activity using functional magnetic resonance imaging-derived reward feedback. Journal of Neuroscience, 27, 7498–7507.
Buzsaki, G. (2006). Rhythms of the brain. New York: Oxford
University Press.
Cannon, R., Congredo, M., Lubar, J., & Hutchens, T.
(2009). Differentiating a network of executive attention: LORETA neurofeedback in anterior cingulate and
dorsolateral prefrontal cortices. International Journal of
Neuroscience, 119(3), 404–441.
Cannon, R., & Lubar, J. [F.] (2007). EEG spectral power and
coherence: Differentiating effects of Spatial-Specific
Neuro-Operant Learning (SSNOL) utilizing LORETA
II. INSTRUMENTATION
neurofeedback training in the anterior cingulate and
bilateral dorsolateral prefrontal cortices. Journal of Neurotherapy, 11(3), 25–44.
Cannon, R., Lubar, J. F., Congedo, M., Gerke, A., Thornton, K., Towler, K., & Hutchens, T. (2007). The effects
of neurofeedback training in the cognitive division of
the anterior cingulate gyrus. International Journal of Science, 117(3), 337–357.
Cannon, R., Lubar, J., Gerke, A., Thornton, K., Hutchens,
T., & McCammon, V. (2006). EEG spectral-power and
coherence: LORETA neurofeedback training in the
anterior cingulate gyrus. Journal of Neurotherapy, 10(1),
5–31.
Cannon, R., Lubar, J., Sokhadze, E., & Baldwin, D. (2008).
LORETA neurofeedback for addiction and the possible
neurophysiology of psychological processes influenced:
A case study and region of interest analysis of LORETA
neurofeedback in right anterior cingulate cortex. Journal
of Neurotherapy, 12(4), 227–241.
Cannon, R., Lubar, J., Thornton, K., Wilson, S., & Congedo, M. (2005). Limbic beta activation and LORETA:
Can hippocampal and related limbic activity be recorded
and changes visualized using LORETA in an affective
memory condition? Journal of Neurotherapy, 8(4), 5–24.
Caria, A., Veit, R., Sitaram, R., Lotze, M., Weiskopf, N.,
Grodd, W., & Birbaumer, N. (2007). Regulation of anterior insular cortex activity using real-time fMRI. NeuroImage, 35, 1238–1246.
Chen, Z. J., He, Y., Rosa-Neto, P., Germann, J., & Evans,
A. C. (2008). Revealing modular architecture of human
brain structural networks by using cortical thickness
from MRI. Cerebral Cortex, 18, 2374–2381.
Collura, T. F., Thatcher, R. W., Smith, M. L., Lambos,
W. A., & Stark, C. R. (2008). EEG biofeedback training using live Z scores and a normative database. In
T. Budzinsky, H. Budzinsky, J. Evans, & A. Abarbanel (Eds.), Introduction to QEEG and neurofeedback:
Advanced theory and applications (pp. 103–142). San
Diego, CA: Academic Press.
Cooper, R., Osselton, J. W., & Shaw, J. G. (1974). EEG technology. London: Butterworth.
deCharms, R. C. (2008). Applications of real-time fMRI.
Nature Neuroscience, 9, 720–729.
deCharms, R. C., Christoff, K., Glover, G. H., Pauly, J. M.,
Whitfield, S., & Gabrieli, J. D. E. (2004). Learned regulation of spatially localized brain activation using realtime fMRI. NeuroImage, 21, 436–443.
Fox, S. S., & Rudell, A. P. (1968). Operant controlled neural event: Formal and systematic approach to electrical
codifing of behavior in brain. Science, 162, 1299–1302.
Freeman, W. J., Burke, B. C., & Homes, M. D. (2003).
Aperiodic phase re-setting in scalp EEG of beta-gamma
oscillations by state transitions at alpha-theta rates.
Human Brain Mapping, 19(4), 248–272.
Freeman, W. J., & Rogers, L. J. (2002). Fine temporal resolution of analytic phase reveals episodic synchronization
by state transitions in gamma EEGs. Journal of Neurophysiology, 87(2), 937–945.
Gomez, J., & Thatcher, R. W. (2001). Frequency domain
equivalence between potentials and currents using
7. QEEG and EEG Biofeedback/Neurofeedback
LORETA. International Journal of Neuroscience, 107,
161–171.
Hagmann, P., Cammoun, L., Gigandet, X., Meuli, R.,
Honey, C. J., Wedeen, V. J., et al. (2008). Mapping the
structural core of human cerebral cortex. PLoS Biology,
6(7), e159.
He, Y., Wang, J., Wang, L., Zhang, C. J., Yan, C., Yang, H.,
et al. (2009). Uncovering intrinsic modular organization
of spontaneous brain activity in humans. PLoS ONE
4(4), e5226.
Hughes, J. R., & John, E. R. (1999). Conventional and
quantitative electroencephalography in psychiatry.
Neuropsychiatry, 11(2), 190–208.
John, E. R. (1977). Neurometrics: Quantitative electrophysiological analyses. In E. R. John & R. W. Thatcher
(Eds.), Functional neuroscience (Vol. 2). Hillsdale, NJ:
Erlbaum.
John, E. R., Karmel, B., Corning, W., Easton, P., Brown,
D., Ahn, H., et al. (1977). Neurometrics: Numerical
taxonomy identifies different profiles of brain functions
within groups of behaviorally similar people. Science,
l96, 1393–1410.
John, E. R., Prichep, L. S., & Easton, P. (1987). Normative
data banks and neurometrics: Basic concepts, methods
and results of norm construction. In A. Remond (Ed.),
Handbook of electroencephalography and clinical neurophysiology: Vol. III. Computer analysis of the EEG and
other neurophysiological signals (pp. 449–495). Amsterdam: Elsevier.
John, E. R., Prichep, L. S., Fridman, J., & Easton, P. (1988).
Neurometrics: Computer assisted differential diagnosis
of brain dysfunctions. Science, 293, 162–169.
Kamiya, J. (1971). Biofeedback training in voluntary control of EEG alpha rhythms. California Medicine, 115(3),
44–49.
Lubar, J., Congedo, M., & Askew, J. H. (2003). Low-resolution electromagnetic tomography (LORETA) of cerebral activity in chronic depressive disorder. International
Journal of Psychophysiology, 49(3), 175–185.
Luria, A. R. (1973). The working brain: An introduction to
neuropsychology. London: Penguin Press.
Matousek, M., & Petersen, I. (1973). Automatic evaluation
of background activity by means of agee-dependent EEG
quotients. Electroencephalography and Clinical Neurophysiology, 35, 603–612.
Niedermeyer, E., & Lopes da Silva, F. (1995). Electroencephalography: Basic principles, clinical applications and related
fields. Baltimore, MD: Williams & Wilkins.
Ozcan, M., Baumgartner, U. L., Vucurevic, G., Stoeter, P.,
& Treede, R. D. (2005). Spatial resolution of fMRI in the
human parasylvian cortex: Comparison of somatosensory and auditory activation. NeuroImage, 25, 877–887.
Pascual-Marqui, R. D., Michel, C. M., & Lehmann, D.
(1994). Low resolution electromagnetic tomography: A
new method for localizing electrical activity in the brain.
International Journal of Psychophysiology, 18, 49–65.
Sauseng, P., & Klimesch, W. (2008). What does phase
information of oscillatory brain activity tell us about
cognitive processes? Neuroscience and Biobehavioral
Reviews, 32(5), 1001–1013.
127
Sterman, M. B. (1973). Neurophysiologic and clinical studies of sensorimotor EEG biofeedback training: Some
effects on epilepsy. Seminars in Psychiatry, 5(4), 507–525.
Teuber, H. (1968). Disorders of memory following penetrating missile wounds of the brain. Neurology, 18(3),
287–298.
Thatcher, R. W. (1989). EEG normative databases and EEG
biofeedback. Journal of Neurotherapy, 2(4), 8–39.
Thatcher, R. W. (1999). EEG database guided neurotherapy. In J. R. Evans & A. Abarbanel (Eds.), Introduction
to quantitative EEG and neurofeedback (pp. 26–44). San
Diego, CA: Academic Press.
Thatcher, R. W. (2000a). EEG operant conditioning (biofeedback) and traumatic brain injury. Clinical EEG and
Neuroscience, 31(1), 38–44.
Thatcher, R. W. (2000b, September). An EEG least action
model of biofeedback. Paper presented at the eighth
annual ISNR conference, St. Paul, MN.
Thatcher, R. W., & Collura, T. (2006, September). Z score
biofeedback and new technology. Paper presented at the
International Society of Neuronal Regulation annual
meeting, Atlanta, GA.
Thatcher, R. W., & Lubar, J. F. (2008). History of the scientific standards of QEEG normative databases. In
T. Budzinsky, H. Budzinsky, J. Evans, & A. Abarbanel (Eds.), Introduction to QEEG and neurofeedback:
Advanced theory and applications (pp. 29–62). San Diego,
CA: Academic Press.
Thatcher, R. W., North, D., & Biver, C. (2005). Evaluation
and validity of a LORETA normative EEG database.
Clinical EEG and Neuroscience, 36(2), 116–122.
Thatcher, R. W., North, D., & Biver, C. (2008). Intelligence
and EEG phase reset: A two-compartmental model of
phase shift and lock. NeuroImage, 42(4), 1639–1653.
Thatcher, R. W., North, D., & Biver, C. (2009a). Self organized criticality and the development of EEG phase
reset. Human Brain Mapping, 30(2), 553–574.
Thatcher, R. W., North, D. M., Neurbrander, J., Biver, C. J.,
Cutler, S., & DeFina, P. (2009b). Autism and EEG phase
reset: Deficient GABA mediated inhibition in thalamocortical circuits. Developmental Neuropsychology, 34(6),
780–800.
Thatcher, R. W., Walker, R. A., Biver, C., North, D., &
Curtin, R. (2003). Quantitative EEG normative databases: Validation and clinical correlation, Journal of
Neurotherapy, 7(3/4), 87–122.
Weiskopf, N., Veit, R., Erb, M., Mathiak, K., Grodd, W.,
Goebel, R., et al. (2003). Physiological self-regulation
of regional brain activity using real-time functional
magnetic resonance imaging (fMRI): Methodology and
exemplary data. NeuroImage, 19(3), 577–586.
Woody, R. H. (1966). Intra-judge reliability in clinical EEG.
Journal of Clinical Psychology, 22, 150–159.
Woody, R. H. (1968). Inter-judge reliability in clinical EEG.
Journal of Clinical Psychology, 24, 251–261.
Yoo, S. S., O’Leary, H. M., Fairneny, T., Chen, N. K.,
Panych, L. P., Park, H., et al. (2006). Increasing cortical activity in auditory areas through neurofeedback
functional magnetic resonance imaging. NeuroReport,
17(12), 1273–1278.
Chapter 8
Introduction to Psychophysiological Assessment
and Biofeedback Baselines
John G. Arena and Mark S. Schwartz
We hope this chapter successfully demystifies psychophysiological assessment and enhances applied
psychophysiologists’ ability to formulate clinical
questions and employ psychophysiological assessments to answer them.
The process of learning how to use psychophysiological assessments to answer relevant clinical
questions first involves the review of some basic
concepts in applied psychophysiology, namely, (1)
the measures that are generally used by applied
psychophysiologists, (2) the basic concepts in
psychophysiology that are most relevant to the
applied psychophysiologist, and (3) the general
conditions that the biofeedback therapist would
likely employ in such assessments. We stress what
is perhaps the most important methodological
issue in any type of assessment, that is, temporal
stability (i.e., reliability) of psychophysiological
measures. We emphasize two conditions that are
of essential importance to any psychophysiological
assessment: baselines and adaptation periods.
response, cardiovascular activity (heart rate, blood
pressure, and vasomotor activity), and respiration
(generally respiration rate and depth). Measures
of electroencephalography (EEG) used in neurofeedback require specialized training and are
outside the expertise of the average biofeedback
clinician (as well as our own); therefore, they are
not discussed in this chapter. For references about
how basic measures are recorded and interpreted,
see Peek (2003, and Chapter 3, this volume) for
SEMG, skin temperature, and electrodermal
responses. For more on SEMG, see Basmajian and
De Luca (1985) and Criswell (2011). For measures
of cardiovascular activity, see Berntson, Quigley,
and Lozano (2007). For respiration and heart rate
variability, see Lorig (2007) and Sutarto, Abdul
Wahab, and Mat Zin (2010).
Basic Concepts in Psychophysiology
Basic concepts in applied psychophysiology and
the scientific method in general with which biofeedback clinicians need to be familiar are autonomic balance, individual response stereotypy,
stimulus–response specificity, the law of initial
values, homeostasis, orienting and defensive
responses, carryover effects, and especially temporal stability of the measures (Andreassi, 2007;
Arena & Schwartz, 2003).
Measures Used
in Psychophysiological Assessments
The measures generally employed in psychophysiological assessments are surface electromyographic
activity (SEMG); skin surface temperature (e.g.,
fingers, hands); and measures of electrodermal
128
8. Psychophysiological Assessment and Biofeedback Baselines
Autonomic Balance
“Autonomic balance” refers to response patterning
of the autonomic nervous system (Abboud, 2010;
Sturgis & Arena, 1984). It has long been believed
that individuals who are exposed to a stimulus of
some sort respond either with sympathetic or parasympathetic response activation. In 1917, Eppinger
and Hess were the first to classify individuals as
vagotonic (parasympathetic responders) and sympatonic (sympathetic) responders. In 1966, Wenger
created a score of autonomic balance based on
how an individual’s electrodermal response, heart
rate, diastolic blood pressure, and salivation output
responded to various stimuli. Autonomic balance
scores have been used in a variety of populations,
including those with anxiety-based disorders,
schizophrenia, hypertension, headache, antisocial personality and attention-deficit/hyperactivity disorder, and low scores tend to be related to
increased susceptibility to both physical and psychological disease, whereas high scores are associated with greater mental and physical health.
Individual Response Stereotypy and Stimulus–
Response Stereotypy
“Individual response stereotypy” and “stimulus–
response specificity” are somewhat complex categorization schemes that also involve the measurement of a number of psychophysiological measures
to look for patterns in the responses. Stimulus–
response specificity refers to different stimuli, such
as a cognitive versus a physical stressor, producing
idiosyncratic patterns of responding. Individual
response stereotypy, on the other hand, is seen
when an individual evidences a single, distinctive
response pattern to all stimuli (Sternbach, 1966).
For example, one individual may characteristically respond to stressful events with increased
heart rate, whereas another might evidence lowered hand surface temperature (stereotypy); alternatively, one individual may reliably respond to a
mental arithmetic task with increased respiration
rate and lowered respiration depth, while responding to an ischemic pain stressor with increased
frontal SEMG.
The Law of Initial Values
The “law of initial values” refers to the effect that
prestimulus values of a particular psychophysiological measure have on that response’s magnitude of psychophysiological reactivity to a specific
129
stimulus (Wilder, 1950). The higher the level of
the measure prior to presentation of a stressful
stimulus, the smaller the increase in response to
the stressor (often referred to as a “ceiling effect”).
Conversely, the higher the level of the measure
prior to presentation of a relaxing stimulus, the
larger the decrease in response to the relaxing
stimulus (prestimulus–response values that are
low prior to the presentation of a relaxing stimulus lower the magnitude of the response and are
often referred to as “floor effects”). While the law
has generally been shown to hold for measures of
respiration and cardiovascular activity (e.g., heart
rate and the vasomotor response), measures such
as salivation and electrodermal response have not
been found to be influenced by prestimulus values.
Homeostasis
“Homeostasis” refers to the tendency of any organism to strive to maintain a state of equilibrium
or rest (Baptista, 2006). Homeostasis is believed
to be maintained by a negative feedback loop, a
hypothesized bodily mechanism that provides
information that directs the physiological system to decrease activity if levels of functioning
are higher than normal, or to increase activity if
levels are diminished compared to normal. Thus,
all organisms strive to return to prestimulus levels of physiological arousal when presented with
any stimulus. Applied psychophysiology research
has demonstrated that there are limits beyond
which increases and decreases in the physiological
response cannot be trained.
Orienting and Defensive Responses
“Orienting and defensive responses” refer to the
way organisms respond to unique and novel stimuli; the response is both behavioral and physiological (Campbell, Wood, & McBride, 1997; Dawson,
Schell, & Filion, 2007). The orienting response
can be viewed as the “What is it?” response. It
typically involves “an increased sensitivity of
the sensory organs, body orientation towards the
stimulus, increased muscle tone with a reduction of irrelevant motor activity, EEG activation,
vasoconstriction of the peripheral vascular system, vasodilatation of the cranial vascular system,
increased skin conductance, respiration amplitude
increase accompanied by decreased respiration
rate, and a slowing of the heart rhythm” (Sturgis
& Arena, 1984, p. 16). Because it is impossible to
ascertain which portions of the initial responses
130
to a stimulus are orienting responses and which
are actual responses to the stimulus, applied psychophysiologists generally disregard the responses
to the beginning stimuli when analyzing response
patterns. The orienting response generally habituates quickly, but it has been shown that responses
to both psychologically and physiologically relevant stimuli habituate at much slower rates. In
contrast to the increased attention toward a stimulus that is the orienting response, the defensive
response is defined as a turning away of attention,
usually from a painful stimulus or a stimulus that
is too intense. Physiologically it is similar to the
orienting response, but with increased heart rate
and constriction of the cranial vascular system. It
is generally believed that the orienting response
habituates more quickly than the defensive
response.
Carryover Effects
The term “carryover effect” is a basic research
methodology notion that refers to the effect that
prior research conditions can have on subsequent
conditions (Box, Hunter, & Hunter, 2005). It is
important to note that in addition to carryover
effects from the actual experience of the condition, a temporal factor may also cause participants to fatigue over time (see below) and show a
change in response pattern unrelated to the condition being evaluated (Sturgis & Arena, 1984).
For example, a biofeedback therapist may present
a variety of stressful conditions in a psychophysiological assessment, and the presentation of stressor
1 may affect stressor 2, and the presentation of
stressors 1 and 2 may affect stressor 3; in addition
to the specific carryover effects of stressors 1 and
2, stressor 3 may have been affected by the patient
becoming fatigued over time.
There are two generally acceptable solutions to
the problem of carryover effects in research methodology. First, and most conservative, is to avoid
the use of a repeated measures design and use
only an independent groups design. The primary
limitation of this design for clinical biofeedback
therapists is the need for large sample sizes and,
of course, the average clinician does not have the
resources, the time, or the patience to conduct such
studies. A second solution to the problem of carryover effects is to use a counterbalanced design,
that is, to vary the order of the conditions in a random manner. This is something that the average
clinician can do, since it does not involve a large
number of patients. One major concern about this
II. INSTRUMENTATION
design is the inherent assumption that carryover
effects are equivalent between the differing conditions. That is, the carryover effect of stressor 1 is
exactly the same as the carryover effects of stressors 2 and 3. The possibility of interactions among
the conditions and differential practice effects is
not controlled. Unfortunately, there has been little research examining the possibility of carryover
effects in applied psychophysiology, and the limited available data are inconclusive.
Temporal Stability of Measures
Temporal stability, or reliability, of measures used
in psychophysiological assessment and treatment
has been a topic of increasing importance in the
past 30 years. If an assessment measure is not stable
over time, it is a poor indicator of what is purportedly tested. Given the wide range of factors that
can affect the magnitude of the various psychophysiological responses, not surprisingly, there can
be difficulties in obtaining stable recordings across
time. Moreover, when limits are set on the reliability of a measure, there is generally an inability to
obtain a high estimate of validity (i.e., whether the
measure actually records a true representation of
the concept supposedly assessed).
To illustrate the difficulties involved in temporal stability of psychophysiological assessment,
consider the following scenario: Ms. X, a single
mom who suffered from chronic headaches, came
to her therapist’s office for a pretreatment psychophysiological assessment. The appointment was at
5:30 P.M. on a cold winter’s day in January. Ms. X
got little sleep the night before because she was
up all night taking care of her 6-year-old son, who
had a stomach virus. Work was extremely difficult
(she worked as a secretary in a lawyer’s office), and
she was guzzling coffee all day just to stay alert.
Ms. X had an argument with a coworker about
who was responsible for a botched copying job just
before leaving at 5:00 P.M. Traffic was horrible,
and she was stuck behind a slow driver in the left
lane. Ms. X arrived at the office just in time, after
finding a parking spot two blocks away, where she
was promptly ushered into the biofeedback and
psychophysiological assessment laboratory and
given a brief explanation by a therapist she had
only met once. Sensors were attached to her forehead, neck, shoulders, two of her fingertips, and
around her chest. She was asked by the therapist,
who was speaking to her over an intercom from an
adjoining room, to perform several tasks, including immersing her hand in a bucket of ice water
8. Psychophysiological Assessment and Biofeedback Baselines
and counting backwards from 999 by 7’s. She wondered the entire time how long the session would
last, because the sitter had to leave by 7:00 P.M.
At the end of the assessment, she was asked if she
had any questions and she responded, “No.” An
appointment was scheduled for treatment and she
rushed home.
Halfway through treatment, on a rather warm
day in mid-March, the therapist retested Ms. X to
“determine if treatment has had any effects yet on
your body’s responses to stressful situations.” This
day, however, in contrast to the testing in January, Ms. X had a great night’s sleep. Her 6-yearold had come home the day before with all A’s
on his report card. The appointment was at 9:30
A.M., and she received permission from her boss
to take the morning off. She found a parking spot
directly in front of the office and while waiting for
10 minutes, planned a shopping trip for later that
morning. She had lost 10 pounds since beginning
therapy, and her clothes were now loose. She was
ushered into the biofeedback and psychophysiological assessment laboratory by her therapist,
and they were both joking, since they were now
well-acquainted with each other. The procedure
was repeated. Ms. X’s reaction to the stressors was
greatly reduced, and the therapist stated, “We now
have hard evidence that the treatment has already
had an effect on your body’s responses to stress.”
Is this actually the case? Can we truly arrive at
that conclusion? Could the reduction in magnitude of the response to the stressors merely have
been a result of repeating the test? Could it have
been a result of more sleep? Weight loss? Differences in time of day? Comfort with the therapist?
Seasonal temperatures? Such questions underscore
the vital importance of temporal stability research
and the relevance of factors such as age, gender,
race, weight loss, situational and trait anxiety, and
so forth.
Review of the Temporal Stability Literature
Because of the topic’s importance, we review the
essential literature concerning temporal stability of psychophysiological responses. We focus on
measures of SEMG and surface skin temperature,
because these are the two non-EEG responses
most often used in biofeedback training.
Sturgis (1980) was one of the first researchers to
investigate temporal stability. She examined the
frontal SEMG response, bilateral cephalic vasomotor response, and digital vasomotor responses
in 10 subjects with migraine and 10 with tension
131
headaches. Overall test–retest reliability of the
measures was 0.31, which, albeit statistically significant, accounted for a small proportion of the
variance.
We (Arena, Blanchard, Andrasik, Cotch,
& Myers, 1983) began our studies of reliability
using a normal population. Fifteen undergraduate subjects were assessed on multiple response
measures (frontal and forearm flexor electromyography (EMG), heart rate, skin resistance level,
hand surface temperature, and cephalic vasomotor response) under multiple stimulus conditions
(baseline, relax whole body deeply, warm hands,
relax forehead, mental arithmetic, positive imagery, stressful imagery, cold pressor) on multiple
occasions (days 1, 2, 8, and 28). Subjects were
screened for medical conditions, and all assessments occurred at approximately the same time
of day. Three forms of reliability coefficients were
computed for each response measure: coefficients
on absolute scores and two coefficients on relative measures—percentage of change from baseline and change scores from baseline to stressful
conditions (the term “relative measures” refers to
any measure other than the actual raw value of the
response; i.e., the actual raw score of the psychophysiological response has been changed or transformed in some manner. This is typically done to
decrease the wide variability that is often found
in psychophysiological responses, as well as to
control, in SEMG studies particularly, differences
between various equipment brands, etc.).
Results indicated that, for absolute values of the
measures, only frontal SEMG seemed consistently
reliable, while hand surface temperature was reliable if sessions were repeated within 1 week. Heart
rate and forearm flexor SEMG were somewhat less
consistently reliable. Lower reliability coefficients
were generally obtained when responses were
treated as relative measures.
We concluded that investigators must first
ascertain the reliability of these measures on their
respective subject population and subsequently
employ in their research only those measures
that are found to be reliable with that population.
Another conclusion was that since frontal SEMG
and hand surface temperature were the primary
biofeedback modalities, and they were fairly reliable, clinicians should merely be wary of falsely
attributing baseline hand temperature increases
solely to biofeedback training, which may result
partly from habituation to the clinical situation.
Speckenbach and Gerber (1999) essentially
replicated our results (Arena et al., 1983; Arena,
132
1984) and those of others (Sturgis, 1980; Schaffer, Sponsel, Kice, & Hollensbe, 1991) concerning
the reliability of blood volume pulse. Burnham,
McKinley, and Vincent (2006) measured intrasession skin temperatures bilaterally in 17 healthy
subjects in the hand, forearm, shoulder, thigh,
shin, and foot using a thermistor and two infrared
thermometers. They found that intrasession reliability was high and similar for each device used
(all r's ≥ .9), also replicating our results.
Other investigators suggested examining a more
complex patterning of the responses. Waters, Williamson, Bernard, Blouin, and Faulstich (1987)
built on the work of Manuck and Schaefer (1978).
These experimenters differentiated groups of
“reactors” and “nonreactors” based on cardiovascular responses to difficult cognitive tasks, and
found stability in these designations when subjects
were retested a week later.
Waters et al. (1987) compared 30 college students, using five stimuli and 10 psychophysiological measures over 2 weeks. The magnitude and
range of correlations were similar to those in the
Arena et al. (1983) study. They also analyzed individual response specificity with the Profile Similarity Index (PSI; Buco & Blouin, 1983), providing
a single index of overall similarity or reliability of
the two response profiles. For reactivity, with the
PSI, at least 87% of the subjects showed similarity.
Probably their most revealing analysis was
derived from comparing subjects’ ranks (on a scale
of 1 to 10) with a ranked hierarchy of standardized
physiological scores for each subject and for the 10
psychophysiological measures. Waters et al. (1987)
averaged the ranks across the stimulus procedures.
Fifteen of 30 subjects ranking 10th in one session,
and 14 ranking ninth, were ninth or 10th in the
second session. Similarly, 29 of the subjects ranked
first or second in the first session were first or second in the second session. Those ranked between
these extremes in the first session varied considerably in the second, and some went to the other
extreme. The researchers concluded that “it is thus
clear that the most extreme responses in an individual's psychophysiological response hierarchy
are the most stable (reliable) across experimental
sessions” (p. 219).
Building on the work of Waters et al. (1987),
Arena, Goldberg, Saul, and Hobbs (1989a) argued
that analysis of both individual response stereotypy and stimulus–response specificity might
provide a perspective on reliability not available
from the traditional Pearson correlational procedures commonly employed, or from an analysis of
II. INSTRUMENTATION
only individual response stereotypy. A multivariate response pattern approach might have some
predictive validity. For example, some (Engel,
1960) might argue that clinical populations would
have more stability than normals in the particular
response system presumed to be abnormal (e.g.,
patients with low back pain in the paraspinal
muscles, patients with headache in the forehead
or upper trapezius muscles), whereas others (Sternbach, 1966) might argue the opposite.
We therefore examined the temporal stability
of three response measures (forehead EMG, hand
surface temperature, heart rate) on 64 college and
community volunteers during four sessions over
a monthlong interval. Each session included an
adaptation period, a baseline condition, a cognitive stressor (serial 7’s), and a physical stressor (a
cold pressor task). Reliability coefficients on the
absolute scores across conditions were, for the
most part, modest and statistically significant.
Treating the responses as relative measures again
produced smaller and less frequently significant
correlational coefficients. The data were also
examined in a multidimensional manner using
z-scores to determine whether each subject showed
any consistencies across sessions with respect to
which response system was maximally aroused.
This analysis led to identification of three groups
of subjects: those who responded primarily within
a single system across sessions regardless of stressor
(individual response stereotypy, 42%), those who
responded differentially across sessions to the two
stressors (stimulus–response specificity, 20%), and
those with profiles not readily classifiable (38%).
Results supported the notion that psychophysiological measures achieve some degree of meaningful reliability over time. We also argued that
identification of clinical patients who fit the stimulus–response specificity pattern may have great
clinical relevance. For example, a headache sufferer who responds to physical stressors with hand
surface temperature, but to mental stressors with
SEMG may require psychophysiological intervention targeting both response systems. Moreover, the clinician may need to investigate these
response patterns in terms of the stimuli most
readily eliciting them. This may explain why some
of our headache patients fail to respond to a singlemodality biofeedback intervention. Presumably, a
less complex therapeutic approach would suffice
for the more common stereotypical responder.
Others have also examined reliability of psychophysiological assessment from a individual
response stereotypy and stimulus–response speci-
8. Psychophysiological Assessment and Biofeedback Baselines
ficity perspective (Berman & Johnson, 1985; Foerester, 1985; Foerster, Schneider, & Walschburger,
1983; Robinson, Whitsett, & Kaplan, 1987) and
have come to similar conclusions. An excellent
review of temporal stability of psychophysiological response patterns can be found in an article by
Heinz, Huber, Schreinicke, and Seibt (2002).
Shaffer et al. (1991) studied the 1-week reliability of resting baseline psychophysiological activity
for several autonomic nervous system variables
assessed for 5 minutes after a 15-minute stabilization period. The 21 male and female undergraduates, ages 18–21, reclined with their legs supported
and their eyes open. The stability was high for skin
conductance level (r = .89), moderate and statistically significant for heart rate (r = .63), abdominal
amplitude (r = .63), finger temperature (r = .54),
and respiration rate (r = .49), and low and nonsignificant for blood volume pulse (r = .23).
As did most of the studies examining at the temporal stability of psychophysiological measures, in
our studies before 1990, we had employed the Pearson product–moment correlation coefficient as
our primary correlational measure of psychophysiological intersession reliability. Statisticians would
argue that the intraclass correlational coefficient
(Kirk, 1995) is a more appropriate reliability statistic when employing more than two test–retest
intervals. Intraclass correlations take into account
changes in values, not just the relative proportion of scores. This is especially useful in psychophysiological measures on which initial intensities commonly vary. More importantly, intraclass
correlations allow simultaneous incorporation of
more than one set of values on the same subjects.
Therefore, Arena and Hobbs (1995) reanalyzed
their 1989 study data using the intraclass correlation and found that with the exception of SEMG
during the physical stressor (cold pressor task),
the absolute values of the responses (forehead
EMG, hand surface temperature, heart rate) had
quite significant reliability (.70 or greater). They
concluded that statistical estimates of psychophysiological response reliability are functions of
the study design and particular reliability analysis
employed.
Gerin and his colleagues (1998), in a seminal
article, looked at the reliability of cardiovascular
responses (blood pressure and heart rate) and the
generalizability of these responses across various
settings. Twenty-four female college students (age
range 17– 26) were given a mental arithmetic task
(serial 13’s) following a 12-minute baseline twice
in the laboratory (to examine test–retest reli-
133
ability), once in a classroom and once at home.
Adequate test–retest reliability was found for the
baseline condition (.81 for systolic blood pressure,
.63 for diastolic blood pressure, and .68 for heart
rate). However, poor reliability was found, using
change scores from baseline to mental arithmetic task (absolute values were not given in the
report) for heart rate response (.09), while systolic
blood pressure (.68) and diastolic blood pressure
(.62) had adequate reliability. When examining
for generalizability, on all three measures, smaller
correlational coefficients were obtained for the
nonlaboratory settings compared to the laboratory
setting. The authors concluded that “this suggests
that even a minor variation in procedure, such as
a change in setting, can affect generalizability”
(p. 209). They further state that
if we are to find predictive power from the laboratory
to the natural environment, there is no dimension of
variability so trivial that it can be dismissed without
investigation. If simply changing the location of the
test site can reduce the lab-to-life associations, then
altering more significant aspects of the test situation,
such as the task or the subject’s motivation, is likely
to do even more damage to the stability of reactivity
as an individual difference. (p. 217)
(For a review of the topic of heart rate variability, we refer the reader to Appelhans and Luecken
[2006] and Wheat and Larkin [2010]).
A number of studies have examined the reliability of heart rate variability and have for the most
part found it to be reliable. To illustrate, Guijt,
Sluiter, and Frings-Dresen (2007) examined the
test–retest reliability of time-domain heart rate
variability and respiration rate measurements in
26 normal subjects over two sessions separated by
a week. Using a portable device, they took measurements during three conditions: lying down,
cycling, and sleeping. Both time-domain heart rate
variability (r's between .74 and .85) and respiration
(r's between .75 and .98) were found to be highly
reliable. Similarly, Carrasco, Gonzalez, Gaitan,
and Yanez (2003) examined heart rate variability by measuring 11 normal subjects for 5 minutes
three times a day over a 5-day period during a
number of conditions (lying on the back, standing, diaphragmatic breathing, exercise, recovery).
Most of the intraclass correlations were quite high
(r's ≥ .68). Others have found a more complicated
pattern. For example, Nussinovitch and colleagues
(2011) recorded heart rate variability from 70
healthy volunteers for 5 minutes (the standard
134
recording time). They then recalculated heart
rate variability based on randomly time-sampled
1-minute and 10-second periods. They found that
“good correlations between the 5-minute electrocardiograms (ECGs) and both the 1-minute and
10-second ECGs were noted for average RR (the
time elapsing between two consecutive R waves
in the electrocardiogram) interval, and root mean
square of successive differences in RR intervals.
No correlation was noted for standard deviation
of the RR interval and several other HRV parameters” (p. 117). Thus, heart rate variability is reliable depending on the time interval sampled and
the type of measurement analysis employed.
Veit, Brody, and Rau (1997) in an interesting
study, examined the stability of cardiovascular
measures (heart rate, systolic and diastolic blood
pressure) in response to a laboratory psychological stressor (mental arithmetic) in 75 adults over
a 4-year test–retest interval. They found adequate
reliability for both absolute and change score from
baseline for heart rate (.81 absolute, .76 change
score) and systolic blood pressure (.52 absolute,
.66 change score) measures. However, the absolute
value correlation for diastolic blood pressure was
.27, and the change score coefficient was only .16.
Researchers and clinicians have now begun
to look at the reliability of SEMG in nontypical
electrode sites, such as the back and jaw, as well as
to examine the reliability of psychophysiological
measures in clinical populations.
Castroflorio and colleagues (2005) measured in
nine healthy subjects the masseter and temporalis
anterior muscles daily for 3 consecutive days. During each session, subjects sustained for 30 seconds
three isometric contractions at 80% of maximal
force. Excellent reliability was found for the maximal force measures.
Recently, Auchincloss and McLean (2009) measured the reliability of pelvic floor muscle (PFM)
SEMG using two different vaginal probes on two
tasks: maximum voluntary contractions and a
coughing task. They found relatively little difference between the two probes and good betweentrial reliability coefficients (ranging from .58 to
.98). Between-days reliability, however, was quite
poor. They concluded that “although it is acceptable to use PFM surface sEMG as a biofeedback
tool for training purposes, it is not recommended
for use to make between-subject comparisons or
to use as an outcome measure between days when
evaluating PFM function” (p. 85).
Arena, Sherman, Bruno and Young (1990) made
bilateral SEMG recordings of paraspinal muscle
II. INSTRUMENTATION
tension in 29 subjects with lower back pain and 20
normal subjects in six different positions (standing, bending from the waist, rising, sitting with
back supported, sitting unsupported, prone) on
two occasions. Measures were highly reliable when
examined with analysis of variance procedures. As
with other research, statistically significant reliability coefficients were obtained when the absolute values of the measures were examined, and
when examined as relative [percent change from
baseline (prone) condition] values, differences
between the two groups were observed: The normal controls were statistically more reliable than
subjects with lower back pain during every condition. This study demonstrates how important it is
to examine reliability of SEMG measures in both
normals and clinical populations.
Lariviere, Arsenault, Gravel, Gagnon, and
Loisel (2002) measured bilateral SEMG from
four back sites in 40 subjects (half controls/half
patients with chronic low back pain) during various tasks, including a dynamometer task, fatigue,
and recovery, twice within 2 weeks. For the most
part, measures were reliable. They found that reliability was highest in the most fatigable muscle.
Callaghan, McCarthy, and Oldham (2009)
measured superficial quadriceps surface SEMG in
29 health control subjects, 74 subjects with patellofemoral pain syndrome, and 55 subjects with knee
osteoarthritis at 60% of maximum voluntary contraction over 3 days. They concluded that “poor
between-days reliability and high measurement
error suggests that surface sEMG should not be
adopted to assess fatigue during multi-joint, submaximal isometric quadriceps testing” (p. 172).
Netto and Burnett (2006) measured the reliability of maximal voluntary isometric contractions and submaximal (60%) isometric contractions bilaterally from eight sites in the C4–C5
level of the neck during a number of conditions
(flexion, extension, left/right lateral bending) in
five healthy male subjects, then retested 2 weeks
later. Both maximal and submaximal contractions had excellent within-day reliability, but only
maximal voluntary isometric contractions had
adequate between-day reliability.
Finally, a number of investigators have also
looked at the temporal stability of an ambulatory
monitoring device for surface SEMG levels. Arena
(2010) has noted that telemedicine and ambulatory monitoring is one of the primary future directions in SEMG research and clinical work. Arena
and colleagues (1994) had 26 healthy controls
wear a lightweight (24 ounce) device that mea-
8. Psychophysiological Assessment and Biofeedback Baselines
sured bilateral upper trapezius EMG, as well as
peak and integral motion, for 5 consecutive days
for up to 18 hours each day. Intraclass correlational
coefficients for the two SEMG variables across the
5 days were both significant, with alpha levels set
at .01. The two SEMG measures were highly correlated (r = .77); the two motion measures were
also highly correlated (r = .60). Reliability coefficients for the SEMG measures were similar to
those found in laboratory studies. They concluded
that the test–retest reliability of the ambulatory
monitoring device was within acceptable limits.
Airaksinen and Airaksinen (1998) also examined
the reliability of an SEMG device for ambulatory
recording and found it to be highly reliable.
135
function of demographic factors (e.g., age, gender, and race), clinical populations, laboratory
versus nonlaboratory settings, caffeine and
nicotine consumption, and psychological characteristics (anger, anxiety, depression, etc.).
8. For the most part, measures of heart rate variability have been found to have adequate test–
retest reliability.
9. We would urge every clinician to gather some
rudimentary baseline data on the reliability of
the measures routinely employed in clinical
practice. This provides some general indication
of equipment reliability, as well as the effect of a
particular clinical setting (type of room where
measures are routinely obtained, pictures, the
therapist variables, quietness, etc.).
Temporal Stability Conclusions
There are a number of conclusions that can be
drawn from the results of the studies reviewed earlier.
1. For forehead EMG, heart rate, blood pressure,
and hand surface temperature, the majority of
the studies indicate at least statistically significant reliability coefficients of modest magnitude.
2. The amount of variance, even in those measures with the correlations of greatest magnitude, suggests that other factors account for
more variance than does the experimental
manipulation (i.e., retesting).
3. Reliability is affected significantly by the statistical approach employed (i.e., analysis of
variance, absolute value Pearson product–
moment correlations, relative value [percent
change from baseline, raw change scores from
baseline], Pearson correlations, intraclass correlations, or analysis of response patterns, such
as individual response stereotypy or stimulus–
response specificity), with most studies suggesting that relative value correlations produce
lower size correlations than absolute value correlations.
4. It is probably prudent to use multiple measures
of arousal rather than rely on any one measure.
5. Hand surface temperature appears to be a very
complex response that may be affected by
repeated measurement.
6. More research is vitally needed, especially on
the temporal stability of measures such as electrodermal response, digital blood volume pulse,
respiration, and SEMG other than forehead.
7. More research is needed on reliability as a
Conditions Generally Employed
in Psychophysiological Assessments
Adaptation Period
The importance of an adaptation period in psychophysiological research has long been a topic
of discussion; unfortunately, there has been little
empirical research to date indicating the optimum
duration of an adequate adaptation period. An
“adaptation period” is defined as the time the subject spends in the experimental situation prior to
the onset of baseline measures or the experimental
conditions, or as Andrasik and Lords (2004) state,
“adaptation refers to a client becoming comfortable and returning to a normal level of functioning” (p. 223; emphasis in original). The function
of an adaptation period in psychophysiological
research and clinical work is threefold:
1. It allows the subject to become familiar with
the novel, experimental situation, as most
people are unaccustomed to having sensors
attached to various parts of their anatomy
while they sit with their eyes closed in a soundand light-attenuated room.
2. It allows presession effects to dissipate, such as
stress, rushing to the appointment, walking up
flights of stairs, and significant temperature
discrepancies between the outdoors and the
office.
3. It allows habituation of the orienting response
and permits the stabilization of psychophysiological responses. If these responses fluctuate
prior to the experimental manipulation or the
recording of tonic levels of physiological functioning, there is uncertainty whether the inde-
136
pendent variable (e.g., diagnosis, experimental
instructions, biofeedback training) led to the
findings, as opposed to random variations secondary to an insufficient period of stabilization. Thus, an adaptation period is especially
salient in early sessions of biofeedback training.
Likewise, an adaptation period is especially
important if advocacy of the law of initial values
yields a need to examine a patient’s physiological
responses using relative rather than absolute scores
(Wilder, 1950). If baseline or prestimulus levels are
unstable, the relative measures—generally raw
change scores pre- minus postmeasure or percent
change scores from baseline—may be drastically
influenced and results potentially vitiated.
There have been only a few studies investigating what constitutes an adequate adaptation
period; unfortunately, all were published before
1990. Meyers and Craighead (1978), in a rather
confusing study, found that respiration rate, finger pulse volume, heart rate, and basal skin resistance can reach stability in an average of about
5–6 minutes, although there was a great deal of
variability between subjects, with some needing
almost no adaptation period and others requiring a
lengthy one. Taub and School (1978), in an anecdotal study, found that some individuals required
as much as 30 minutes to stabilize on hand surface
temperature response. Frontal SEMG stabilization occurred in an average of 11 minutes across
a very small group of 17 undergraduate nonclinical students (Sallis & Lichstein, 1979). However,
among these subjects, there was “considerable
idiosyncrasy of the sEMG adaptation response”
(p. 339). This suggests there would probably be
much variability among clinical patients, especially for those with high levels of forehead muscle
tension. Lichstein, Sallis, Hill, and Young (1981)
reported a gender effect for heart rate response, in
which males were adapted from onset and females
required 13 minutes for an adequate adaptation
period. They also found that adaptation periods of
7 and 13 minutes were necessary for, respectively,
skin resistance level and frontal EMG.
Many factors merit consideration during planning and implementation of adaptation and baseline (see below) recordings (as well as for psychophysiological feedback). Even after making these
global changes, individual differences in responses
and patterns still exist, and clinical judgment must
enter into the picture. One size (or adaptation)
does not fit all.
II. INSTRUMENTATION
In practice, the clinician usually tailors the
cognitive preparation and adaptation time to the
patient and situation, although there is no standard. Practitioners need to be aware of the potential impact of instructional set on their patients
and should consider standardizing it as much as
possible and always documenting the instructions
they use. Other practitioners will appreciate a
clinical or research report that includes this information. It can ease replication and application to
clinical practice.
The total duration of the stabilization phase
varies and depends on several factors, including
time in the waiting room. Other practical factors
include the physical condition of the patient arriving at the office, the physiological activity monitored, the therapist's purpose, and the number of
prior sessions. Rashed, Leventhal, Madu, Reddy,
and Cardoso (1997) suggested that cardiovascular responses (heart rate, blood pressure, and
hand surface temperature) to cold pressor stress
are significantly attenuated by exercise. Kim et
al. (2006) note that this phenomenon does not
hold for anaerobic weight-bearing exercise. Thus,
an adaptation period is especially important when
incorporating such a task into a psychophysiological assessment.
Practitioners should monitor and record
moment-by-moment physiological functioning
to check the potential effects of brief orienting responses and should, for example, watch
for events such as abrupt noises. Brief orienting
responses are of potential clinical use as well. For
example, the person who is more physiologically
responsive to low-level environmental stimuli may
require different procedures. Habituation is rapid
after orienting responses but can affect summary
scores of adaptation and other periods.
Psychophysiological arousal can occur with perceived and bona fide threatening stimuli present.
Habituation of this type of response is usually slow
and variable (Sturgis & Grambling, 1988). Several
potentially threatening factors associated with
office visits can increase muscle tension and autonomic arousal. An example occurs in a person who
feels threatened by physicians or mental health
professionals. Sitting quietly for several minutes,
reclining, and being connected to the instruments
are other examples of stimuli that can be threatening for some persons. Practitioners should check
these factors when planning and interpreting
adaptation and other baseline periods. Repeated
exposure to potentially threatening stimuli within
8. Psychophysiological Assessment and Biofeedback Baselines
a longer initial session or repeated sessions might
be necessary to obtain adequate adaptation. Additional research is needed to develop more standard
guidelines for adaptation.
We suggest that one commonly needs at least
5 minutes for adaptation. However, sitting quietly
for up to 20 minutes may be necessary for some
patients. For example, Taub and School's (1978)
anecdotal report suggests that for some persons,
even 30 minutes is not enough for stabilization
of hand temperature. Shorter times are probably
enough for most people seen in clinical practice.
Adaptation often occurs between 3 and 5 minutes
with instruments attached, especially for patients
who have waited several minutes in a waiting room.
Neutral conversation is acceptable if the goals
are adjustment of body position and adjustment
to the instruments. Therapists should consider
omitting conversation if the goal is allowing the
physiological systems of interest to settle down.
During this phase, therapists typically give no
specific instructions to the patient except to sit
quietly and get comfortable. Arena and his colleagues (1983) usually tell their patients, “Sit quietly with your eyes closed for the next couple of
minutes” (p. 450). Other therapists may wish to
have patients sit with their eyes open (see discussion in baseline section, below). The basic goal is
to get the patient to sit quietly and get used to the
clinical or experimental situation.
Another perfectly acceptable criterion is a floating adaptation period, which has no prescribed
length. Rather, the clinician or investigator has a
preset criterion for stabilization of each response
(e.g., heart rate must remain plus or minus 3 beats/
minute for a full minute, or SEMG response cannot fluctuate by more than 5% for a full minute),
advancing to another condition once that criterion is met. This saves time with those individuals
who are already stable, is tailored to the patient’s
physiological responding, and ensures that all
patients achieve stabilization. However, disadvantages also exist: The therapist must focus deeply on
a patient’s responding and risk the patient needing,
as Taub and School (1978) noted, 30 minutes to
stabilize. Fortunately, the latter problem is solved
with a modified floating adaptation period (i.e.,
patients meet the floating criterion or 10-minute
time limit, whichever comes first). Regardless of
the adaptation period used, we urge clinicians to
employ a specific strategy consistently. In psychophysiological assessment and treatment, as in most
things, consistency is half the battle.
137
Psychophysiological Baselines
Although the actual instructions during a baseline period are usually identical to those provided
for adaptation, the two conditions serve different
functions. A “baseline period” is defined as the
period following adaptation, in which psychophysiological response measures have stabilized
(prior to the onset of any experimental or clinical
manipulation; e.g., a stressor condition or biofeedback). The purpose of this condition is to observe
and measure resting basal physiological activity. We believe that this condition is essential,
because, nearly always, the practitioner compares
the baseline or resting condition values to the
experimental or treatment conditions. As noted
earlier, relative values, most generally raw change
scores from baseline or percent change from baseline, are dependent on a baseline condition. A
baseline is essential in cardiovascular research, in
which nearly all measures of cardiovascular reactivity use relative scores.
Some (Sella, 2005) have argued that for SEMG
research, the term “baseline” should not be used;
rather, “resting tonus” should be used in its place.
We (Arena, 2005) disagree with this approach
based on the fact that “baseline” is a well-established term that has been used for over a century in both behavioral and psychophysiological
research/clinical practice, as well as the fact that
the concept of baseline refers to a measurement,
whereas “resting tonus” is the phenomenon that
one is measuring.
Whether a patient’s eyes are open or closed
during baseline is an area of disagreement among
clinicians and researchers. There are no available data on this subject, so personal preference
determines choice. Here, the two authors of this
chapter differ slightly. Arena conducts baselines
with the patient’s eyes closed during nearly all psychophysiological assessments, and most biofeedback training sessions. He reserves an eyes-open
baseline period for biofeedback when a patient
expresses a preference for visual feedback or generalization training (usually, after the patient is
sufficiently skilled in producing the biofeedback
response, instruction advances to reproduce more
challenging “real world” factors). Schwartz (1995)
believes that eyes-closed baselines are suitable for
conditions such as insomnia and are less realistic
for assessing baseline physiology for headaches and
other symptoms that occur with eyes open. For
example, he asserts that using only an eyes-closed
138
baseline can lead a therapist to conclude incorrectly that there is no excess muscle activity. Commonly, more muscle activity in the head and facial
muscles exists when the patient’s eyes are open.
Schwartz (1995, p. 152) stated that
Therapists should get baseline data with eyes open
when patients' symptoms start with their eyes open
and when biofeedback with eyes open is planned.
Observing lower arousal with eyes open than with
eyes closed is potentially useful. It provides cues
about what it means for patients to close their eyes.
It raises questions to answer about what patients are
thinking about and doing when they close their eyes.
When eyes are kept open, therapists should consider
instructing patients to include time calmly gazing
at an object such as a picture or plane. They should
remind patients to avoid staring or examining the
object they are looking at as well.
Regardless of which baseline recording strategy
one chooses, both Arena and Schwartz urge careful consideration and consistent application during clinical/research practice. This ensures a large
database, and the larger the database is (and we
would argue that clinical experience creates a very
strong database, indeed), the more sure one can be
about one’s observations and conclusions.
As noted earlier, there are few data about what
constitutes an adequate baseline, and much of the
research confuses baseline with adaptation periods. For example, Hastrup (1986) reviewed the
methodology and duration of baseline conditions
in an exhaustive review of cardiovascular reactivity studies and found nearly no agreement among
the studies in terms of methodology and duration of the baseline periods. She recommended
that the baseline period be at least 15 minutes in
duration for cardiovascular reactivity experiments
to ensure the lowest possible baseline recordings.
Jennings, Kamarck, Stewart, Eddy, and Johnson
(1992) found similar results, with five of 24 studies
reviewed having baseline periods of 10 minutes or
more, five having between 6 and 10 minutes, and
the remaining 24 with less than 5 minutes’ duration. Gerin, Pieper, and Pickering (1994) found
that a 1-minute baseline was sufficient for stabilization of heart rate and systolic and diastolic blood
pressure.
For those interested in conducting methodologically complex psychophysiological research,
Jamieson (1999) wrote an excellent article about
baseline differences in psychophysiological recording, indicating that (1) relative values such as
change scores are confounded with baseline
II. INSTRUMENTATION
whenever data are skewed, and (2) when baseline
differences are real, analysis of covariance has a
directional bias that magnifies differences in one
direction and minimizes those in the other direction. Jamieson provides suggestions for identifying and correcting these problems. Finally, Piferi,
Kline, Younger, and Lawler (2000), arguing that
simply resting quietly does not ensure equivalency
between individuals, present data suggesting that
showing a relaxing video of the sea achieves a
greater degree of relaxation and a more accurate
recording of baseline measures than does the
traditional baseline condition, at least for measures of cardiovascular reactivity. We would posit,
however, that the baseline condition should not
“obtain the lowest possible resting rates along
the same point on the continuum of excitement”
(p. 215), and Piferi and her colleagues are actually
creating a relaxation condition (see below).
Clinical Considerations
in Psychophysiological Baselines
For resting baselines and for office-based stressors,
clinical practitioners need to be very cautious when
interpreting physiological data. Comparisons of
resting baselines across sessions are complex for
many patients. Knowledge of this fact is important
when clinicians generalize to other situations and
compare data across sessions. Many practitioners
view each session's resting baselines as largely new
situations, at least for most autonomic-mediated
variables.
Marked shifts in muscle and autonomic activity
often occur after the patient sits quietly for several
minutes. Muscle activity can steadily or suddenly
drop. Finger temperature can gradually or suddenly
increase. Heart rate can plummet. Therefore, baseline periods should be considered up to 15 minutes
in a very early biofeedback session and in some
therapy sessions that can capture these changes. It
also helps to check for physiological changes that
occur in the extended relaxation periods outside
the therapist's office. There is no fixed or proper
time for all people and all circumstances.
Realizing that these changes can occur before
therapy can help to increase a patient's confidence. It is also important to document the lack
of change, especially when changes begin to occur
later during feedback and nonfeedback phases.
This information is useful to the therapist, because
such changes in an initial session do not mean
that one needs therapy less. Among the important
therapeutic goals are shortening the time before
8. Psychophysiological Assessment and Biofeedback Baselines
the therapeutic changes occur and increasing the
degree and replicability of such changes.
Shorter baseline phases, such as 1 to 3 minutes,
also are feasible and proper under some conditions.
For example, if we consider the patient sitting quietly with eyes closed, his or her muscle activity may
remain low and steady, with very little variability
for about 1–2 minutes among multiple muscles in
the head and neck. The muscle activity will probably not change significantly over the next few
minutes. Cost containment and other pragmatic
factors (time constraints, patient’s becoming restless, etc.) argue for the shortest baseline phases
that can typically answer evaluative and therapy
questions.
There are circumstances in which the clinician
wants or needs additional baseline data. Obtaining such data can entail more than one baseline
session and extending some baseline phases even
beyond 15 minutes. For example, when there
is much variability within or between sessions,
one can justify longer resting baselines. Another
example occurs for disorders such as sleep-onset
insomnia, when the relaxation sessions at home
are long.
Some patients show increasing roused activity,
cooling hands, increasing pulse, and/or restlessness during the first few minutes of a baseline.
More than several minutes of a baseline might be
unnecessary and counterproductive. Even about
5 minutes may be enough. In such a case, the
therapist should consider that relaxation-induced
anxiety (RIA) may be present (for a detailed discussion of this topic, see Arena & Blanchard,
1996; Schwartz, Schwartz, & Monastra, 2003; or
Wegner, Broome, & Blumberg, 1997). When this
occurs, the clinician should consider the therapeutic goal to be gradually increasing periods of sitting
quietly without increases in physiological arousal.
Pretreatment and periodic physiological baselines are less practical under some circumstances.
For example, there are limitations in the schedules
of some patients, as in the case of the patient who
lives a few hundred miles from the therapist and
there is no qualified professional to whom one can
refer the patient. The therapist is consulting only
for one session that focuses on the intake interview
and patient education for treatments thought to
help reduce physical symptoms. There is time for a
brief biofeedback session, but not enough time for
a desired baseline. The therapist decides instead
to get only a brief baseline of about 5 minutes
and invest the remaining instrumentation time
to providing feedback. The therapist instructs the
139
patient in relaxation techniques and provides education booklets and audiotapes. The patient then
goes home and practices as instructed.
The lack of physiological baseline data does not
always compromise therapy. One properly adjusts
priorities, maintains the patient's best interests,
and can initiate therapy. If this patient returned
for further therapy, one could still get a physiological baseline. Practitioners can discuss the ideal
with the patient and note in their reports the reasons for proceeding differently.
Conversely, there are conditions for which one
can justify multiple physiological baseline segments or sessions. Such a situation occurs when
one suspects that the patient has symptoms that
fluctuate in intensity at different times. Examples
of such times are soon after specific eliciting or
emitting events (e.g., eating, upsetting discussions,
physical activity, and certain times of day). Therapists should consider scheduling office sessions to
coincide with or immediately follow such events.
The absence of excess tension or arousal during
a resting baseline does not mean the person has
adequate control. This also is true for patients who
show a lack of significant reactivity to a stressor.
Therapists should always consider such factors as
the possible effects of medications, the office environment, baseline conditions, and the limitations
of simulated stressors. For example, some people
do not react to office stressors. There are individual differences in how therapists present stressors.
Presentation style can affect the stressor's effect.
Physiological tension and arousal, reactivity,
and slow recoveries in one office session often do
suggest similar functioning in daily life. That is,
one can generalize from the psychophysiological
assessment setting to real-life settings. However,
office sessions do have limits. Tension, physiological reactivity, and slow recovery in one session do
not mean that the person reacts the same way in
other situations. At best, these are snapshots or
glimpses of a person's psychophysiological activity in daily life. Office procedures are sources of
hypotheses and information for productive discussions. However, they are not always reliable evidence for a patient's daily functioning.
Conditions Involving Assessment
of Self‑Control Abilities
Relax Deeply Condition
Because many psychophysiological interventions
are believed to work through the final common
140
pathway of relaxation, obtaining some measure
of patients’ ability to relax on their own prior to
any treatment and periodically during treatment
is often useful. The instructions that Arena and
colleagues have used for this condition (Andrasik, Blanchard, Arena, Saunders, & Barron,
1982; Arena, Blanchard, Andrasik, Applebaum,
& Myers, 1985) are simply, “Please try to relax as
deeply as you possibly can.” Schwartz (1995) uses
the following instructions:
“Now, rest quietly a little longer. Use whatever methods you think best to relax. Focus on the muscles of
your face, head, and shoulders. Let yourself go and
release the tension in different parts of your body. If
you feel that you have to move, scratch, sneeze, or
something else, go ahead and do it. This phase lasts
a few minutes. Don’t think of problems or upsetting
events and do not worry about how well you are
doing. Whatever degree of relaxation achieved is all
right.” (p. 154)
Warm Hands Condition
This condition is directly relevant to thermal biofeedback training, which involves teaching hand
warming through mental means. Obtaining some
measure of a patient’s ability to increase hand temperature prior to treatment is useful, particularly
when assessing whether the skills have developed
or been learned. There are many ways to test or
prepare the patient for psychophysiological learning. The most common, by far, is a “self-control”
condition that is interspersed between a baseline
and a feedback segments. During the self-control
period, the patient is asked to control the desired
psychophysiological response (in this instance,
hand temperature: “Please try to warm your hands
through purely mental means”) without any feedback. If the patient can control the response, the
clinician may infer that between-session learning has occurred. Such a phase can be routinely
added after the second or third biofeedback session. Sometimes this condition is presented after
the biofeedback portion. If a patient can control
the response, then the practitioner may infer that
within-session learning has occurred.
“Generalization” involves preparing the patient
to, or determining whether or not the patient can,
apply the learning that may have occurred during
the biofeedback session to the “real world.” One
can partially or tentatively infer this from the
previous procedures and also obtain temperature
measurements in other situations.
II. INSTRUMENTATION
Relax Muscles Condition
This condition is directly relevant to SEMG biofeedback, which involves teaching reduction of
muscle activity. Obtaining some measure of a
patient’s ability to decrease muscle activity prior to
treatment is useful. The instructions are adapted
to the muscle group recorded (e.g., “Please try to
relax the muscles of your forehead for the next
couple of minutes”).
Personally Meaningful Positive
Imagery Condition
This condition is often included as a control condition that is compared to the negative imagery
condition. This is frequently useful, because
individuals’ abilities to imagine vary greatly, and
without this phase, a practitioner may mistakenly
believe that a patient has no reactivity to a negative imagery condition when, in reality, he or she
is merely poor at imaging. Consider asking the
patient for a vividness score of the scene after the
condition is over. Prior to the assessment, consider
asking him or her to describe a very pleasant scene
experienced previously, with descriptors of images
targeting a majority of the five senses. Also, consider requesting that the scene rate a 9 or 10 on a
1- to 10-point scale of pleasantness. Information
is recorded and modified to refine the scene until
consensus is reached. During the assessment, the
instructions used are as follows: “I’d like you to try
to imagine, to picture in your mind’s eye.” The
pleasant scene is then read. In addition to variable
imaging skills, another problem is the absence of
control for the experimenter effect. That is, some
therapists or experimenters are rather low key and
likely to read the description of the image in a
monotone, while others will read it with dramatic
flair. Thus, some researchers and practitioners provide more standardized negative imagery tasks that
use tape-recorded instructions/imagery. However,
standardization, or a one-size-fits-all approach is
imperfect, as we indicate below.
Stressor Reactivity or the Stimulation Phase
It is often useful to introduce cognitive and physical stressors to check for psychophysiological reactivity and rate of recovery. Both reactivity effects
and recovery can help to identify causes and correlates of biobehavioral disorders and can potentially help to predict those persons at risk for these
disorders (Haynes, Gannon, Orimoto, O’Brien, &
8. Psychophysiological Assessment and Biofeedback Baselines
Brandt, 1991). These authors also note that assessing reactivity can help develop effective interventions. Lovallo (2005) presents a complex theory
for how psychophysiological reactivity can lead
to cardiovascular disease and posits a number of
treatment implications. This section covers reactivity, while the next section focuses on recovery.
Although many providers use stress stimuli
routinely before starting therapy, some do not.
However, using stress stimuli may help therapists
answer some patients' questions about the ways
their physiology reacts to stimuli, while other
patients benefit from seeing evidence of their reactivity and recovery.
For some patients there is very little, if any,
excess tension or arousal while they are sitting
quietly or relaxing. The therapist may suspect that
this is not typical for a specific patient. Stimulation allows one to examine the patient during and
following a stressor presentation, and compare and
contrast him or her to normal individuals and to
those with the same biobehavioral problems.
There are two interdependent pathways in
which physiological stress occurs (Haynes et
al., 1991). One is the autonomic nervous system
(ANS) pathway, especially the sympathetic nervous system (SNS) division. The other is the
hypothalamic–pituitary–adrenocortical
system
pathway. The hypothalamus organizes the ANS
pathway with input from cortical and subcortical
brain structures. This tends to have a rapid onset
and short equilibrium time that is the duration of
maximum effect. The effects are mostly the results
of nerve endings releasing epinephrine and norepinephrine and from the adrenal medulla. In
the second pathway, the hypothalamus also regulates the release of adrenocorticotropic hormone
(ACTH) from the pituitary gland. This promotes
the release of cortisol from the adrenal cortex.
These effects are slower and have a longer equilibrium latency or time until maximum effects. The
time is longer than that resulting from epinephrine and norepinephrine.
Thus, the duration of a stressor strongly influences its impact. Many studies show that shortduration stressors elevate neurotransmitters, but
longer-duration stressors suppress them. Longerduration stressors deplete norepinephrine, lift the
inhibition of ACTH, release cortisol, and suppress
the immune system. Transient stressors often used
in laboratory studies and clinical practice may not
be sufficient for health-inhibiting effects.
The nature of the stress—physiological or psychological—is important. For measuring primarily
141
psychological stressors, therapists should be aware
that serum cortisol responds more to subjectively
distressing, uncontrollable, and psychologically
prominent stress (Dienstbier, 1989). In contrast,
the ANS-mediated catecholamine responses, such
as epinephrine and norepinephrine, respond to
nearly all stimuli such as startle, cognitive, exercise, and mild electric shock (Haynes et al., 1991).
See Asterita (1985), Haynes, Falkin, and SextonRadek (1989) and Kronenberg, Melmed, Polonsky,
and Larsen (2008) for more detailed discussions of
this topic.
Stress and stimulation constitute part of evaluations and treatments for conditions other than
those treated with relaxation. For example, when
evaluating patients with fecal incontinence, therapists should use simulated stimulation to the lower
bowel. This checks for reactivity of the internal
and external anal sphincters. It also checks for
ineffective tensing of the gluteal and abdominal
muscles.
For patients with urinary incontinence, therapists sometimes introduce fluid into the bladder
to check for sphincter control. During this procedure, practitioners also check for ineffective tensing of abdominal muscles. Other examples involve
patients undergoing muscle reeducation. Therapists often ask patients to hold, carry, walk, push,
or bend to evaluate their muscle actity.
Response magnitude such as peak reactivity
during a stressor is a commonly used psychophysiological response parameter for assessing the effects
of stress (Haynes et al., 1991). Researchers and
practitioners use cognitive and physical stressors to
examine ANS- or central nervous system (CNS)mediated reactivity. A variety of stimuli are in use
in clinical practice. Practitioners assume that the
stimuli are stressful. However, for some individuals, this may not be the case. In some cases, it is
merely orienting or mild stimulation. The following are a few stimuli in clinical and research use
and abbreviated sample instructions. For each, if
instructions are not included with the description of the stressor, the therapist should assume an
introductory phrase (e.g., “In a few moments I will
ask you to . . . ” or “When I ask you to, please . . . ”).
1. Mental arithmetic. “When I tell you to, please
start at (an arbitrarily chosen large number; e.g.,
986) and count backward by 7’s (or 8’s, 9’s, or 13’s)
keeping your eyes closed.” Alternatively, “Please
read silently (or aloud) each math problem and
write down (or call out) the answer” (e.g., 121 +
767 = ?; 326 – 74 = ?; 18 × 12 = ?; Linden, 1991).
142
This is probably the most commonly used office
stressor in research and clinical practice. Advantages (Linden, 1991) of a mental math stressor
are ease of administration and lack of equipment
requirements (unlike video games, reaction times,
or cold pressor tasks). At most, only a method to
visually present equations is necessary. No ethical concerns should arise from an Institutional
Review Board (IRB), Human Subjects Research
Review Committee, Ethics Committee, or Clinical Practices Committee. In addition, the technique of mental arithmetic offers a wide range of
variations for adaptations to specific patients and
for repeated presentations. However, a potential
problem for comparing studies and procedures
stems from the lack of universally accepted standardized procedures, as indicated by Strike and
Steptoe (2003): “It should be noted that there are
no standard agreed protocols for mental stress testing. It cannot be assumed that mental arithmetic,
for example, is the same challenge in different
studies. Few studies have collected subjective ratings, behavioural performance measures, or other
indicators of stressfulness” (p. 697).
Linden (1991) provides a useful review and
a series of studies of the effects of vocal versus
written versions, noise distraction, and different
types of math tasks. The most arousing, at least
for cardiovascular reactivity, are those involving
vocal responding, noise distraction, and solving
visually presented equations (Linden, 1991). One
may expect some attenuation of the reactivity
with repeated presentations of the same math task
(Sharpley, 1993). Thus, the therapist should consider different math tasks if he or she uses repeated
stressor presentations involving math.
2. Tense muscles. “Make a fist.” “Clench your
teeth.” “Try to open this tightly closed jar.” “Shrug
your shoulders.” “Bend slightly at the waist.” “Hold
this package with both hands.”
3. Personally meaningful negative imagery. Similar to personally meaningful positive imagery (discussed earlier) except the patient is asked to imagine something very unpleasant.
4. Memory tasks. “Remember this story exactly
as I say it.”
5. Hyperventilation. “Inhale and exhale very
quickly and deeply for 2 minutes.” Or, “Inhale and
exhale through both your mouth and nose. Each
time you inhale try to fill your lungs completely.
Each time you exhale try to empty your lungs completely.” Or, “Inhale every time I say ‘In.’ Exhale
II. INSTRUMENTATION
every time I say ‘Out.’ ” (One can use an audiotape
to signal inhalations and exhalations. See Gevirtz
and Schwartz, 2003, for discussion of hyperventilation, Hyperventilation Provocation Test, and cautions.)
6. Prerecorded loud noises or other unpleasant
sounds. Examples are a baby crying, car horns
repeatedly blowing, or listening to people screaming at each other. For combat-related posttraumatic stress disorder, often sounds of war, such as
helicopters and machine gun fire, are used.
7. Cold exposure and cold pressure. “When I say
‘Start,’ I’d like you to place your right hand up to
your wrist in the bucket of ice water. Then close
your eyes. Please keep your hand in the ice water
until it hurts so badly (or becomes too uncomfortable) that you want to remove it or until I tell you
to remove it. Any questions? OK. Start.” For excellent reviews of the methodology of the cold pressor test, certainly the most widely used physical
stressor in psychophysiology today, as well as the
physiology involved, see Mitchell, McDonald, and
Brodie (2004) and Velasco, Gomez, Blanco, and
Rodriguez (1997).
8. Action and challenging video games.
9. Slides of stressful scenes or videotaped trauma.
“Look at these slides (or video).”
10. Difficult quizzes. “Complete this quiz. Most
people get a score of at least .”
11. Your Everyday Life Pressures and Holmes–
Rahe Visualizations. Rosenthal et al. (1989) developed two brief and practical stressor tasks for
research that are of potential clinical use. The
first involves a number of stressful vignettes based
on the Holmes and Rahe (1967) Social Readjustment Scale, which ranks a number of stressful life
events in terms of severity. Instructions are as follows: “Please close your eyes. Visualize yourself in
the following situation(s). Try to see yourself in
the situation. Feel just how this situation hits you.
Really get into it! Try to make it as real and vivid
as you can—include the sights, sounds, smells and
emotions. Imagine how you would react as clearly
as possible.”
The second stressor task, which involved generally stressors of lesser intensity, was termed a YELP
(Your Everyday Life Pressures) task. The YELP
task involved eight selected vignettes “depicting
frustrating, disappointing, or otherwise noxious”
situations. Many doctoral-level clinicians selected
these eight YELP tasks from among 48 potential
8. Psychophysiological Assessment and Biofeedback Baselines
items. Interested readers should refer to Rosenthal et al. (1989) or Schwartz (1995) for additional information regarding the YELP task items,
including sample vignettes.
12. Ischemic (blood pressure tourniquet) pain
(e.g., Johnson & Tabasam, 2003; Pinerua-Shuhaibar et al., 1999).
13. Exercise step-up test (e.g., Feinstein et al.,
1999; Lim, Shields, Anderson, & McDonald,
1999).
14. Stroop Color Test. One of the oldest reaction time tasks in psychology (MacLeod, 1991),
the Stroop Color Test has been in existence since
1929. In this task, patients are asked to read the
name of a color that is presented to them. If the
color of the ink matches the name of the color, the
naming takes less time and people are more accurate than if the name of the color does not match
the color of the ink. The therapist or experimenter
can make the task even more stressful by having
people return to the beginning of the task if they
make a mistake or by stating, “Most people are
faster than you on this task,” and so forth.
Obviously, one does not use all or most of
these techniques with each patient. Research and
clinical practice usually include from one to three
stressors. Most require at least 1 minute and usually up to 4 minutes for each presentation. There
are individual differences in reactivity, hence the
rationale for using multiple stimuli of different
types.
Instructions probably have an arousal effect for
at least some physiological responses, such as heart
rate (Furedy, 1987; Sharpley, 1993). This probably results from several factors, such as attending
to the instructions and anxiety associated with
the uncertainty and challenge of the task. This
additional arousal effect can confound the assessment of reactivity of the stressor. Therefore, the
therapist should consider measuring the reactivity
during the instructions and separating this from
the stressor task data. With computer-based psychophysiological systems one can create periods or
trials designated as instructions.
Some practitioners insert another period of
about 1 minute after the instructions and before
instructing the patient to start the task. They
assume that this allows the patient to relax and
allows for measuring the effects of instructions
and anticipation. However, some patients may
prematurely start some cognitive task during this
143
period. Therapists can circumvent this easily by
instructions such as “In a few moments I will ask
you to. . . . Keep relaxing until I say to begin.” For
counting backwards, therapists should wait until
after the interspersed postinstruction period to
give the numbers.
One vitally important area of research that has
been nearly overlooked is which stressor is better
for which type of patient under what condition.
Yoshida and colleagues (1999) compared the ability of cold pressor, hyperventilation, mental arithmetic, and exercise step-up stressors to induce
an angina attack in 29 patients with vasospastic
angina pectoris. They found that the hyperventilation task was least effective (13%), with cold
pressor and mental arithmetic equally effective (27
and 28%, respectively) and the step-up test most
effective (55%). Similar research is needed for
other types of disorders, such as headache, panic,
Raynaud’s disease, and so forth.
Poststress Adaptation Periods
Psychophysiological and other biobehavioral disorders often have important physiological components. Implicated as causal factors are environmental and other stressors. However, the interactions
among behavioral, cognitive, and physiological
factors are complex.
The strength of the relationship between the
magnitude of reactivity and other indices of psychological functioning is often modest (Haynes
et al., 1991). Haynes et al. also observed a modest
ability for reactivity to distinguish between persons with a disorder and those without it. Hence,
there is interest in both the rate and degree of
recovery to help explain etiology and plan clinical
interventions.
The goal of many clinical interventions is
changing the psychophysiological response to
stress (Cacioppo, Berntson, & Anderson, 1991;
Haynes et al., 1989; Schwartz & Andrasik, 2003).
Therefore, psychophysiological poststress recovery
is of crucial importance.
Implications include etiology and treatment of
psychophysiological disorders. Specifically, recovery indices may help identify causal mechanisms
of many biobehavioral disorders. They may help
identify persons at risk for these biobehavioral
disorders and help in the development of effective
interventions and evaluations of treatment. For
example, if a patient reacts in a normal fashion
during a variety of stressors but takes much longer
than normal to return to baseline levels following
144
these stressors, two reasonable inferences can be
made. One is that this individual has a physiological impairment or abnormality that causes a return
to baseline slower than that of normal individuals.
Two, he or she needs to develop psychological and
psychophysiological coping strategies immediately
following a stressor to reduce physiological arousal
as quickly as possible and return to a basal quiescent state.
Within-study differences between stressor and
poststress recovery results constitute a very important index of the importance and potential use
of poststress recovery. This implies that these
two indices stem from different mechanisms. Of
the 180 statistical analyses reported by Haynes
et. al. (1991), 81 showed nonsignificant effects of
stressors. Of these 81 individuals, 74% showed
significant recovery phase effects. Conversely,
when stressor effects were significant in 74 analyses, recovery phases showed nonsignificant effects
for 42% of the same variables. Stress effects and
recovery very often differ in terms of sensitivity
and potential utility.
Impaired recovery or slowness of recovery after
psychophysiological reactivity to one or more
stressors is the focus of using poststress recovery
stages. Nearly all theories of psychophysiological
disorders include an impaired recovery process as
one of their central tenets or in their definition
of psychophysiologic abnormality. Specific definitions of poststress recovery vary in the literature.
Arena et al. (1989a) defined “recovery” as a return
to the quiescent baseline state after stress-induced
reactivity. The SEMG recovery was a return to 5%
of the mean of initial resting baseline. Hand temperature recovery was a return to within 5% of the
baseline mean. Heart-rate recovery was recovery
to within 2 beats/minute of the baseline.
Other definitions of recovery include “changes
in stressor-induced responses following stressor
termination” and “the rate and degree to which a
psychophysiological response approaches pre-stress
levels following a stressful experience” (Haynes et
al., 1991, p. 356). These definitions allow for nonlinear and bidirectional changes. It is different
from a return to a prestressor quiescent baseline
state. The time course of recovery is the magnitude of the response over time after stopping a
stressor. It is sometimes nonlinear and may diverge
from prestressor levels. For example, arousal sometimes increases or becomes unstable.
Very few studies specifically address the optimal time period for assessing poststress recovery
(Arena, 1984; Arena, Bruno, Brucks, & Hobbs,
II. INSTRUMENTATION
1992). The physiological variables in these two
studies included cephalic vasomotor response,
frontal and forearm flexor EMG, hand temperature (left hand), heart rate, and skin resistance.
Arena (1984) studied 15 college undergraduates
(about age 20) and reported that a 3-minute poststress period was adequate to return to a basal quiescent state for most of several psychophysiological measures. Frontal EMG, however, needed more
than 3 minutes to recover. This study “indicated
good intrasession reliability on all measures except
frontal EMG, where there was inadequate intersession reliability” (Arena, 1984, p. 247).
A follow-up study (Arena et al., 1992) did not
completely replicate or support the major findings of the earlier study. This study examined
heart rate, hand surface temperature, and frontal
SEMG following a cognitive (serial 7’s) and physical (cold pressor task) stressor, and had 6-minute
poststress adaptation periods. Most subjects (about
78%) returned to baseline within 6 minutes for
heart rate. Average times were 3.7 and 2.9 minutes for the two poststress periods. However, for
forehead EMG, only 48% returned to baseline in
the 6 minutes. The average time was 4.6 minutes
and 5.1 minutes, respectively, for the two poststress
periods.
For hand surface temperature, 6 minutes
“was clearly inadequate” for most subjects. Only
about 38% returned to baseline in this period.
The average time was about 4.5 minutes and 5.6
minutes, respectively, during the two poststress
periods. The percentages returning to baseline
appeared higher in the first poststress period. For
example, hand temperatures returned to baseline
in nearly 48% of subjects after the first cognitive
stressor (mental arithmetic), compared to only
about 26% in the second poststress period (cold
pressor task).
Although the earlier study indicated that about
3 minutes are required for cardiovascular and
heart rate modalities to return to a prestress basal
level, the latter study indicated that much longer
times are necessary. Arena et al.’s (1992) speculation about the differences focuses on the larger
sample size and the wider age range of the second
study. The age range was 17–75, and the mean age
was nearly 33.
These two studies used serial 7’s from a large,
random, three-digit number as the cognitive
stressor and cold pressor (to the right hand) as the
physical stressor. Exposure was 4 minutes for the
cognitive stressor and up to that for the physical
stressor. This cognitive stressor is probably milder
8. Psychophysiological Assessment and Biofeedback Baselines
than that used by many others, and the cold pressor is a more intense physical stress than most
clinicians use (and likely to evoke cardiovascular
effects that prolong recovery vs. most other officebased physical stressors). The subjects were college
students and community volunteers of various
ages; hence, caution should be used when generalizing to patients. However, a reasonable assumption is that implications from the more recent
study are more applicable to patients.
Except for the limitations described earlier
concerning these studies, we believe this type of
research is important and useful. Findings support the need for recovery periods. They indicate differences in the durations of these periods
among modalities and document the duration of
the recovery periods under specified conditions.
Poststress periods should be at least 6 minutes if
patients must return to a baseline. This is feasible in clinical practice, using two or three stress
periods during an evaluation session. However,
it is typically impractical during routine therapy
sessions if multiple stressors or intense stressors
are used. The cognitive stress in these studies is
only serial 7’s for 4 minutes. This is not universally
stressful for everyone. As we have noted concerning laboratory and office-based stressors,
there are many other inferences that applied psychophysiology researchers make. One is that laboratory
stressful conditions are comparable to stressors found
in the everyday world. Having both a) placed my
hand in a bucket of ice water up to my wrist and kept
it there until I couldn’t stand it any longer, and b)
been in an airplane for three hours with screaming
children in the seats directly in front and in back of
me, I can tell you that equivalence of laboratory and
“real world” stressors is a very dubious proposition,
indeed. (Arena, 2000, p. 22)
Clinical Vignettes
of Psychophysiological Assessments
We present a number of situations in which the
biofeedback therapist might use psychophysiological assessments to answer clinical questions. We
highlight the pain literature (headache and lower
back pain), because of our expertise. First, though,
we present some caveats.
Arena (2000) purported that “much of the
research and clinical pain work that utilize psychophysiological assessments or ‘stress profiling’—
including those that employ surface sEMG measures—are based on inferences which have not
145
been empirically tested and nearly all psychophysiological assessments have not been empirically
demonstrated to have any clinical utility” (p. 21).
He further stated:
The biggest inference that clinicians routinely proceed upon is that conducting a psychophysiological
assessment or a “stress profile” will give them important information in helping to determine how to proceed in treatment. . . . This assumption has never
been empirically tested. That is, no study has shown
that if you have a pain patient who demonstrates one
particular sEMG abnormality, compared to a different sEMG abnormality or no abnormality, that that
person does better in one type of treatment compared to a different type of treatment. Such research
is vitally important and must be conducted if our
field is to continue to grow and flourish. (pp. 22–23)
Such research is especially important when
standard treatments are expected to produce very
high rates of success, as is the case with psychophysiological treatments for headache (Andrasik, 2010; Arena & Blanchard, 2005). When we
achieve such high rates of success by providing
everyone the same standard treatment, the psychophysiological assessment must add significant
predictive value, or the justification for its use is
lacking.
Simply because such research has not been
conducted, though, does not mean that clinicians
should stop conducting psychophysiological assessments. Clinicians are often making inferences and
going where “no literature has gone before.” The
inferences that we and other clinicians make are
our best judgments, based on clinical experiences,
assessment literature, and the indication that the
presence of certain abnormalities or findings warrants a specific treatment direction. We should be
humble and understand the limits of our interpretations. A little tentativeness goes a long way.
With that very important caveat discussed, we
proceed with some clinical examples.
Straightforward SEMG Assessment of a Tension
Headache Patient
A therapist wants to ascertain in the intake
whether Mrs. Smith, a tension headache patient,
would likely benefit from biofeedback, and, if so,
would feedback from one muscle site be more
likely to achieve treatment success than feedback
from another muscle site. Her reported tension
and pain are primarily in the forehead and in the
upper back and/or posterior neck. In a psychophys-
146
iological assessment, the therapist measures forehead SEMG and bilateral upper trapezius SEMG
levels during two baseline conditions, a 5-minutes
eyes-open and a 5-minute eyes-closed condition,
preceded by a 10- to 15-minute adaptation period
to the room and sensors with eyes open. The therapist decides to stop at this point and not continue
with other aspects of an assessment, as he observes
that Mrs. Smith has approximately four times the
normal SEMG levels in both forehead and upper
trapezius muscle groups, based on office normative
data; the therapist does not notice any left–right
upper-trapezius muscle differences. He tentatively
concludes that (1) SEMG biofeedback would likely
help Mrs. Smith due to her elevated muscle tension, and (2) he can provide feedback from either
the forehead or upper trapezius regions and would
most likely choose to provide feedback from the
forehead as it is the standard placement for tension
headache. Being a careful clinician, he conducts
the same assessment prior to the first biofeedback session. Because he obtains the same results
he has a greater level of certainty concerning his
findings,and thus feels more confident in following his treatment plan. There is probably no need
for additional assessment at this point.
Throwing a Monkey Wrench
into the Straightforward SEMG Assessment
of a Patient with Tension Headache
Let us suppose the same scenario as that presented
earlier, except that the therapist on the second
assessment comes up with different results than
the therapist on the first assessment. Now, forehead SEMG levels remain at the same magnitude,
but trapezius readings are different: The right trapezius levels are about four times higher than his
normative group, but Mrs. Smith’s left trapezius is
10 times the normative levels. What can the therapist do in this situation?
There are multiple possible strategies and no
clearly right or wrong answers to this situation.
First, the therapist could decide that he needs
more information and conduct a more detailed
psychophysiological assessment, including stressor
conditions and assessment of various postures and
positions, either immediately or at the next session.
While this would give more information, it would
add significant cost to the treatment regimen and,
while it could simplify the treatment picture, it
could also cloud it even more. Second, he could
follow his original treatment plan, as forehead
SEMG abnormalities were found, and forehead
II. INSTRUMENTATION
SEMG biofeedback is a standard treatment for
tension headache (Arena & Blanchard, 2005). He
is still left, however, with the nagging question of
what do about the trapezius findings. He could still
give trapezius feedback should Mrs. Smith prove
refractory or have insufficient headache relief from
forehead feedback. Third, he could continue with
his plan of forehead feedback and monitor (but
not have the patient attempt to control) bilateral
trapezius levels. If on repeated biofeedback sessions this asymmetry continues, he may wish to
change the focus of the feedback to correcting the
left–right trapezius asymmetry if it does not dissipate (as, following a general relaxation theory of
frontal biofeedback, it is likely to do). Fourth, he
could decide to use the second assessment results
and change the focus of Mrs. Smith’s feedback to
correcting the left–right trapezius asymmetry. He
could then give forehead feedback should Mrs.
Smith prove refractory or obtain insufficient headache relief from trapezius feedback. Fifth, he could
decide to give Mrs. Smith feedback from all three
SEMG sites. This has the advantage of giving her
a more clear picture of her psychophysiological
abnormalities but might provide too much information, interfering with the psychophysiological
learning process. Sixth, he might devote part of
each session to each muscle area or focus on the
different areas in different sessions.
If the reader is getting confused by the wide
variety of treatment options posed by the psychophysiological assessment results, then we have
been successful in our endeavor to point out that
even simple assessments often do not have simple
answers. However, it is important to note that all
the previous options are perfectly defensible. Dr.
Arena would probably choose the third option,
and Dr. Schwartz, the fifth or sixth. We urge you
to document in your report the rationale for selecting a particular treatment direction so others can
understand your thought processes.
A More Involved SEMG Psychophysiological
Assessment of a Patient with Tension Headache
The psychophysiological assessment of Mrs. Smith
assumes no abnormalities during the adaptation and baseline conditions. The therapist plans
another assessment the next week that comprises
the same adaptation and baseline conditions,
three stressors (personally meaningful stressful
imagery, mental arithmetic–serial 9’s, and a cold
pressor task) of 4 minutes each interspersed with
3-minute poststress adaptation periods, followed
8. Psychophysiological Assessment and Biofeedback Baselines
by assessments in six different positions (standing,
bending from the waist, rising, sitting with back
unsupported, sitting with back supported, and
prone). Mrs. Smith again is within normal limits
during both baselines, but during the personally
meaningful stressful imagery, she has about six
times the normal forehead SEMG levels. During
both cognitive stressors, she takes longer to return
to baseline on forehead SEMG, and during the
standing and sitting unsupported positions she has
approximately 10 times the upper trapezius SEMG
levels than do normals. There are no left–right
upper-trapezius muscle differences during any condition.
Based on these data, one can tentatively conclude that
• EMG biofeedback would likely help Mrs. Smith
due to the psychophysiological assessment displaying abnormal patterns of SEMG responding.
• Since forehead SEMG levels were more abnormal during the stressor and poststress recovery
conditions, the therapist decides to use forehead
SEMG biofeedback initially.
• Following Mrs. Smith learning the forehead
SEMG biofeedback response, the therapist will
switch to the upper-trapezius muscle group.
• When he teaches generalization of the biofeedback response, the therapist will make sure he
trains Mrs. Smith to reduce her muscle tension
levels when she is standing and sitting, with her
back both supported and unsupported, as abnormalities in trapezius SEMG were noted during
the standing and sitting unsupported positions.
Given the fact that an impaired recovery process was found during both cognitive stressors
for forehead EMG, the therapist will repeatedly
practice with Mrs. Smith to rapidly decrease
her forehead SEMG levels, and he will emphasize the importance for her to reduce forehead
muscle tension levels immediately following a
stressful situation in her daily living.
Repeated assessment of the stressor and position
conditions would be helpful, but cost-effectiveness
and practicalities of clinical work probably preclude another assessment. Also, given that the
portion of the assessment the therapist repeated
was unchanged, there is more surety in the stability of the psychophysiological responding. One
could reasonably question why the therapist did
not include measures of other responses such as
heart rate, respiration, and hand surface tem-
147
perature. Inclusion of these measures might have
enhanced the clinical picture and allowed examination of stimulus–response specificity and individual response stereotypy. His line of reasoning
was likely that he has found in his clinical practice
that these measures do not add anything to assessment of patient with tension headache, and that
they would be prohibitive in terms of time.
The Refractory Migraineur: Scenario I
A therapist conducts 12 sessions of thermal biofeedback with Mr. Jones, a 49-year-old accountant
with migraines since age 16. Mr. Jones mastered
hand warming in the office, as well as outside
the office, but he has not experienced any headache relief. Therefore, his therapist conducts a
psychophysiological assessment with Mr. Jones
that comprises a variety of psychophysiological
measures (forehead, upper trapezius, posterior
neck and frontal–posterior neck SEMG, pulse,
hand temperature from multiple sites, respiration,
cephalic blood volume pulse, and electrodermal
response) in a variety of conditions (adaptation,
baseline, a relax body condition, and stressors such
as an exercise step-up test, fists, mental arithmetic, and personally meaningful negative imagery,
with poststress adaptation periods following each
stressor). Mr. Jones responds generally within normal limits during all conditions; however, when
examined for individual response stereotypy and
stimulus–response specificity during each stressor,
Mr. Jones responds with maximal arousal in some
of the SEMG measures, and he is able to relax
these responses the least during the relax body
condition. On the basis of the psychophysiological assessment, the therapist began SEMG biofeedback from those sites, including frontal and
focused daily relaxation instructions for those sites
and reinforced frequent daily practice of varying
durations.
This is a logical and defensible treatment plan
based on the psychophysiological data. One could
postulate that the assessment was contaminated
by the thermal biofeedback, and perhaps this
was the reason Mr. Jones did not respond abnormally during the assessment. One could also
advance the hypothesis that this explained his
not responding stereotypically with hand surface
temperature. The reader may also express concern
that the therapist did not question why the treatment was ineffective sooner (although some therapists have had patients who did not receive any
headache relief until 12 or more sessions). Also,
148
although repeated assessments might have shed
further light on Mr. Jones’s psychophysiology, costeffectiveness and clinical practicalities probably
precluded another assessment.
The Refractory Migraineur: Scenario II
In this scenario, the therapist found that Mr. Jones
responded within normal limits in the psychophysiological assessment. There were no patterns
in the examination of individual response stereotypy and stimulus–response specificity. Defensible
possible responses include
• Refer him because all likely possible approaches
are exhausted, and he is not improving.
• Repeat the psychophysiological assessment with
different stressors in hope of identifying abnormalities.
• Begin searching for other causes of the head
pain, such as secondary gains (days off from
work, children keeping quiet when headache
is present, etc.), dietary factors, psychological
characteristics (e.g., depression, anger, or anxiety).
• Assume the stressors were insufficient, and consider other relevant stressors for another assessment (e.g., playing a tape of Mr. Jones arguing
with his wife).
• Attempt a psychophysiological assessment in
the natural environment---through ambulatory recordings, accompanying him to work and
conducting an assessment in his office, etc.
• Check on and focus instructions on daily relaxation practice in terms of frequency, durations,
and timing.
• Consider including other muscle sites and other
skin temperature sites.
Routine SEMG Assessment of a Patient with Low
Back Pain
Some research on low back pain shows significant
paraspinal SEMG differences between patiens
with low back pain and nonpain controls, as well
as between low back patients with pain of differing etiologies (Arena, Sherman, Bruno, & Young,
1989b, 1991; Geisser et al., 2005; Kankaanpaa,
Taimela, Laaksonen, Hanninen, & Airaksinen,
1998). Arena often conducts psychophysiological
assessments of his back pain patients, using rightand left-sided L4–L5 paraspinal activity and bilateral biceps femoris (a muscle located in the back
II. INSTRUMENTATION
of the thigh; see Arena and Blanchard, 2002, for
a detailed figure depicting the sensor placement)
activity in six different positions. He looks for
three possible muscle tension abnormalities:
1. Unusually low muscle tension levels (perhaps
from nerve damage with resultant muscle atrophy).
2. Unusually high muscle tension levels (the most
frequent abnormality).
3. Asymmetry, in which one side of the back or
thigh muscles has normal muscle tension levels, while the other has unusually low or high
readings.
He might use biofeedback to decrease muscle
tension in the respective muscle groups. If an
asymmetry is found, he will use biofeedback to
help patients increase and/or decrease the abnormal sides. One goal is balanced bilateral values
within normal ranges.
For example, Mr. Doe is a 26-year-old bank
clerk and an amateur weightlifter. His paraspinal
SEMG levels are five to six times normal on all
positions except lying prone, which does not show
any abnormalities. There are also no abnormalities
with biceps femoris EMG. While sitting with his
back unsupported, his left paraspinal muscles are
over 30 times normal, and his right side is about
five times normal. This is the only time any left–
right asymmetry was found during the assessment.
Based on the psychophysiological assessment, the
therapist conducts bilateral paraspinal SEMG biofeedback while Mr. Doe sits with his back unsupported to correct for the asymmetry.
An Evaluation of a Low Back Pain Patient
in the Work Setting
We cannot overemphasize the importance of tailoring psychophysiological assessments and treatments to each patient. Clinicians must be creative and flexible. Take the example of Mr. Doe
and assume that he went through the regimen as
described and achieved a 40% reduction in his
back pain, but not as much as he and his therapist
had wanted and thought possible. The therapist
then explored Mr. Doe’s lifestyle in greater detail.
The therapist focused on the weightlifting as an
obvious and probable area of inappropriate muscle
usage but knows that Mr. Doe has been lifting
weights for only 2 years, whereas his back problems
began 4 years ago. He knows that Mr. Doe began
8. Psychophysiological Assessment and Biofeedback Baselines
working at a bank 5 years ago. Further inquiry
now about his job reveals that one of his major job
functions is to assist customers in accessing their
safe deposit boxes.
The therapist accompanies Mr. Doe to work and
notes that many of the safe deposit boxes are quite
high, and although there is a ladder in the vault,
Mr. Doe uses the ladder only for those boxes he
cannot reach. Moreover, their are other safe deposit
boxes that are nearly at floor level, and to access
them, Mr. Doe must either bend, squat, kneel, or
sit on the floor (the latter of which he never does).
A workplace psychophysiological assessment with
a portable SEMG of the paraspinal muscles when
Mr. Doe was either reaching up for a safe deposit
box or bending or squatting near the floor revealed
massive paraspinal SEMG activity.
The therapist then
• Instructed Mr. Doe to use the ladder when
accessing any safe deposit box that was higher
than head level.
• Used the portable SEMG biofeedback to
instruct Mr. Doe on proper ways to access boxes
on the lower wall levels.
• Provided Mr. Doe with office-based paraspinal
feedback of simulated access to safe deposit
boxes, reinforcing what was learned during the
evaluation in the vault.
• Instructed Mr. Doe to become much more
aware of his muscle tension levels whenever he
performed his safe deposit box job function, and
to conduct relaxation exercises after accessing
the boxes, in addition to maintaining the correct form and posture.
As a result, Mr. Doe’s back pain was reduced
significantly. Although the therapist’s standard
psychophysiological assessment included bending
from the waist and rising conditions, it was not
comprehensive enough to identify Mr. Doe’s problems. This was because bending from the waist
condition meant going from straight to about 30
degrees. Generally, this is all that the typical lower
back pain patient can do without experiencing significant increases in pain levels. Mr. Doe’s youth
and overall physical fitness allowed him to bend
much lower than the typical back pain patient,
demonstrating the importance of tailoring psychophysiological assessment and treatment to the
needs of each patient. (Please note that Mr. Doe is
not an actual patient but a combination of three
patients for illustrative purposes.)
149
Conclusion
Our intent in this chapter was to present to many
a reasonably detailed “how-to” on conducting
psychophysiological assessments in applied psychophysiology, as well as provide an understanding of the major pitfalls and questions that arise
when employing such procedures, especially with
patients with headaches or other chronic pain.
Our hope is to demystify somewhat the concept
of psychophysiological assessment. Good clinical
vvcommon sense and a basic understanding of the
general concepts of psychophysiology are required.
We urge clinicians to utilize psychophysiological
assessments when they have questions about their
patients’ treatment plans, or when they are confused about what may be causing or maintaining
their symptoms. Such evaluations often shed light
on complex clinical questions.
At the same time that we advocate practitioners
employment of such techniques, we caution readers to recognize limitations of psychophysiological
assessments and avoid using them to obtain simple
answers to complex questions. We are especially
concerned that the applied psychophysiologist will
give the same assessment to all patients, not tailoring it to the individual patient and disorder. We are
even more concerned about practitioners who give
every patient a psychophysiological assessment.
Such a practice is neither necessary nor fruitful.
We cannot rely solely on equipment for decision
making and practical application. As anyone
familiar with computers knows, the technology is
only as good as our related abilities and knowledge.
Still, after the best equipment is purchased, clinicians must formulate appropriate questions and
develop valid methods to answer them.
References
Abboud, F. (2010). In search of autonomic balance: The
good, the bad and the ugly. American Journal of Physiology: Regulatory, Integrative and Comparative Physiology,
298, R1449–R1467.
Airaksinen, O., & Airaksinen, K. (1998). Ambulatory
device for surface EMG recordings. Acupuncture and
Electro-Therapeutics Research, 23, 9–17.
Andrasik, F. (2010). Biofeedback in headache: An overview
of approaches and evidence. Cleveland Clinic Journal of
Medicine, 77, S72–S76.
Andrasik, F., Blanchard, E. B., Arena, J. G., Saunders, N.
L., & Barron, K. D. (1982). Psychophysiology of recur-
150
rent head pain: Methodological issues and new empirical findings. Behavior Therapy, 13, 407–429.
Andrasik, F., & Lords, A. (2004). Biofeedback. In L. Freeman (Ed.), Mosby's complementary and alternative medicine (2nd ed., pp. 207–236). St. Louis, MO: Mosby.
Andreassi, J. L. (2007). Psychophysiology: Human behavior
and physiological response. Mahwah, NJ: Erlbaum.
Appelhans, B., & Luecken, L. (2006). Heart rate variability
as an index of regulated emotional responding. Review of
General Psychology, 10, 229–240.
Arena, J. G. (1984). Inter- and intra-reliability of psychophysiological post-stress adaptation periods. Journal of
Behavioral Assessment, 6, 247–260.
Arena, J. G. (2000). Surface electromyographic psychophysiological assessment for chronic pain disorders:
The necessity for clinical and research inferences to be
empirically verified. Biofeedback, 28, 21–24.
Arena, J. G. (2005). A rose by any other name: The case for
baselines in psychophysiological research. Biofeedback,
33, 115–116.
Arena, J. G. (2010). Future directions in surface electromyography. Biofeedback, 38, 78–82.
Arena, J. G., & Blanchard, E. B. (1996). Biofeedback and
relaxation therapy for chronic pain disorders. In R. J.
Gatchel & D. C. Turk (Eds.), Chronic pain: Psychological perspectives on treatment (pp. 179–230). New York:
Guilford Press.
Arena, J. G., & Blanchard, E. B. (2002). Biofeedback training for chronic pain disorders: A primer. In R. J. Gatchel
& D. C. Turk (Eds.), Chronic pain: Psychological perspectives on treatment (2nd ed., pp. 159–186). New York:
Guilford Press.
Arena, J. G., & Blanchard, E. B. (2005). Assessment and
treatment of chronic benign headache in the primary
care setting. In W. O’Donohue, N. Cummings, D. Henderson, & M. Byrd (Eds.), Behavioral integrative care:
Treatments that work in the primary care setting (pp. 293–
313). New York: Allyn & Bacon.
Arena, J. G., Blanchard, E. B., Andrasik, F., Applebaum, K.,
& Myers, P. E. (1985). Psychophysiological comparisons
of three kinds of headache sufferers during and between
headache states: Analysis of post-stress adaptation periods. Journal of Psychosomatic Research, 29, 427–441.
Arena, J. G., Blanchard, E. B., Andrasik, F., Cotch, P. A.,
& Myers, P. E. (1983). Reliability of psychophysiological
assessment. Behaviour Research and Therapy, 21, 447–
460.
Arena, J. G., Bruno, G. M., Brucks, A., & Hobbs, S. H.
(1992, March). What is an adequate psychophysiological poststress adaptation period? In Proceedings of the
23rd annual meeting of the Association for Applied Psychophysiology and Biofeedback (pp. 90–92). Wheat Ridge,
CO: Association for Applied Psychophysiology and Biofeedback.
Arena, J. G., Bruno, G. M., Brucks, A. G., Searle, J. D.,
Sherman, R. A., & Meador, K. J. (1994). Reliability of
an ambulatory electromyographic activity device for
musculoskeletal pain disorders. International Journal of
Psychophysiology, 17, 153–157.
II. INSTRUMENTATION
Arena, J. G., Goldberg, S. J., Saul, D. L., & Hobbs, S. N.
(1989a). Temporal consistency of psychophysiological
stress profiles: Analysis of individual response stereotypy and stimulus response specificity. Behavior Therapy,
20, 609–618.
Arena, J. G., & Hobbs, S. N. (1995). Temporal stability of
psychophysiological stress profiles: A re-analysis using
intraclass correlation coefficients. Psychological Reports,
76, 171–175.
Arena, J. G., & Schwartz, M. S. (2003). Psychophysiological assessment and biofeedback baselines: A primer. In
M. S. Schwartz & F. Andrasik (Eds.), Biofeedback: A
practitioner’s guide (3rd ed., pp. 128–148). New York:
Guilford Press.
Arena, J. G., Sherman, R. A., Bruno, G. M., & Young, T. R.
(1989b). Electromyographic recordings of five types of
low back pain subjects and non-pain controls in different positions. Pain, 37, 57–65.
Arena, J. G., Sherman, R. A., Bruno, G. M., & Young, T.
R. (1990). Temporal stability of paraspinal electromyographic recordings in low back and non-pain subjects.
International Journal of Psychophysiology, 9, 31–37.
Arena, J. G., Sherman, R. A., Bruno, G. M., & Young, T.
R. (1991). Electromyographic recordings of five types of
low back pain subjects and non-pain controls in different positions: Effect of pain state. Pain, 45, 23–28.
Asterita, M. F. (1985). The physiology of stress. New York:
Human Sciences Press.
Auchincloss, C., & McLean, L. (2009). The reliability of
surface EMG recorded from the pelvic floor muscles.
Journal of Neuroscience Methods, 182, 85–96.
Baptista, V. (2006). Starting physiology: Understanding homeostasis. Advances in Physiology Education, 30,
263–264.
Basmajian, J. V., & De Luca, C. J. (1985). Muscles alive:
Their functions revealed by electromyography (5th ed.).
Baltimore, MD: Williams & Wilkins.
Berman, P., & Johnson, H. (1985). A psychophysiological
assessment battery. Biofeedback and Self-Regulation, 10,
203–221.
Berntson, G. G., Quigley, K. S., & Lozano, D. (2007). Cardiovascular psychophysiology. In J. T. Cacioppo, L. G.
Tassinary, & G. G. Berntson (Eds.), Handbook of psychophysiology (3rd ed., pp. 182–210). Cambridge, UK:
Cambridge University Press.
Box, G., Hunter, J. S., & Hunter, W. G. (2005). Statistics for
experimenters: Design, innovation and discovery (2nd ed.).
Hoboken, NJ: Wiley.
Buco, S. M., & Blouin, D. C. (1983). Use of similarity coefficients when populations parameters are unknown. Paper
presented at the 29th annual meeting of the Southeastern Psychological Association, New Orleans, LA.
Burnham, R., McKinley, R., & Vincent, D. (2006). Three
types of skin surface thermometers: A comparison
of reliability, validity and responsiveness. American
­Journal of Physical Medicine and Rehabilitation, 85, 553–
558.
Cacioppo, J. T., Berntson, G. G., & Anderson, B. L. (1991).
Psychophysiological approaches to the evaluation of
8. Psychophysiological Assessment and Biofeedback Baselines
psychotherapeutic process and outcome: Contributions
from social psychophysiology. Psychological Assessment,
3, 321–336.
Callaghan, M., McCarthy, C., & Oldham, J. (2009). The
reliability of surface electromyography to assess quadriceps fatigue during multi joint tasks in healthy and painful knees. Journal of Electromyography and Kinesiology,
19, 172–180.
Campbell, B., Wood, G., & McBride, T. (1997). Origins of
orienting and defensive responses: An evolutionary perspective. In P. J. Lang, R. F. Simmons, & M. Balabam
(Eds.), Attention and orienting: Sensory and motivational
processes (pp. 41–68). Mahwah, NJ: Erlbaum.
Carrasco, S., Gonzalez, R., Gaitan, M., & Yanez, O. (2003).
Reproducibility of heart rate variability from short-term
recordings during five manoeuvres in normal subjects.
Journal of Medical Engineering and Technology, 27, 241–
248.
Castroflorio, T., Icardi, K., Becchino, B., Merlo, E., Debernardi, C., Bracco, P., et al. (2005). Reproducibility of surface EMG variables in isometric sub-maximal contractions of jaw elevator muscles. Journal of Electromyography
and Kinesiology, 16, 498–505.
Criswell, E. (2011). Cram's introduction to surface electromyography (2nd ed.). Sudbury, MA: Jones & Bartlett.
Dawson, M., Schell, A., & Filion, D. (2007). The electrodermal system. In J. T. Cacioppo, L. G. Tassinary, & G.
G. Berntson (Eds.), Handbook of psychophysiology (3rd
ed., pp. 159–181). Cambridge, UK: Cambridge University Press.
Dienstbier, R. A. (1989). Arousal and physiological toughness: Implications for mental and physical health. Psychological Review, 96, 84–100.
Engel, B. T. (1960). Stimulus–response and individual
response specificity. Archives of General Psychiatry, 2,
305–313.
Eppinger, H., & Hess, L. (1917). Vagotonia. New York: Nervous and Mental Disease Publishing Company.
Feinstein, R. A., Hains, C. S., Hemstreet, M. P., TurnerHenson, A., Redden, D. T., Martin, B., et al. (1999).
A simple “step-test” protocol for identifying suspected
unrecognized exercise-induced asthma (EIA) in children. Allergy and Asthma Proceedings, 20, 181–188.
Foerster, F. (1985). Psychophysiological response specificities: A replication over a 12-month period. Biological
Psychology, 21, 169–173.
Foerster, F., Schneider, H., & Walschburger, P. (1983). The
differentiation of individual-specific, stimulus-specific,
and motivation-specific response patterns in activation
processes: An inquiry investigating their stability and
possible importance in psychophysiology. Biological Psychology, 17, 1–26.
Furedy, J. J. (1987). Beyond heart rate in the cardiac psychophysiological assessment of mental effort: The
T-wave amplitude component in the electrocardiogram.
Human Factors, 29, 183–194.
Geisser, M., Ranavaya, M., Haig, A., Roth, R., Zucker, R.,
Ambroz, C., et al. (2005). A meta-analytic review of
surface electromyography among persons with low back
151
pain and normal, healthy controls. Journal of Pain, 16,
711–726.
Gerin, W., Christenfeld, N., Pieper, C., DeRafael, D. A., Su,
O., Stroessner, S. J., et al. (1998). The generalizability of
cardiovascular responses across settings. Journal of Psychosomatic Research, 44, 209–218.
Gerin, W., Pieper, C., & Pickering, T. (1994). Anticipatory
and residual effects of an active coping task on pre- and
post-stress baselines. Journal of Psychosomatic Research,
38, 139–149.
Gevirtz, R., & Schwartz, M. K. (2003). The respiratory system in applied psychophysiology. In M. S. Schwartz & F.
Andrasik (Eds.), Biofeedback: A practitioner’s guide (3rd
ed., pp. 212–244). New York: Guilford Press.
Guijt, A., Sluiter, J., & Frings-Dresen, M. (2007). Test–
retest reliability of heart rate variability and respiration
rate at rest and during light physical activity in normal
subjects. Archives of Medical Research, 38, 113–120.
Hastrup, J. L. (1986). Duration of initial heart rate assessment in psychophysiology: Current practices and implications. Psychophysiology, 23, 5–17.
Haynes, S. N., Falkin, S., & Sexton-Radek, K. (1989). Psychophysiological measurement in behavior therapy. In
G. Turpin (Ed.), Handbook of clinical psychophysiology
(pp. 175–214). London: Wiley.
Haynes, S. N., Gannon, L. R., Orimoto, L., O’Brien, W.
H., & Brandt, M. (1991). Psychophysiological assessment of poststress recovery. Psychological Assessment, 3,
356–365.
Heinz, A., Huber, B., Schreinicke, J., & Seibt, R. (2002).
Temporal stability of psychophysiological response patterns: Concepts and statistical tools. International Journal of Psychophysiology, 44, 57–65.
Holmes, T. H., & Rahe, R. H. (1967). The social readjustment scale. Journal of Psychosomatic Research, 11, 213–
218.
Jamieson, J. (1999). Dealing with baseline differences: Two
principles and two dilemmas. International Journal of
Psychophysiology, 31, 155–161.
Jennings, J. R., Kamarck, R., Stewart, C., Eddy, M., &
Johnson, P. (1992). Alternate cardiovascular baseline
assessment techniques: Vanilla or resting baseline. Psychophysiology, 29, 730–742.
Johnson, M., & Tabasam, G. (2003). An investigation
into the analgesic effects of interferential currents and
transcutaneous electrical nerve stimulation on experimentally induced ischemic pain in otherwise pain-free
volunteers. Physical Therapy, 83, 208–233.
Kankaanpaa, M., Taimela, S., Laaksonen, D., Hanninen,
O., & Airaksinen, O. (1998). Back and hip extensor
fatigability in chronic low back pain patients and controls. Archives of Physical Medicine and Rehabilitation, 79,
412–417.
Kim, H., Sohn, D., Lee, E., Kim, Y, Oh, B., & Park, Y.
(2006). Coronary blood flow after cold exposure and
the effects of warm-up exercise. Journal of the American
Society of Echocardiology, 19, 386–390.
Kirk, R. E. (1995). Experimental design: Procedures for the
behavioral sciences (3rd ed.). Belmont, CA: Brooks/Cole.
152
Kronenberg, H. M., Melmed, S., Polonsky, K. S., & Larsen,
P. R. (Eds.). (2008). Williams textbook of endocrinology
(11th ed.). Philadelphia: Saunders/Elsevier.
Lariviere, C., Arsenault, A., Gravel, D., Gagnon, D., &
Loisel, P. (2002). Evaluation of measurement strategies
to increase the reliability of sEMG indices to access back
muscle fatigue and recovery. Journal of Electromyographic
Kinesiology, 12, 91–102.
Lichstein, K. L., Sallis, J. F., Hill, D., & Young, M. C. (1981).
Psychophysiological adaptation: An investigation of
multiple parameters. Journal of Behavioral Assessment,
3, 111–121.
Lim, P., Shiels, P., Anderson, J., & MacDonald, T. (1999).
Dundee step test: A simple method of measuring the
blood pressure response to exercise. Journal of Human
Hypertension, 13, 521–526.
Linden, W. (1991). What do arithmetic stress tests measure?: Protocol variations and cardiovascular responses.
Psychophysiology, 28, 91–102.
Lorig, T. S. (2007). The respiratory system. In J. T. Cacioppo,
L. G. Tassinary, & G. G. Berntson (Eds.), Handbook of
psychophysiology (3rd ed., pp. 231–244). Cambridge, UK:
Cambridge University Press.
Lovallo, W. R. (2005). Cardiovascular reactivity: Mechanisms and pathways to cardiovascular disease. International Journal of Psychophysiology, 58, 119–132.
MacLeod, C. M. (1991). Half a century of research on the
Stroop Effect: An integrative review. Psychological Bulletin, 109, 163–203.
Manuck, S. B., & Schaefer, B. C. (1978). Stability of individual differences in cardiovascular reactivity. Physiology and Behavior, 21, 675–678.
Meyers, A. W., & Craighead, W. E. (1978). Adaptation
periods in clinical psychophysiological research: A recommendation. Behavior Therapy, 9, 355–362.
Mitchell, L., McDonald, R., & Brodie, E. (2004). Temperature and the cold pressor test. Journal of Pain, 5, 233–237.
Netto, K., & Burnett, A. (2006). Reliability of normalization methods for sEMG analysis of neck muscles. Work,
26, 123–130.
Nussinovitch, U., Elishkevitz, K., Katz, K., Nussinovitch,
M., Segev, S., Volovitz, B., et al. (2011). Reliability of
ultra-short ECG indices for heart rate variability. Annals
of Noninvasive Electrocardiology, 16, 117–122.
Peek, C. J. (2003). A primer of biofeedback instrumentation. In M. S. Schwartz & F. Andrasik (Eds.), Biofeedback: A practitioner’s guide (3rd ed., pp. 43–102). New
York: Guilford Press.
Piferi, R. L., Kline, K. A., Younger, J., & Lawler, K. A.
(2000). An alternative approach for achieving cardiovascular baseline: Viewing an aquatic video. International Journal of Psychophysiology, 37, 207–217.
Pinerua-Shuhaibar L., Prieto-Rincon, D., Ferrer, A.,
Bonilla, E., Maixner, W., & Suarez-Roca, H. (1999).
Reduced tolerance and cardiovascular response to ischemic pain in minor depression. Journal of Affective Disorders, 56, 119–126.
Rashed, H. M., Leventhal, G., Madu, E. C., Reddy, R., &
II. INSTRUMENTATION
Cardoso, S. (1997). Reproducibility of exercise-induced
modulation of cardiovascular responses to cold stress.
Clinical Autonomic Research, 7, 93–96.
Robinson, J., Whitsett, S., & Kaplan, B. (1987). The stability of psychophysiological reactivity over multiple sessions. Biological Psychology, 24, 129–139.
Rosenthal, T. L., Montgomery, L. M., Shadish, W. R.,
Edwards, N. B., Hutcherson, H. W., et al. (1989). Two
new, brief, practical stressor tasks for research purposes.
Behavior Therapy, 20, 545–562.
Sallis, J. F., & Lichstein, K. L. (1979). The frontal electromyographic adaptation response: A potential source of
confounding. Biofeedback and Self-Regulation, 4, 337–339.
Schwartz, M. S. (1995). Biofeedback: A practitioner’s guide
(2nd ed.). New York: Guilford Press.
Schwartz, M. S., & Andrasik, F. (Eds.). (2003). Biofeedback:
A practitioner’s guide (3rd ed.). New York: Guilford Press.
Schwartz, M. S., Schwartz, V. J., & Monastra, V. (2003).
Problems with relaxation and biofeedback-assisted
relaxation, and guidelines for management. In M. S.
Schwartz & F. Andrasik (Eds.), Biofeedback: A practitioner’s guide (3rd ed., pp. 251–264). New York: Guilford
Press.
Sella, G. (2005). A rose is but a rose . . . Biofeedback, 33,
114–115.
Shaffer, F., Sponsel, M., Kice, J., & Hollensbe, J. (1991,
March 15–20). Test–retest reliability of resting baseline
measurements. In Proceedings of the 22nd Annual Meeting of the Association for Applied Psychophysiology and
Biofeedback, New Orleans (pp. 128–130). Wheat Ridge,
CO: Association for Applied Psychophysiology and Biofeedback.
Sharpley, C. F. (1993). Effects of brief rest periods upon
heart rate in multiple baseline studies of heartrate reactivity. Biofeedback and Self-Regulation, 18, 225–235.
Speckenbach, U., & Gerber, W. D. (1999). Reliability of
infrared plethysmography in BVP biofeedback therapy
and the relevance for clinical application. Applied Psychophysiology and Biofeedback, 24, 261–265.
Sternbach, R. S. (1966). Principles of psychophysiology. New
York: Academic Press.
Strike, P., & Steptoe, A. (2003). Systematic review of mental stress-induced myocardial ischemia. European Heart
Journal, 24, 690–703.
Sturgis, E. T. (1980, November). Physiological lability and
reactivity in headache activity. Paper presented at the
Association for Advancement of Behavior Therapy 14th
annual convention, New York.
Sturgis, E. T., & Arena, J. G. (1984). Psychophysiological
assessment. In M. Hersen, R. Eisler, & P. M. Miller (Eds.),
Progress in behavior modification (Vol. 17, pp. 1–30). New
York: Academic Press.
Sturgis, E. T., & Gramling, S. (1988). Psychophysiological
assessment. In A. S. Bellack & M. Hersen (Eds.), Behavioral assessment: A practical handbook (3rd ed., pp. 213–
251). New York: Pergamon Press.
Sutarto, A. P., Abdul Wahab, M. N., & Mat Zin, N. (2010).
Heart rate variability (HRV) biofeedback: A new train-
8. Psychophysiological Assessment and Biofeedback Baselines
ing approach for operator’s performance enhancement.
Journal of Industrial Engineering and Management, 3,
176–198.
Taub, E., & School, P. J. (1978). Some methodological considerations in thermal biofeedback training. Behavior
Research Methods and Instrumentation, 10, 617–622.
Veit, R., Brody, S., & Rau, H. (1997). Four-year stability of
cardiovascular reactivity to psychological stress. Journal
of Behavioral Medicine, 20, 447–460.
Velasco, M., Gomez, J., Blanco, M., & Rodriguez, I. (1997).
The cold pressor test: Pharmacological and therapeutic
aspects. American Journal of Therapeutics, 4, 34–38.
Waters, W. F., Williamson, D. A., Bernard, B. A., Blouin,
D. C., & Faulstich, M. E. (1987). Test–retest reliability of
psychophysiological assessment. Behaviour Research and
Therapy, 25(3), 213–221.
153
Wegner, D. M., Broome, A., & Blumberg, S. (1997). Ironic
effects of trying to relax under stress. Behaviour Research
and Therapy, 35, 11–21.
Wenger, M. A. (1966). Studies of autonomic balance: A
summary. Psychophysiology, 2, 173–186.
Wheat, A., & Larkin, K. (2010). Biofeedback of heart rate
variability and related physiology. Applied Psychophysiology and Biofeedback, 35, 229–242.
Wilder, J. (1950). The law of initial values. Psychosomatic
Medicine, 12, 392–400.
Yoshida, K., Utsunomiya, T., Morooka, T., Yazawa, M.,
Kido, K., Ogawa, T., et al. (1999). Mental stress test is an
effective inducer of vasospastic angina pectoris: Comparison with cold pressor, hyperventilation and master
two-step exercise test. International Journal of Cardiology, 70, 155–163.
Chapter 9
Consumer‑ and Home‑Based Biofeedback
Mark S. Schwartz and Frank Andrasik
Rationale and Caveat
devices totally on their own without at least a professional’s recommendation, evaluation, and guidance.
There are now several biofeedback1 instruments2
designed for consumers, and some are marketed
directly to the public. Some include computerbased systems. Others label their devices other
than biofeedback. The creation of these devices,
mostly in the last several years, their widespread
use during the last few years, increased sophistication, inclusion in many research studies, use by
many practitioners, and recommendation by many
practitioners for patients are among the factors
involved in the rationale for reviewing them in
this new chapter.
Inclusion of this chapter is a major departure
from the viewpoints long held by us and many
other practitioners that such devices should not
be marketed directly to the public. Rather, these
devices were best viewed as supplements to officebased interventions when practitioners felt them to
be appropriate for use. The chief rationale for this
position was the complexity of biofeedback and
the resultant need for supervision by appropriately
trained therapists. Thus, this topic was purposely
not addressed in prior editions of this book. This
view is likely still shared by many practitioners and
remains valid for many, perhaps most, patients.
We still support the view that obtaining consultation, evaluation, and interventions by a properly
credentialed professional has many advantages to
and is preferable to laypersons purchasing these
A Very Brief History of Consumer‑Based
Biofeedback Products
For many years a few devices were (and continue to
be) marketed directly to the public. Other modalities (e.g., single-channel surface electromyograph
[SEMG], skin temperature) have also long been
available to the public and marketed to practitioners to recommend to selected patients to supplement office-based interventions.
Alpha electroencephalographic (EEG) biofeedback devices were available many years ago when
alpha biofeedback was part of the formative history of biofeedback. However, we have neither
seen them advertised nor have they been available
from major manufacturers or distributors for many
years.
The issues initially associated with devices
available to consumers centered more on the
potential for inappropriate and unsupervised
intervention, presumably by persons searching for
help with symptoms and conditions that might not
be correctly evaluated, diagnosed, and treated by a
layperson or by insufficiently educated and inappropriately credentialed professionals. Quality or
accuracy of the measurements was typically much
154
155
9. Consumer‑ and Home‑Based Biofeedback
less of a focus, because the early devices were far
less sophisticated than the more elaborate devices
and software–hardware systems available only to
professionals; thus, they were not seen as being
competitive.
So, What Has Changed?
We believe that several factors, summarized in
Table 9.1, have influenced the changes. These proposed factors have impacted the availability and
acceptance of biofeedback devices and systems to
consumers.
What Has Not Changed?
Direct marketing to the public still means that we
assume the following:
• There is widespread unsupervised use of such
devices.
• Many people are buying these devices unneces-
sarily. This includes people for whom biofeedback is not indicated or for whom the modality
selected is incorrect.
• Many people are using these devices incorrectly.
• Many people do not benefit, or they have new
or worsening symptoms.
• However, some of the devices purchased directly
by the public are done so at the specific recommendations of health care professionals who
are properly credentialed in biofeedback, and
who supervise the consumer or provide proper
instructions before or soon after the purchase.
• There is the potential that some people, failing
to find benefit, become disappointed and are
thereore less likely to seek professional help that
could be beneficial in the future. One might
argue that the patient is worse off than he or
she was prior to trying the device.
• Thus, we assume that some consumers for
whom biofeedback is indicated have become
discouraged because of misuse or insufficient
use of devices.
TABLE 9.1. Factors Influencing the Development and Growth of Consumer Devices
Factors
Sources and comments
Increased awareness of biofeedback by the
public
Lay articles, “word of mouth,” and health care
provider recommendations
Increased positive attitudes about biofeedback
by the public
Same as above
Increased positive attitudes about biofeedback
by health care professionals
Published research, credible biofeedback providers
Increased acceptance and recommendation by
major medical institutions
Staff physicians and psychologists, websites
Increased sophistication of devices
Advances by biomedical engineers
Increased information, acceptance, and
recommendations by medical websites
Medical advice websites that are independent of
major medical institutions
Increased sophistication of computer software
Advances by biomedical medical programmers
Increased expense of conventional health care
Limited third-party coverage for biofeedback
Health insurance companies are sometimes reluctant
to reimburse office-based biofeedback due to multiple
factors, including costs and varied types of providers
Number of providers of biofeedback
Growth continues but remains slow, partly due to
limited educational/training opportunities
Continued evolving/growth of self-help zeitgeist
Widespread acceptance of complementary and
alternative medicine
156
Are there exceptions, caveats, cautions, and/
or contraindications for biofeedback devices
being marketed directly to the public? Yes. The
most obvious ones to us involve EEG biofeedback devices. We share the serious concerns of
Hammond and Kirk (2007), who go so far as to
express “alarm” that “dealers, manufacturers, and
trainers . . . have been supplying EEG biofeedback
equipment direct to lay persons” (p. 140). Their
concern is because
• Some of these individuals open their own practices and advertise to the public that they can
treat a variety of disorders, all of which are serious and complicated, and for which they have
no advanced degrees or health care licenses for
independent practice.
• Purchasers in the public who attempt to selfmanage their treatment without supervision
from a competent professional3 risk negative
side effects and adverse reactions ranging from
“very mild, transient . . . to very serious conditions” (p. 140). Everyone in the public who is
considering the purchase of EEG biofeedback
devices, especially without properly credentialed and licensed health care professionals
supervising their therapy, needs to read this
article.
Devices Selected
for Brief Review
Consumer devices discussed in this chapter
include those that are:
• Marketed to or available to the public4 without
the necessity of a health care professional’s recommendation or prescription.
• Recommended by practitioners to supplement
office-based interventions.
We focus on three devices (RESPeRATE,
EmWave, Wild Devine) and a cloud-based system
(BFA Monitor). We intentionally do not cover this
exhaustively due primarily to space limitations.
Excluded from this chapter are:
• Devices that are small enough to be portable or
ambulatory but are solely available to practitioners.
• Devices such as liquid crystal temperature
devices and small thermometers attached to fingers that have been used for decades to supple-
II. INSTRUMENTATION
ment practitioner services. They are available
to the public and are considered biofeedback
devices by some professionals, although they do
not meet the criteria of the official definition5
of biofeedback.
• All EEG biofeedback devices.
• Temperature-sensitive devices.
• Exclusively electrodermal devices.
Issues to Consider when Buying,
Recommending, and/or Using
Consumer‑Available Devices
Issues include (1) advantages, (2) ease of use, (3)
effectiveness for claimed/expected changes, (4)
readability of instructions and explanations, (5)
potential for misuse, (6) risks of negative side
effects, (7) potential dangers, (8) reliability of the
device, (9) accuracy, (10) research, (11) cost, and
(12) likelihood of adherence to recommended use
by user.
Product6 Availability
and Contact Information
RESPeRATE
Resperate, Inc.
220 Meridian Blvd., Suite 07735
Minden, NV 89423
Phone: 800-220-1925
Website: www.resperate.com
EmWave
HeartMath, LLC
14700 West Park Avenue
Boulder Creek, CA 95006
Phone: 831-338-8700 or 800-450-9111
Website: www.heartmath.com
Also available from biofeedback distributors.
Wild Divine
Wild Divine, Inc.
9550 South Eastern Avenue, Suite 253
Las Vegas, NV 89123
Phone: 866- 594-9453
Website: www.wilddivine.com
Products include The Journey to the Wild Devine,
Relaxing Rhythms, and Wisdom Quest.
157
9. Consumer‑ and Home‑Based Biofeedback
Biofeedback distributors (e.g., http://bio-medical.
com, www.stens-biofeedback.com) carry all these
and other devices.
Are These Devices Referred to as Biofeedback
by Their Manufacturers?
These three devices/systems are not referred to
as biofeedback in their published materials and
on their websites (see Note 4). For the RESPeRATE, the 2002 510(k) summary from InterCure
in Lod, Israel, referred to it as a “biofeedback
device . . . intended for use as a relaxation treatment for the reduction of stress by leading the
user through interactively guided and monitored
breathing exercises.” The device shares the same
general intended use in relaxation and/or stress
reduction and the same indications for use. Thus,
the RESPeRATE is appropriate as a breathing
biofeedback-assisted device for use with any symptoms or disorders for which this type of breathing
therapy is recommended, although it is currently
approved for hypertension by the U.S. Food and
Drug Administration (FDA), which also notes its
value for stress reduction.
The Wild Divine website specifically addresses
this issue on its home page. It is careful to point
out that its products are different from biofeedback devices, which are considered “clinical medical instruments.” The home page emphasizes that
its products provide “active feedback with rich,
graphical feedback integrated into the training
program (not biofeedback), emphasizing that its
products “are for educational, entertainment, and
leisure use” and “should not be used in place of
professional medical care.” It “strongly urges” users
“to discuss any and all alternative medical therapies” with their “doctor or health professional”
(www.wilddivine.com home page).
Advantages of These Devices
• Small size. All are very portable and can be carried either in one's pocket (EmWave), briefcase
(Wild Devine), or small bag (RESPeRATE).
• Ease of use. All are relatively easy to learn. The
RESPeRATE is probably the easiest to learn.
• Time to learn. Typically, a few hours, which
includes reading the manual and interacting
with the devices.
• Provision of readily understandable feedback.
• Reasonably priced. Most cost in the range from
about $150 to $350.
• Durable.
• Reliable.
• Accurate.
Some Cautions and Possible Contraindications
for Using, at Least, the Heart Rate Variability
(HRV) Devices
• A patient who is using a pacemaker that regulates heart rate.
• A patient who uses some heart-regulating medications (especially high dosages).
• A patient with chronic low circulation.
• A low finger temperature, because it might
interfere with obtaining a pulse reading.
• A patient who engages in significant movement.
Additional Issues with RESPeRATE
1. Strap location. It is surprising that parts of
the website instruct the user to place the strap
around the chest rather than the usual and preferred abdomen.
2. Frequency of use. The manufacturer’s website and other websites and printed materials recommend 15 minutes at least three to four times per
week. They then state that “typically, within 30
days of use, a sustained reduction in blood pressure
can be achieved.” Like other physical exercises,
regular use is required to maintain the benefits.
Concern: This implies that three or four 15-minute periods of relaxed breathing are sufficient to
significantly reduce blood pressures and generalize
the reductions among persons with hypertension.
Although not explicitly stated, the implication
is that one does not need to use relaxed breathing more than a total of 1 hour a week to achieve
clinically significant results. We believe that many
health care professionals who treat people with
hypertension consider that much more frequent
relaxation, including relaxed breathing, is much
preferred and probably needed.
3. The manufacturer’s website states that “ten
clinical studies have proven the effectiveness of
RESPeRATE on lowering blood pressure.” Concern: The disagreement here is with the term
“proven.” “Supported,” yes, but research does not
prove anything.
4. The manufacturer’s website highlights and
implies support from the Mayo Clinic. Concern:
The concern of one author (Schwartz) is the focus
on using the Mayo Clinic name. The company
has repeatedly used the Mayo Clinic name as an
158
implied endorsement. In prior versions of their
website, the implied endorsement was even more
explicit. This is inconsistent with the author’s
(Schwartz) experience with Mayo Clinic.
Uses and Applications
In our opinion, several uses and applications are
implied. These devices provide biofeedback to
help users learn and to facilitate skillful and efficient slow, deep, rhythmic diaphragmatic breathing,7 a major intervention/therapy procedure for
patients/clients/persons with a wide variety of
symptoms, disorders, and purposes (e.g., anxiety
disorders involving psychophysiological arousal—
panic, general anxiety, posttraumatic stress disorder [PTSD], phobias, acute stress, migraine headaches, hyperventilation, functional chest pain,
functional nausea/vomiting, asthma).
RESPeRATE is specifically designed for respiratory biofeedback that measures breathing (e.g.,
rate of inhalations and exhalations, depth and
location) and uses associated electronics and software, combined with a strain gauge transducer
that measures the displacement of the abdomen.
It is designed to help entrain slower and deeper
relaxed breathing.
The Iom Active Feedback Hardware measures
and provides feedback for HRV and skin conductance level.
EmWave2 measures HRV via pulse plethysmography (PPG) on a finger or ear lobe. Users learn to
use breathing and “positive emotions” to achieve
“coherence” between breathing and cardiac
rhythms. Feedback includes visual and audio stimuli, and computer-based software graphic feedback.
Cloud‑Based Distance Biofeedback8
This is another exciting, innovative approach with
great heuristic value. This cloud-based biofeedback service could work with any kind of modality. The most common modalities currently being
used in this system involve cardiac blood volume
(e.g., HRV) and EEG, but other modalities are
planned. The software is partly on the patient/
client/subject’s home computer and mostly on the
cloud-based location. The practitioner/therapist
uploads to the cloud program his or her evaluation
and intervention content and program (e.g., text,
photos, audio clips, video clips). The cloud-based
II. INSTRUMENTATION
program analyzes data, provides feedback, stores
the data, sends automatic reports to the patient,
and communicates the results to the practitioner.
The patient logs into the website of the practitioner’s office, clinic, hospital, or research laboratory.
Then, the patient follows the instructions while
his or her psychophysiological data are transmitted to the cloud-based program. Clients’ identities
are coded to maintain confidentiality. Each practitioner knows the code of each client, and the
program provides the practitioner with frequent
updates regarding the client’s adherence and psychophysiological data.
Research Questions
Applied research with these devices and systems is
still in the nascent stage. We include a few questions that we would like to see researched in the
next few years (many of which are similar to those
raised by Glasgow and Rosen, 1978, in their seminal review of behavioral bibliotherapy):
• Does use of one or more of these devices without
professional supervision and guidance produce
different results compared to those obtained
with professional supervision and guidance?
Are outcomes comparable to those found when
the same modality is provided with more elaborate instrumentation in office sessions?
• Does the use of the RESPeRATE preceding
EmWave result in faster acquisition of desired
HRV outcomes compared to starting with
EmWave?
• Are there meaningful differences in outcomes
between devices?
• Do clients express preferences for particular
devices and, if so, what client characteristics are
associated with these preferences?
• How important is it to follow the manufacturer’s
guidelines for use? For example, do outcomes
differ if practitioners recommend a different frequency of use?
• If practitioners vary their level of involvement,
could certain combinations provide more favorable cost–benefit returns? For example, would
differences emerge when patients use the device
totally independently rather than combined
with varied amounts of therapist support (prudent limited office treatment vs. multiple office
sessions in conjunction with home use)?
159
9. Consumer‑ and Home‑Based Biofeedback
• What characteristics of users are associated with
effective and ineffective use of these devices?
• How often do patients/clients actually practice
relaxed breathing alone and/or HRV aside from
the number of times they use the devices when
treating their symptoms (e.g., hypertension)?
Conclusions and Recommendations
All of the biofeedback devices and systems
described in this chapter are excellent and welldesigned. The devices are logical and probably
useful for treating a wide variety of symptoms and
disorders for which at least slow, deep, diaphragmatic breathing and/or HRV have been found
clinically useful. These devices can be used in the
professional’s office, loaned or “rented” for home
use by the professional, or purchased as a supplement to professionally supervised intervention.
In sum:
• Consumer- and home-based biofeedback devices
are here to stay.
• There are many logical, safe, effective, and supported uses and applications of these devices.
• There are limitations, cautions, and contraindications that vary across modalities and specific
devices, especially when used without supervision by appropriately educated and credentialed
health care professionals.
• These devices will become more numerous,
and, with increased software and programming,
will continue to become more sophisticated.
• Well-controlled research (we hope) will continue to increase, provide supportive evidence,
and lead to development of guidelines for use
and misuse.
Notes
1. Some of these instruments are not called “biofeedback” by the manufacturers, yet they are clearly and
unequivocally biofeedback instruments or devices. “A rose
by any other name . . . ”
2. The terms “instrument” and “device” are used interchangeably in this chapter. Some companies use the term
“device” presumably for commercial value.
3. Of course, negative side effects and adverse reactions can also occur in persons being treated by health care
professionals who are properly credentialed and licensed,
especially if they have less training or experience, although
presumably and logically the risks are considerably less.
Also, when adverse events occur within a therapeutic relationship, they are more likely to be noticed and acted upon.
Further discussion of this topic, however, goes beyond the
scope of this chapter.
4. These devices are also recommended by practitioners
to supplement office-based interventions, and loaned to or
rented to recipients to supplement office-based interventions.
5. See www.aapb.org, www.bcia.org, and www.isnr.org.
6. The order of presentation of devices in this chapter
is essentially random.
7. Referred to in this chapter as “relaxed breathing.”
8. See www.bfa-global.com (BFA-GLOBAL, Biofeedback Analytics Ltd., Phone: +972-52-3334888, Email:
Info@BFA-Global.com).
References
Glasgow, R. E., & Rosen, G. M. (1978). Behavioral bibliotherapy: A review of self-help behavior therapy manuals.
Psychological Bulletin, 85(1), 1–23.
Hammond, D. C., & Kirk, L. (2007). Negative effects and
the need for standards of practice in neurofeedback. Biofeedback, 35(4), 139–145.
P a r t III
Adjunctive/Complementary
Interventions
Ch a p ter 10
Dietary Considerations
Keith I. Block, Charlotte Gyllenhaal, and Mark S. Schwartz
Many health care professionals believe, and many
patients report, that certain dietary elements or
patterns aggravate or trigger certain physical or
psychological symptoms. Many of these symptoms
are also treated with applied psychophysiological
interventions such as biofeedback (e.g., migraine
headaches, anxiety, irritable bowel syndrome,
insomnia). The idea that specific foods act as triggers has received closer examination in recent
years, and doubts have been raised about the
importance of some classic migraine trigger foods.
At the same time, other aspects of diet are beginning to look more relevant in migraine, and implications of diet both for intensification of migraine
and other headaches, and for long-term health of
migraine patients are becoming more important.
We emphasize in this chapter the relevance of diet
in migraine, although not to the exclusion of other
conditions.
brain chemicals, with irritation and inflammation
of the trigeminal nerve. Among the brain chemicals involved are Substance P (a well-known pain
mediator), nitric oxide, glutamate, and calcitonin
gene-related peptide. Magnesium deficiency may
play a role in migraine. There are two distinct
types of migraines: migraines with aura (a period
of visual disturbance that precedes the migraine)
and those without. This further complicates our
understanding of the disease. Among the processes involved is “cortical spreading depression,”
a wave of abnormal activity that passes through
the visual cortex and is thought to be the cause of
migraine aura. The etiology of migraines is therefore complex, and one would consequently expect
that multiple interventions might be necessary to
manage them.
The management of migraines includes several
medications that are used to treat acute migraine
pain, including nonsteroidal anti-inflammatory drugs (NSAIDs), aspirin, acetaminophen,
and caffeine. All of these have undesirable side
effects. Triptans can prevent migraines but should
be avoided by patients with vascular disease or
uncontrolled hypertension (Gilmore & Michael,
2011). Patients may also want to avoid medications due to pregnancy or other concerns with
medication side effects, so biofeedback and other
nonpharmacological interventions remain of
great interest.
Mechanisms of Migraine
Models for the etiology of migraine are diverse. Taylor (2011) presents migraine as an inflammatory
pain syndrome, at least in part. Both neurogenic
and vascular processes are thought to contribute
to migraine (Shevel, 2011). Release of serotonin
or other brain chemicals may cause vasoconstriction and vasodilation, leading to further release of
163
164
Prevalence of Migraine Triggers
Researchers have surveyed headache patients to
determine how many of them experience migraines
triggered by diet and other factors. In a study in
Italy, one-third of headache patients reported susceptibility to various foods (Savi, Rainero, Valfrè, Gentile, Lo Giudice, & Pinessi, 2002). Both
patients with tension headaches and those with
migraines reported sensitivity to chocolate, alcoholic drinks, and cheeses. On the other hand, in
a study in Denmark, Hauge, Kirchmann, and Olesen (2011) found that red wine was the only food
trigger listed by a group of patients with migraines
with aura, although several lifestyle factors were
listed. In reviewing this literature, Wöber and
Wöber-Bingöl (2010) comment that nearly every
aspect of life has been named as a migraine trigger.
They emphasize not only alcohol and withdrawal
from caffeine as triggers but also skipping meals
and possibly dehydration. Based on a survey of 120
headache patients, Wöber, Holzhammer, Zeitlhofer, Wessely, & Wöber-Bingöl, (2006) conclude
that the great majority of triggers occur occasionally and inconsistently. The major triggers related
to diet include allergenic foods, vasoactive amines,
glutamates, specific food items (e.g., hot dogs and
ice cream), alcohol, and caffeine.
Migraine Diets
Several proposed diets have aimed to remove possible migraine triggers from nutritional intake. The
triggers investigated include allergens, vasoactive
amines, caffeine, particular foods such as ice cream
and hot dogs, and others. In a 50-patient randomized study, Zencirci (2010) attempted to remove
multiple types of triggers. In this study, patients
in Turkey were given a diet excluding the major
types of foods thought to trigger migraines (caffeine sources, alcohol, cheeses, processed meats,
certain beans, monosodium glutamate [MSG] and
other food additives, etc.), along with a medication
and supplement regimen (metropolol, vitamin B2,
and naproxen at the start of headaches). A comparison group received only the medication plus
supplements. Monthly numbers of migraines were
similar in the two groups at the start, but by the
end of the study, patients in the diet group had
2.7 migraines per month, while those in the medication-only group had 5.15 migraines per month,
a significant difference. Monthly analgesic intake
also decreased by about half in the diet group. The
III. ADJUNCTIVE/COMPLEMENTARY INTERVENTIONS
presentation of results in this study was rather
abbreviated, and there were no data on whether
patients in the two groups adhered to their allocated programs, making the evaluation of these
intriguing results difficult. The lack of dietary
adherence data also applies to the other studies of
whole-diet interventions in migraine and reduces
the credibility of these studies. Additionally, most
of these studies involve rather small numbers
of patients and should be replicated with larger
groups.
Allergenic Foods
There were no controlled studies showing that
vascular headaches represent an allergic reaction
until the work of Egger, Wilson, Carter, Turner,
and Soothill (1983) supported an allergic pathogenesis for migraines among many children ages
3–16. Children were put on a diet low in foods
commonly known to be allergenic for 2 weeks.
Of the 88 children who completed the diet, 78%
reported full recovery from migraines. Forty of the
children then underwent double-blind challenges
at weekly intervals with allergenic foods. Of these
40, 39% reported migraines from cow’s milk, 31%
from wheat, 36% from eggs, and 17% from corn—
all foods commonly found to be highly allergenic.
Adverse reactions, however, were idiosyncratic for
each child. Other commonly allergenic foods found
to trigger migraines were orange, tomato, rye, fish,
and soy. Some foods suspected to contain vasoactive substances also caused migraines, including
chocolate, cheese, coffee, and malt (Egger, 1991).
This study lacked controls in the first phase, and
this work has been been criticized for other issues
in experimental design, although the doubleblind challenges lend some further credibility to
the results. A systematic review (Damen, Bruijn,
Koes, Berger, Passchier, & Verhagen, 2006) found
conflicting evidence concerning diets excluding
allergens for pediatric migraines, while behavioral
therapies including relaxation and biofeedback
were effective. It is possible, however, that dietary
manipulations in adults may be more successful than those in children. Alpay, Ertas, Orhan,
Üstay, Lieners, and Baykan (2010) constructed
individualized provocation and elimination diets
based on immunoglobulin G (IgG) allergy testing
for 30 adults with migraine. These patients each
participated in a nonintervention run-in period of
6 weeks, an elimination diet period in which foods
to which they were allergic were omitted, and a
165
10. Dietary Considerations
provocation period in which the allergenic foods
were included. Diets had the same number of calories and were nutritionally adequate. Significant
reductions in migraine days (from 10.0 to 7.5 days)
and number of attacks were seen in the elimination period (9.0 to 6.2). This crossover trial represented a stronger design than earlier trials. Like
other trials, it lacked adherence data but at least
suggested a stronger rationale for following a lowallergen diet.
Vasoactive Amines and Related Compounds
“Amines” are chemical substances that result from
the breakdown of proteins. A number of amines
are suspected of acting on blood vessels to produce migraines (termed “vasoactive” or “biogenic
amines”), and foods containing these amines
are often avoided by migraine patients. Biogenic
amines have been studied by administering the
amines in capsules, so that they can be examined in double-blind, placebo-controlled trials.
However, Jansen, van Dusseldorp, Bottema, and
Dubois (2003) reviewed the literature on welldesigned trials of biogenic amines and found that
reliable scientific evidence for the effects of these
compounds on migraines and other conditions is
lacking. There were two conclusive trials of tyramine that showed no association with migraine.
Phenylethylamine, a constituent of chocolate,
has also been suspected of causing migraines,
but Jansen et al. found no relationship between
the amount of phenylethylamine in chocolate
and headache. Curiously, cocoa was found to
decrease expression of inflammatory molecules in
the trigeminal nerve, which would argue against
the migraine-triggering properties of chocolate (Cady & Durham, 2010). Holzhammer and
Wöber (2006), however, also doubt the ability of
amines to cause migraines. More recently, in Italy,
D’Andrea, Nordera, Perini, Allais, and Granella
(2007) discussed the potential for tyramine, octopamine, synephrine, and related compounds to act
as migraine and cluster headache triggers based on
the discovery of trace amine-associated receptors
in the brain (D’Andrea, Nordera, Perini, Allais, &
Granella, 2007). They recorded elevated concentrations of such amines in the blood platelets of
migraine patients (D’Andrea et al., 2006). MSG
and other glutamates have also been suspected of
triggering migraines; however, Freeman (2006)
reviewed the literature on MSG effects and did not
find any consistent evidence that it produces head-
aches or other symptoms. It is worth noting that by
using a double-blind n-of-1 trial, a migraine patient
validated occurrence of migraines due to gelatin
(hydrolyzed protein) capsules used to administer
potential triggers in double-blind studies (Strong,
2000), casting some doubt on the negative findings reviewed earlier. Given the complexity of
migraine’s pathophysiology and this uncertainty,
it is possible that some patients do genuinely react
to tyramine, the most widely distributed of the
biogenic amines, or to other compounds such as
MSG. Table 10.1 shows tyramine contents of a
variety of foods, mostly cheeses. MSG is common
in many processed foods, under a variety of names,
including hydrolyzed vegetable protein (HVP),
natural flavor, flavoring, and kombu extract (see
Block, Schwartz, & Gyllenhaal, 2003, for more
information on MSG, HVP, and other substances,
as well as older studies involving whole-diet interventions).
Specific Foods: Ice Cream and Hot
Dog Headaches
In an experimental protocol devised to study icecream headaches, Selekler, Erdogan, and Budak
(2004; Selekler & Budak, 2004) found that a cold
stimulus could produce headaches, with migraine
patients more likely to report symptoms than tension headache patients. This was attributed to
irritation of the trigeminal nerve due to cold, and
deficits in central pain control. Hot dog headache
may be a reaction to sodium nitrite, and ingestion
of nitrite by persons sensitive to it may result in
headaches. Henderson and Raskin (1972) gave
10 milligrams of sodium nitrite or placebo to one
patient with migraine. Headache occurred in eight
of 13 trials of sodium nitrite, and no headaches
occurred after placebo. Ten healthy volunteers had
no headaches after nitrite or placebo. Given the
possible role of nitric oxide in migraine pathophysiology, some individuals may be sensitive to hot
dogs and other nitrite-containing processed meats.
Alcohol
In a systematic review of the role of alcohol in
triggering migraines, Panconesi (2008) concluded
that while about one-third of migraine patients
report alcohol as an occasional headache trigger
in retrospective studies, only 10% report that it
frequently triggers migraine, with red wine being
166
III. ADJUNCTIVE/COMPLEMENTARY INTERVENTIONS
TABLE 10.1. Tyramine-Containing Foods, Beverages, and Condiments, and Tyramine Concentrations
(in Micrograms per Gram or Micrograms per Milliliter, Unless Otherwise Noted)
Cheeses
Cheddar (Highest: old, center-cut Canadian and New York State, approx. 1500; Canadian aged in ale, approx.
1000) (Others: English, up to 953; New Zealand, 471–580; Australian, 226; other Canadian, most 120–192)
Gruyère (American, 516; British, 11–1184; Finnish, 102)
Stilton (American, 466; English, 2170)
Emmenthal (225); Emmentaler (225–1000)
Brie (American, 180; Danish, nil)
Camembert (American, 86; Danish, 23–1340; Mycella [Camembert type], 1304; Cuban, 34–425)
Roquefort (French, 27–520)
Blue (Danish, 31–256; French, 203; Bourmandise [blue type], 216)
Boursault (French, 1116)
Parmesan (Italian, 65; American, 4–290)
Processed (American, 50; Canadian, 26)
Romano (Italian, 238)
Provolone (Italian, 38)
Cracker Barrel (Kraft brand, American, 214)
Brick (natural Canadian, 524)
Mozzarella (Canadian, 410)
Gouda (Canadian, 20; Cuban, 40–280)
Cream cheese and cottage cheese (nil except South African cottage, 6.6)
Cuban (Fontina, 54–167; Broodkaase, 0–163; Dambom, 26–100; Carré, 52–200)
Spanish (Mahon old, 369; Cáceres cured, 225; Cáceres, 102; Malaga fresh, 22)
Dry/fermented sausage
Salami (hard, average 210, up to 392; farmer, average 314; Genoa, average 534, up to 1237) Sausage (summer, up to
184; dry, up to 244; semidry, up to 85.5)
Others (pepperoni, average, up to 195; smoked landjaeger, up to 396; dry fermented, 102–1506)
Others
Herring (Marinated [pickled], 3030; Canadian, 470)
Caviar (estimated high, but no published analysis found)
Sour cream, yogurt (variable but often nil, especially from reputable brands)
Chicken liver (cooked, not kept refrigerated, 94–113; fresh cooked, nil)
Beef liver (fresh or frozen, approx. 5)
Raspberries, fresh (13–92)
Yeast products (English: unspecified brand, 2100; Marmite, 1087–1639; Yex, 506; Befit, 419; Barmene, 152; Yeastrel,
101) (Canadian: unspecified brand, 66–84; plain, nil)
Bananas (peel, 63–65; pulp, 7)
Soya beans (fermented, Singapore, 713 [50 q]; Taiwan, 878 [10 ml]; Swiss soya sauce, Dr. Dunners, 293 [10 ml],
Formosa soya bean condiment, 939 [20 g]; Korea soya bean paste fermented, 206 [50 g] a
Red wine (highly variable, some nil)
Note. For a patient-friendly summary of avoiding high-tyramine foods, see www.mc.vanderbilt.edu/documents/neurology/files/Tyramine%20Menu%20Book%2006227101.pdf. See also Block et al. (2003, Table 9.1) for many other foods and beverages, most of which
have very small amounts of tyramine. All tyramine < 1 mg/portion unless specified. Portions of most meat/vegetables, 50–75 g
(1.8–2.6 oz.). Normal portions of cheese are variable. As a guide, Vidaud, Chaviano, Gonzales, and Garcia Roche (1987) note that a
normal portion of Camembert cheese is about 23 g (0.8 oz., avoirdupois) and of Gouda, about 30 g (1.06 oz.).
The data on concentrations are from the following sources: Maxwell (1980), citing the concentrations as reported by Horowitz,
Lovenberg, Engelman, and Sjoerdsma (1964), McCabe (1986), Sen (1969), Boulton, Cookson, and Paulten (1970), Coffin (1970),
Hedberg, Gordon, and Glueck (1966), Marley and Blackwell (1970), Orlosky (1982), Udenfriend, Lovenberg, and Sjoerdsma (1959);
Cuban and Spanish data from Vidaud, Chaviano, Gonzales, and Garcia Roche (1987), Rice, Eitenmiller, and Kohler (1975, 1976),
Rice and Kohler (1976), Rivas-Gonzalo, Santos-Hernandez, and Marine-Font (1983).
Vidaud et al. (1987) reviewed eight studies with ranges of 13–2000 milligrams per kilogram. All but one study (Asatoor, Levi, &
Milne, 1963) reported average tyramine content of less than 211.
aThe amounts given here are those consumed in a normal serving.
167
10. Dietary Considerations
the most common type of alcohol selected. Prospective studies, which are more reliable, cast
doubt on the hypothesis that alcohol is an important trigger, at least in adults. Hauge et al. (2011),
in their survey study, reported that no migraine
patients claimed that alcohol triggered more than
50% of their migraine attacks. Migraine patients
in general consume less alcohol than other populations. Milde-Busch et al. (2010) found, in a
cross-sectional study of over 1200 adolescents,
that high consumption of cocktails was associated
with both migraine and tension headaches. While
such a study design is less reliable than prospective
studies, this suggests that refraining from alcohol
may be a useful suggestion for adolescents (and, of
course, it is health-promoting for other reasons).
Caffeine1
Rationale
Caffeine is the world’s most popular drug and the
focus of considerable attention by applied psychophysiology and health care professionals. One
major rationale for discussing caffeine in this chapter is that as a powerful psychotropic stimulant, it
elicits or aggravates physiological symptoms associated with many disorders: migraine headaches,
anxiety, Raynaud’s disease, nausea and other gastrointestinal symptoms (irritable bowel syndrome
[IBS]), hypertension, premenstrual syndrome, hot
flashes, urinary incontinence, sleep-onset insomnia, and conditions associated with breathing
problems, increased skeletal muscle contractions,
and/or increased heart rate. Another major concern is that caffeine can interfere with developing
and applying psychophysiological self-regulation.
Thus, caffeine use is inconsistent with the overall goals of reducing sympathetic and general and/
or focal muscle tension. The effects of withdrawal
constitute another cause for concern and rationale
for including caffeine here.
Metabolism and Toxicity
In usual doses, caffeine stimulates the central nervous system. The effects vary but often results in
excess stimulation and tension. Orally consumed
caffeine spreads rapidly through the body. Body
tissues and fluids quickly absorb it, resulting in
peak plasma levels usually from about 15 minutes
to within an hour, but longer, up to 2 hours, in
some cases (Arnaud, 1993).
Half-life estimates vary from about 2 to 12 hours
in healthy, nonmedicated adults, with an average
typically in the 4- to 6-hour range (Grosso &
Bracken, 2005). Half-life is much longer in pregnant women after the first trimester, especially
later stages, about 9–11 hours at about 17 weeks
of gestation and longer, about 11–18 hours, by the
end of pregnancy (Aldridge, Bailey, & Neims,
1981; Carrillo & Benitez, 2000). Among people
with chronic liver disease the half-life of caffeine
is much longer and up to 96 hours with severe
liver disease (Carrillo & Benitez, 2000). Smoking
nearly doubles the metabolic rate, thus shortening
the half-life of caffeine. It does this via a hydrocarbon that increases liver enzyme activity (Kalow &
Tang, 1991; Parsons & Neims, 1978).
Interactions with many medications and
dietary chemicals affect the metabolism of caffeine and hence the half-life elimination of caffeine. We suggest reading Carrillo and Benitez
(2000) and Culm-Merdek, von Moltke, Harmatz,
and Greenblatt (2005). Among the many interactions, note that fluvoxamine (Luvox®), a selective serotonin reuptake inhibitor (SSRI), lengthens the half-life to an estimated 56 hours (95%
confidence interval [CI] 26–76 hours; CulmMerdek et al., 2005).
Eating certain vegetables, such as dill weed, celery, parsley, parsnips, and carrots (members of the
Apiaceae or carrot family), slows caffeine metabolism and therefore increases half-life. Eating other
vegetables, especially radish sprouts, broccoli,
cauliflower, and cabbage (part of the Cruciferae
or cabbage family), increases metabolism of caffeine and therefore decreases half-life (Grosso &
Bracken, 2005; Lampe et al., 2000).
Caffeine Effects
Headaches
The role of caffeine in migraine and tension
headaches is paradoxical. Consuming caffeine
causes vasoconstriction: It can therefore be useful in stopping headaches, and it is a constituent of several acute migraine medications. This
vasoconstriction, however, is later followed by a
rebound vasodilation, which may also cause headache. Researchers have recently become interested
in the role of caffeine in chronic daily headache
or chronic migraine, and in medication overuse
headache (Aguggia & Saracco, 2010). We discuss
these topics below.
168
Anxiety
Moderate caffeine intake may actually be useful
in some psychiatric disorders, having been associated with fewer depressive symptoms, lower suicide
risk and less cognitive failure due to its stimulant
and mood-elevating effects (Lara 2010). Anxiety
in response to caffeine is associated with a polymorphism in the adenosine A2A receptor; thus,
some persons are genetically more susceptible to
caffeine effects than others. Caffeine-induced
anxiety is especially noted in people with panic
disorder and social anxiety disorder (Nardi et al.,
2009). However, in many people, even those with
susceptible genotypes, tolerance to the anxiogenic
effects of caffeine occurs when they take caffeine
regularly (Rogers et al., 2010).
Blood Pressure
Caffeine doses of 250–500 milligrams administered acutely can raise systolic and diastolic
blood pressure significantly by a few to several
millimeters of mercury (Lane, Phillips-Bute, &
Pieper, 1998), independent of posture, activity, or
perceived stress. However, with chronic intake of
caffeine, the effect on blood pressure is attenuated
(Geleijnse, 2008). Caffeine intake in general does
not raise risks of cardiovascular disease, in part
because coffee appears to protect against development of Type 2 diabetes, possibly due to chlorogenic acid content (Riksen, Rongen, & Smits,
2009). Coffee intake does appear to be associated
with an increased risk of elevated blood pressure
in older men who are overweight or obese (Giggey,
Wendell, Zonderman, & Waldstein, 2011).
Insomnia
Caffeine management (e.g., elimination, substantial reduction, or at least limited use only in the
early morning), is commonly recommended for all
or most persons with sleep-onset insomnia (e.g.,
Juliano & Griffith, 2005). Caffeine is an adenosine receptor antagonist that is thought to be a factor in interfering with sleep onset, as measured by
subjective sleepiness and electroencephalographic
(EEG) theta activity (Landolt et al., 2004). Even
individuals accustomed to caffeine may experience sleep problems from it. It is notable that the
DSM-5 includes a diagnosis of caffeine-induced
sleep disorder (American Psychiatric Association, 2013). Possible genetic factors are the focus
of research such as that by Luciano et al. (2007),
III. ADJUNCTIVE/COMPLEMENTARY INTERVENTIONS
who concluded that the “heritability of coffeeattributed sleep disturbance . . . was approximately
0.40, with three fourths of this genetic variance
explained by genes unrelated to the general sleep
disturbance factor” (p. 1378). This can partly
explain the individual variations of effects of caffeine on sleep-onset insomnia and probably other
symptoms. However, practitioners do not usually
have genetic information available (yet); thus, the
standard recommendations to manage caffeine
intake are still prudent.
Urinary Incontinence
Approximately one-fourth of cases of urge incontinence (although not stress or mixed incontinence)
appeared to be related to high (> 450 mg daily) caffeine intake in a recent prospective survey (Jura,
Townsend, Curhan, Resnick, & Grodstein, 2011).
Caffeine reduction is standard advice in management of urinary problems. A randomized trial in
patients presenting for urinary problems found significant reductions in caffeine intake after counseling, and subsequently in urgency and frequency
outcomes (Bryant, Dowell, & Fairbrother, 2002),
validating this advice.
Caffeine Contents of Beverages and Foods
There is much variability in caffeine amounts
measured in the many published studies of caffeine in coffee and tea (Bracken et al., 2002).
As these investigators also note, the variability
occurs even in the same participants and within
the same day and the same brewing conditions
when they were brewing coffee or tea. The caffeine content in coffee depends on the strain of
coffee bean, the condition of the beans (whether
they are green or roasted), and the type of coffee
(e.g., drip, instant, espresso), brewing time,2 exposure surface area of the tea or coffee to the water
(i.e., smaller tea leaves and finer coffee grounds
result in more release of caffeine), and the ratio of
ground coffee to water (see http://blackbearcoffee.
com/resources/83 for standard and connoisseur's
ratios). In view of the extensive data for caffeine
contents that are available very readily on the
Internet, the data and Internet links presented in
Table 10.2 provide estimates of caffeine content for
a few types of beverages.
A practitioner need not know the exact amount
of caffeine in a specific patient’s coffee. It does not
pay to invest expensive professional time exploring the strain of coffee beans and brewing time
169
10. Dietary Considerations
TABLE 10.2. Caffeine Contents (in Milligrams
per Ounce)
Coffee
Generic
Drip
Instant
Espresso
Coffee shopsa
Expresso
McDonald’s
13–25 (est. 20)
3–22 (est. 12)
30–90
9–20
75
9
Teas
Black more than green, instant,
and premixed
Black
Green
Iced (e.g., Lipton)
3–17
8–15 (est. 10)
1–4
≤2
Soft drinksb /carbonated beverages
Most
“Energy” drinks (most)
“Energy” shots
2–6 (est. 3 or 4)
9–50
40–50
Internet links
www.cspinet.org/new/cafchart.htm
www.mayoclinic.com/health/caffeine/an01211
www.energyfiend.com/the-caffeine-database
Note. Content is based on most values given on websites.
aFor example, Starbucks, Einstein, Dunkin’Donuts (except
espresso).
bFDA official limit for cola and pepper soft drinks = 71 mg in
12 oz., 6 mg/oz.
used by the patient. One can estimate the caffeine
intake from coffee from the general type (e.g.,
instant, drip) and the number of ounces drunk.
One should also inquire about the size of the cup
used, because most people use the term “cup”
when they are actually drinking from larger-size
mugs. A mug can vary from about 5 to 12 ounces,
or even more.
Caffeine Withdrawal
The factor that motivates a substantial percentage
of habitual caffeine consumers is the need to avoid
withdrawal symptoms, particularly headaches, created by physical dependence on caffeine (Shapiro,
2008). The comprehensive review and literature
analysis by Juliano and Griffith (2005) regarding
caffeine withdrawal provided an empirical vali-
dation of specific symptoms and signs (see Table
10.3) and an appraisal of the important features
of caffeine withdrawal syndrome. Headaches were,
as expected, the most common symptom, with
an incidence of 50%. In general, regular use of
as little as 100 milligrams of caffeine per day can
induce a form of physical dependence, interruption
of which elicits withdrawal symptoms (Juliano &
Griffith, 2005).
Also note the following selected symptoms
noted by Block et al. (2003): increased heart rate,
analgesic use, cerebral blood flow volume, EEG
theta activity primarily in the occipital area, muscle pain/stiffness, and irritability.
Headaches that occur after general anesthesia
are often related to preoperative caffeine withdrawal after moderate daily intake of caffeine.
(Fennelly, Galletly, & Purdie, 1991). Furthermore,
the probability of headaches developing preoperatively and postoperatively increases with consumption of more caffeine preoperatively (e.g.,
increases of 100 milligrams increased headaches
16% postoperatively).
A patient’s medication alone can contain
enough caffeine to create or add to the problem.
Even some migraine medications contain caffeine.
Table 10.4 presents the caffeine content in several
prescription and over-the-counter (OTC) preparations.
TABLE 10.3. Caffeine Withdrawal Symptoms
from a Critical Review by Juliano and Griffiths
(2005)
10 caffeine withdrawal symptoms fulfilling validity
criteria
•• Headache
•• Fatigue
•• Decreased energy/activeness
•• Decreased alertness
•• Drowsiness
•• Decreased contentedness
•• Depressed mood
•• Difficulty concentrating
•• Irritability
•• Foggy/not clearheaded
Additional withdrawal symptoms judged likely to
represent valid categories
•• Flu-like symptoms
•• Nausea/vomiting
•• Muscle pain/stiffness
Note. Caffeine withdrawal symptoms typically start 12–24
hours after caffeine abstinence and reach maximum intensity
after 20–51 hours. Duration can be 2–9 days.
170
III. ADJUNCTIVE/COMPLEMENTARY INTERVENTIONS
TABLE 10.4. Caffeine Content in Milligrams
per dose of Selected Prescription and
Nonprescription Preparations (by Trade Names)
Prescription
Cafergot
Ergocaf
Esgic
Ezol
Fioricet
Fiorinal
Medigesic
Norgesic
Norgesic Forte
Orphengesic
Orphengesic
Forte
Repan tablets
Synalgos-DC
Tencet
Triad
Novartis
Rugby Labs
Forest; Gilbert
Stewart Jackson
Novartis
Novartis
U.S. Pharm
3M
3M
Various
Various
100
100
40
40
40
40
100
30
60
60
40
Everett
Wyeth
Roberts
Forest
32
30
40
40
Whitehall
Thompson Medical
Thompson Medical
Thompson Medical
Mentholatum
Chilton
Bristol-Myers
32
100
200
200
32
75
65
Bristol-Myers
Bristol-Myers
Vita Elixer
Bayer
65
65
250
32.4
Nonprescription
Anacin
Aqua Ban
Aqua Ban Plus
Caffedrine
Cope
Energets
Excedrin Extra
strength
ASA free
Migraine
Lerton
Midol Original
(cramps)
No Doz
Quick-Pep
Stay-Alert
Stay Awake
Summit Extra
strength
Vanquish
Vivarin
Wakespan
Squibb
Thompson Medical
Apothecary
Whiteworth
Pfeiffer
Bayer
Smith, Kline,
Beecham
Weeks & Lea
100 and 200
150
200
250
65
200
200
250
Caffeine is also found in several of the popular herbal medicines and other dietary supplements for weight control, muscular development,
and increased feelings of energy. The herbs that
contain caffeine are kola nut, guarana, and yerba
maté; green tea extracts are also sold as supplements, with and some without caffeine. “Energy
drink” supplements in particular may contain high
levels of caffeine.
Thus, we recommend that practitioners:
• Ask about OTC and prescription drugs, dietary
supplements, and herbal preparations to estimate caffeine intake from these sources.
• Provide information about the potential for caffeine withdrawal symptoms.
• Recommend significantly reducing or often
stopping caffeine consumption.
• Provide guidance regarding gradual reduction of
caffeine consumption.
Website Link
www.encyclopedia.com/doc/1g2-3403100095.html
International Differences in Caffeine Content
and Selected Implications
International practitioners should note that beverages, especially coffee, vary in their caffeine content in different countries. This has implications
also for athletes and applications of biofeedback to
athletic performances. Athletes also vary in their
metabolism and caffeine excretion, and may therefore vary, and exceed, regulated amounts of caffeine allowed in competition. This can occur even
with only about 3–6 cups of brewed coffee of average strength. This excess is more likely to occur
if the urine sample is obtained at the time that
caffeine concentration is highest. Further contributing to excess caffeine in the body of an athlete
or patient is the interactions of medications with
caffeine (Carillo & Benitez, 2000).
Meal Patterns and Headache
Missing meals has commonly been recognized as
a trigger for migraines. This is corroborated by the
observation of more than double the migraines in
Muslim patients fasting during Ramadan (AbuSalameh, Plakht, & Ifergane, 2010). In addition,
Blau (2005) observed that one-third of migraine
patients reported that water deprivation can trigger migraines. A small randomized trial of advice
to increase water intake demonstrated a significant decrease in migraines, a result that needs further validation (Spigt et al., 2005).
Other Risk Factors for Migraine
Recent research has begun to investigate the role
of dietary elements in two types of intensification of headache: medication overuse headache
171
10. Dietary Considerations
and chronic migraine or chronic daily headache.
Chronic migraine and chronic daily headache
are characterized by headache at least 15 days per
month.
Medication overuse headache can arise in
response to continuing use of various medications,
including triptans, ergot alkaloids, barbiturates,
and caffeine. Migraine patients are more likely
than others to develop this type of chronic headache, which appears to be related to increases in
calcitonin gene-related peptide or cortical spreading depression that may be triggered by these
medications (Meng, Dodick, Ossipov, & Porreca,
2011). Risk factors for the transformation of episodic to chronic migraine include excessive medication, stress, sleep disorders, caffeine overuse, and
obesity (Bigal & Lipton, 2009).
The understanding that inflammation is
important in the pathophysiology of migraine suggests a new dimension to dietary considerations.
Migraine patients have higher levels than healthy
controls of the biomarker C-reactive protein,
which indicates chronic inflammation, as well
as other biomarkers of endothelial (blood vessel)
dysfunction (Tietjen et al., 2009). NSAIDs can
be used to control inflammation, but they carry
a high risk of gastric ulcers, so safer interventions
are preferable. A trial of omega-3 fatty acid supplementation (the anti-inflammatory constituents
of fish oil) in migraine patients did not reduce
migraines (Pradalier et al., 2001), but the effects
of these could have been counteracted if subjects’
diets were high in omega-6 fatty acids, which are
common in Western diets. A Mediterranean diet
supplemented with olive oil or nuts, however,
was able to reduce C-reactive protein and other
inflammatory biomarkers (Estruch, 2010).
Growing evidence also indicates that migraine
is related to disorders of glucose metabolism,
including insulin resistance, the inability of insulin to deliver glucose to body tissues that is seen
in Type 2 diabetes and prediabetes. Migraine
patients were observed to be much more likely
than normal controls to have signs of insulin
resistance, and these were associated with indications of abnormalities with nitric oxide—one of
the mediators of migraine (Gruber et al., 2010).
Migraine patients have a higher than expected
prevalence of metabolic syndrome, which is associated with insulin resistance (Guldiken et al.,
2009). Both insulin resistance and inflammation place patients at elevated risk of cardiovascular disease, and migraine patients are indeed at
higher than average risk for stroke. Insulin resis-
tance can be treated with diet, and a variety of
diets have been found effective in reducing insulin resistance, including vegetarian (Kahleova et
al., 2011), high-fiber (Li et al., 2010), and other
diets. Loss of body fat through diet or exercise is
important in reducing insulin resistance (Camhi,
Stefanick, Katzmarzyk, & Young, 2010).
Therapeutic Strategies
Different therapeutic strategies with regard to diet
can be discussed with patients who have episodic
migraines, medication overuse headaches, and
chronic daily migraine. Table 10.5 summarizes
dietary interventions, which are listed in order of
aggressiveness.
Practitioners can initially approach treatment of
patients with episodic migraine using techniques
of psychophysiological self-regulation. Because of
the questionable evidence for the prevalence of
dietary triggers of migraine, patients can be asked
to avoid triggers that they already know are bothersome for them, as well as caffeine and alcohol.
Issues relative to adequate hydration and regularity of meals or fasting can also be addressed. In
regard to general health, it would be prudent to ask
patients whether they have talked to their physicians about cardiovascular risks, and to advise
them to consult a dietitian to develop a dietary
plan that will decrease cardiovascular risks and
either reduce or prevent overweight. With adolescents experiencing migraine or other headaches,
TABLE 10.5. Dietary Interventions for Migraine
•• Avoid foods patient suspects as migraine triggers.
•• Avoid alcohol, especially for adolescents.
•• Gradually reduce and/or eliminate caffeine.
•• Remain well hydrated.
•• Avoid cold foods and specific items such as hot dogs/
processed meats.
•• Eat meals regularly, and approach fasting with
caution.
•• Address overweight (consult with dietitian).
•• Address inflammation and insulin resistance
(consult with integrative health professional or
dietitian).
•• Explore use of dietary supplements (consult with
integrative health professional).
•• Reduce vasoactive amine intake and intake of other
chemicals such as MSG (consult with dietitian).
•• Allergenic elimination diet (consult with dietitian).
Note. Least aggressive interventions are listed first.
172
particular attention to caffeine and alcohol consumption is recommended.
For patients with possible medication overuse
headache, or patients who consume excessive
caffeine-containing beverages or medications
daily, a more assertive dietary intervention is
wise, in addition to instruction in psychophysical self-regulation techniques. Using the tables in
this chapter, the practitioner can help the patient
determine dietary and medication sources of caffeine, then encourage slow reduction or elimination of caffeine. Reducing caffeine intake gradually prevents caffeine withdrawal symptoms. This
strategy can also be used with patients who have
other problems related to caffeine intake. With
migraine patients specifically, it is important to
institute nondrug therapies, so that medication
overuse headache or excessive caffeine use does
not lead to chronic migraine. Addition of counseling about dehydration, fasting, and known triggers
may assist in establishing nondrug control. If the
combined efforts are unsuccessful, consultation
with other health professionals to address possible
overweight, inflammation, and insulin resistance
may be recommended. Internet sites offer calculation of body mass index to determine overweight,
and physicians may be consulted about laboratory
tests that indicate inflammation or insulin resistance. A dietitian can assist with interventions for
weight reduction or laboratory-confirmed abnormalities. It is also recommended that the patient
seek information about dietary supplements that
show evidence of reducing migraines, including
magnesium, butterbur, feverfew, coenzyme Q10
and alpha-lipoic acid (Sun-Edelstein & Mauskop,
2009); an integrative medical doctor, pharmacist,
naturopath, or knowledgeable dietitian should be
consulted about these, and they are more likely
to be effective when used with other lifestyle
changes.
For patients with chronic daily headache, a vigorous intervention may be implemented, based on
psychophysiological self-regulation, exploration of
caffeine overuse, and elimination of known triggers (e.g., alcohol, dehydration, and fasting). The
patient should also be referred to a dietitian to
address overweight, inflammation, and insulin
resistance as soon as possible (the practitioner may
consider consulting with the dietitian to explain
the significance of these conditions for migraine).
Dietary supplements can be considered. If these
efforts do not begin to ameliorate headache within
a few months, the practitioner could consider helping the patient to analyze his or her diet for other
III. ADJUNCTIVE/COMPLEMENTARY INTERVENTIONS
migraine triggers, such as tyramine, other amines
and related compounds, allergens (consult with
a dietitian), and processed meats, and encourage
elimination of these on a trial basis. A full elimination diet could be attempted with support from
a dietitian. It is interesting to note that many of
the foods that are pointed out as migraine triggers
(e.g., cheese, processed meats, chocolate candy,
dairy products, fatty foods and food additives; as
listed by Zencirci, 2010) are foods that contribute
to inflammation, insulin resistance, and weight
gain.
The lack of solid high-level evidence for many
of the foods commonly considered migraine triggers, and the emerging data on effects of dietary
patterns on chronification of migraine suggest
that a new approach to the understanding of diet
and migraine is needed and appears to be evolving. Certainly, better-controlled research studies
involving larger populations are needed to evaluate dietary interventions. Along with well-validated techniques of psychophysical self-regulation,
however, these interventions offer the potential
to assist many patients experiencing migraines
to reduce the incidence or severity of headaches.
These nondrug methods also may help many
patients avoid the complications of medication
overuse headaches and chronic daily headache.
Notes
1. There are numerous credible and comprehensive
websites that include or focus on caffeine. This readily
available and extensive information has reduced the need
and justification for including much content on caffeine
in this chapter. Rather, we provide a summary of selected
information and links to the websites. See www.marksschwartzphd.com for the hyperlinks (e.g., http://emedicine.
medscape.com/article/821863-overview).
2. Longer brewing (e.g., 10 vs. 5 minutes) can increase
caffeine by about 4–15 milligrams per 150 milliliters.
References
Abu-Salameh, I., Plakht, Y., & Ifergane, G. (2010).
Migraine exacerbation during Ramadan fasting. Journal
of Headache and Pain, 11, 513–517.
Aguggia, M., & Saracco, M. G. (2010). Pathophysiology of migraine chronification. Neurological Sciences,
31(Suppl. 1), S15–S17.
Aldridge, A., Bailey, J., & Neims, A. H. (2008). The disposition of caffeine during and after pregnancy. Seminars in
Perinatology, 5(4), 310–314.
Alpay, K., Ertas, M., Orhan, E. K., Üstay, D. K., Lieners, C.,
10. Dietary Considerations
& Baykan, B. (2010). Diet restriction in migraine, based
on IgG against foods: A clinical, double-blind, randomized, crossover trial. Cephalagia, 30, 829–837.
American Psychiatric Association. (2013). Diagnostic and
statistical manual of mental disorders (5th ed.). Arlington
VA; Author.
Arnaud, M. J. (1993). Metabolism of caffeine and other
components of coffee. In S. Garattini (Ed.), Caffeine,
coffee and health (pp. 43–95). New York: Raven Press.
Asatoor, A. M., Levi, A. J., & Milne, M. P. (1963). Trancylpromine and cheese. Lancet, 2, 733–734.
Bigal, M. E., & Lipton, R. B. (2009). What predicts the
change from episodic to chronic migraine? Current
Opinion in Neurology, 22, 269–276.
Blau, J. N. (2005). Water deprivation: A new migraine precipitant. Headache, 45, 757–759.
Block, K. I., Schwartz, M. S., & Gyllenhaal, C. (2003).
Dietary considerations: Rationale, issues, substances,
evaluation, and patient education. In M. S. Schwartz
& F. Andrasik (Eds.), Biofeedback: A practitioner’s guide
(3rd ed., pp. 181–211). New York: Guilford Press.
Boulton, A. A., Cookson, B., & Paulten, R. (1970). Hypertensive crisis in a patient on MAOI antidepressants following a meal of beef liver. Canadian Medical Association Journal, 102, 1394–1395.
Bracken, M., Triche, E., Grosso, L., Hellenbrand, K.,
Belanger, K., & Leaderer, B. (2002). Heterogeneity in
assessing self-reports of caffeine exposure: Implications
for studies of health effects. Epidemiology, 13, 165–171.
Bryant, C. M., Dowell, C. J., & Fairbrother, G. (2002). Caffeine reduction education to improve urinary symptoms.
British Journal of Nursing, 11, 560–565.
Cady, R. J., & Durham, P. L. (2010). Cocoa-enriched
diets enhance expression of phosphatases and decrease
expression of inflammatory molecules in trigeminal
ganglion neurons. Brain Research, 1323, 18–32.
Camhi, S. M., Stefanick, M. L., Katzmarzyk, P. T., & Young,
D. R. (2010). Metabolic syndrome and changes in body
fat from a low-fat diet and/or exercise randomized controlled trial. Obesity (Silver Spring), 18, 548–554.
Carillo, J. A., & Benitez, J. (2000). Clinically significant
pharmacokinetic interactions between dietary caffeine
and medications. Clinical Pharmacokinetics, 39, 127–153.
Coffin, D. C. (1970). Tyramine content of raspberries and
other fruit. Journal of the Association of Official Analytic
Chemists, 53, 1071–1073.
Culm-Merdek, K. E., von Moltke, L. L., Harmatz, J. S., &
Greenblatt, D. J. (2005). Fluvoxamine impairs singledose caffeine clearance without altering caffeine pharmacodynamics. British Journal of Clinical Pharmacology,
60(5), 486–493.
Damen, L., Bruijn, J., Koes, B. W., Berger, M. Y., Passchier,
J., & Verhagen A. P. (2006). Prophylactic treatment of
migraine in children. Part 1: A systematic review of nonpharmacological trials. Cephalalgia, 26, 373–383.
D'Andrea, G., Granella, F., Leone, M., Perini, F., Farruggio, A., & Bussone, G. (2006). Abnormal platelet trace
amine profiles in migraine with and without aura. Cephalalgia, 26, 968–972.
173
D'Andrea, G., Nordera, G. P., Perini, F., Allais, G., &
Granella, F. (2007). Biochemistry of neuromodulation
in primary headaches: Focus on anomalies of tyrosine
metabolism. Neurological Sciences, 28(Suppl. 2), S94–
S96.
Egger, J. (1991). Psychoneurological aspects of food allergy.
European Journal of Clinical Nutrition, 45(Suppl. 1),
35–45.
Egger, J., Wilson, J., Carter, C. M., Turner, M. W., &
Soothill, J. F. (1983). Is migraine food allergy?: A doubleblind controlled trial of oligoantigenic diet treatment.
Lancet, 2, 865–869.
Estruch, R. (2010). Anti-inflammatory effects of the Mediterranean diet: The experience of the PREDIMED
study. Proceedings of the Nutrition Society, 69, 333–340.
Fennelly, M., Galletly, D. C., & Purdie, G. I. (1991). Is caffeine withdrawal the mechanism of postoperative headache? Anesthesia and Analgesia, 72, 449–453.
Freeman, M. (2006). Reconsidering the effects of monosodium glutamate: A literature review. Journal of the American Academy of Nurse Practitioners, 18(10), 482–486.
Geleijnse, J. M. (2008). Habitual coffee consumption and
blood pressure: An epidemiological perspective. Vascular Health and Risk Management, 4, 963–970.
Giggey, P. P., Wendell, C. R., Zonderman, A. B., & Waldstein, S. R. (2011). Greater coffee intake in men is associated with steeper age-related increases in blood pressure. American Journal of Hypertension, 24, 310–315.
Gilmore, B., & Michael, M. (2011). Treatment of acute
migraine headache. American Family Physician, 83(3),
271–280.
Grosso, L. M., & Bracken, M. B. (2005). Caffeine metabolism, genetics, and perinatal outcomes: A review of
exposure assessment considerations during pregnancy.
Annals of Epidemiology, 15(6), 460–466.
Gruber, H. J., Bernecker, C., Pailer, S., Fauler, G., Horejsi, R., Möller, R., et al. (2010). Hyperinsulinaemia in
migraineurs is associated with nitric oxide stress. Cephalalgia, 30, 593–598.
Guldiken, B., Guldiken, S., Taskiran, B., Koc, G., Turgut,
N., Kabayel, L., et al. (2009). Migraine in metabolic syndrome. Neurologist, 15, 55–58.
Hauge, A. W., Kirchmann, M., & Olesen, J. (2011). Characterization of consistent triggers of migraine with aura.
Cephalalgia, 31, 416–438.
Hedberg, D. L., Gordon, M. W., & Glueck, B. C. (1966).
Six cases of hypertensive crisis in patients on tranylcypromine after eating chicken livers. American Journal of
Psychiatry, 122, 933–937.
Henderson, W. R., & Raskin, N. (1972). Hot dog headache:
Individual susceptibility to nitrite. Lancet, 2, 1162–1163.
Holzhammer, J., & Wöber, C. (2006). [Alimentary trigger
factors that provoke migraine and tension-type headache]. Schmerz (Berlin, Germany), 20, 151–159.
Horowitz, D., Lovenberg, W., Engelman, K., & Sjoerdsma,
A. (1964). Monoamine oxidase inhibitors, tyramine
and cheese. Journal of the American Medical Association,
188(13), 1108–1110.
Jansen, S. C., van Dusseldorp, M., Bottema, K. C., &
174
Dubois, A. E. (2003). Intolerance to dietary biogenic
amines: A review. Annals of Allergy, Asthma and Immunology, 91, 233–240.
Juliano, L. M., & Griffith, R. R. (2005). Caffeine. In J. H.
Lowinson, P. Ruiz, & R. B. Milliman (Eds.), Substance
abuse: A comprehensive textbook (pp. 403–421). New
York: Lippincott, Williams & Wilkins.
Jura, Y. H., Townsend, M. K., Curhan, G. C., Resnick, N.
M., & Grodstein, F. (2011). Caffeine intake, and the risk
of stress, urgency and mixed urinary incontinence. Journal of Urology, 188, 1775–1780.
Kahleova, H., Matoulek, M., Malinska, H., Oliyarnik, O.,
Kazdova, L., Neskudla, T., et al. (2011). Vegetarian diet
improves insulin resistance and oxidative stress markers more than conventional diet in subjects with Type 2
diabetes. Diabetic Medicine, 28, 549–559.
Kalow, W., & Tang, B. K. (1991). Caffeine as a metabolic
probe: Exploration of the enzyme-inducing effect of cigarette smoking. Clinical Pharmacology and Therapeutics,
49, 44–48.
Lampe J. W., King, I. B., Li, S., Grate, M. T., Barale, K.
V., Chen, C., et al. (2000). Brassica vegetables increase
and apiaceous vegetables decrease cytochrome P4501A2
activity in humans: Changes in caffeine metabolic
ratios in response to controlled vegetable diets. Carcinogenesis, 21(6), 1157–1162.
Landolt, H. P., Rétey, J. V., Tönz, K., Gottselig, J. M., Khatami, R., Buckelmüller, I., et al. (2004). Caffeine attenuates waking and sleep electroencephalographic markers
of sleep homeostasis in humans. Neuropsychopharmacology, 29, 1933–1939.
Lane, J. D., Phillips-Bute, B. G., & Pieper, C. F. (1998). Caffeine raises blood pressure at work. Psychsomatic Medicine, 60, 327–330.
Lara, D. R. (2010). Caffeine, mental health and psychiatric
disorders. Journal of Alzheimer’s Disease, 20(Suppl. 1),
S239–S248.
Li, S., Guerin-Deremaux, L., Pochat, M., Wils, D., Reifer,
C., & Miller, L. E. (2010). NUTRIOSE dietary fiber supplementation improves insulin resistance and determinants of metabolic syndrome in overweight men: A double-blind, randomized, placebo-controlled study. Applied
Physiology, Nutrition and Metabolism, 35, 773–782.
Luciano, M., Zhu, G., Kirk, K. M., Gordon, S. D., Heath,
A. C., Montgomery, G. W., et al. (2007). “No thanks, it
keeps me awake”: The genetics of coffee-attributed sleep
disturbance. Sleep, 30, 1378–1386.
Marley, E., & Blackwell, B. (1970). Interactions of monoamine oxidase inhibitors, amines, and foodstuffs.
Advances in Pharmacology and Chemotherapy, 8, 185–
239.
Maxwell, M. B. (1980). Reexamining the dietary restrictions with procarbazine (an MAOI). Cancer Nursing, 3,
451–457.
McCabe, B. J. (1986). Dietary tyramine and other pressor amines in MAOI regimens. Journal of the American
Dietetic Association, 86(8), 1059–1064.
Meng, I. D., Dodick, D., Ossipov, M. H., & Porreca, F.
III. ADJUNCTIVE/COMPLEMENTARY INTERVENTIONS
(2011). Pathophysiology of medication overuse headache: Insights and hypotheses from preclinical studies.
Cephalalgia, 31(7), 851–860.
Milde-Busch, A., Blaschek, A., Borggräfe, I., Heinen, F.,
Straube, A., von Kries, R. (2010). Associations of diet
and lifestyle with headache in high-school students:
Results from a cross-sectional study. Headache, 50,
1104–1114.
Nardi, A. E., Lopes, F. L., Freire, R. C., Veras, A. B., Nascimento, I., Valença, A. M., et al. (2009). Panic disorder
and social anxiety disorder subtypes in a caffeine challenge test. Psychiatry Research, 169, 149–153.
Orlosky, M. (1982). MAO inhibitors in sickness and in
health. Massachusetts General Hospital Newsletter: Biological Therapies in Psychiatry, 5, 25–28.
Panconesi, A. (2008). Alcohol and migraine: Trigger factor, consumption, mechanisms: A review. Journal of
Headache Pain, 9(1), 19–27.
Parsons, W. D., & Neims, A. H. (1978). Effect of smoking
on caffeine clearance. Clinical Pharmacology and Therapeutics, 24, 40–45.
Peterlin, B. L., Rapoport, A. M., & Kurth, T. (2010).
Migraine and obesity: Epidemiology, mechanisms, and
implications. Headache, 50, 631–648.
Pradalier, A., Bakouche, P., Baudesson, G., Delage, A.,
Cornaille-Lafage, G., Launay, J. M., et al. (2001). Failure
of omega-3 polyunsaturated fatty acids in prevention of
migraine: A double-blind study versus placebo. Cephalalgia, 21, 818–822.
Rice, S. L., Eitenmiller, B. R., & Kohler, P. E. (1975). Histamine and tyramine content of meat products. Journal of
Milk Food Technology, 38, 256–258.
Rice, S. L., Eitenmiller, B. R., & Kohler, P. E. (1976). Biologically active amines in food: A review. Journal of Milk
Food Technology, 39, 353–358.
Rice, S. L., & Kohler, P. E. (1976). Tyrosine and histidine
decarboxylase activities of Pediococcus cerevisiae and
Lactobacillus species and the production of tyramine in
fermented sausages. Journal of Milk Food Technology, 39,
166–169.
Riksen, N. P., Rongen, G. A., & Smits, P. (2009). Acute and
long-term cardiovascular effects of coffee: Implications
for coronary heart disease. Pharmacology and Therapeutics, 121, 185–191.
Rivas-Gonzalo, J. C., Santos-Hernandez, J. G., & MarineFont, A. (1983). Study of the evolution of tyramine
content during the vinification process. Journal of Food
Science, 48, 417–418.
Rogers, P. J., Hohoff, C., Heatherley, S. V., Mullings, E. L.,
Maxfield, P. J., Evershed, R. P., et al. (2010). Association of the anxiogenic and alerting effects of caffeine
with ADORA2A and ADORA1 polymorphisms and
habitual level of caffeine consumption. Neuropsychopharmacology, 35, 1973–1983.
Savi, L., Rainero, I., Valfrè, W., Gentile, S., Lo Giudice,
R., & Pinessi, L. (2002). Food and headache attacks: A
comparison of patients with migraine and tension-type
headache. Panminerva Medica, 44, 27–31.
10. Dietary Considerations
Selekler, H. M., & Budak, F. (2004). Idiopathic stabbing
headache and experimental ice cream headache (shortlived headaches). European Neurology, 51, 6–9.
Selekler, H. M., Erdogan, M. S., & Budak, F. (2004). Prevalence and clinical characteristics of an experimental
model of “ice-cream headache” in migraine and episodic tension-type headache patients. Cephalalgia, 24,
293–297.
Sen, N. P. (1969). Analysis and significance of tyramine
content in foods. Journal of Food Science, 34, 22–26.
Shapiro, R. E. (2008). Caffeine and headaches. Current
Pain and Headache Reports, 12, 311–315.
Shevel, E. (2011). The extracranial vascular theory of
migraine—a great story confirmed by the facts. Headache, 3, 409–417.
Spigt, M. G., Kuijper, E. C., Schayck, C. P., Troost, J., Knipschild, P. G., Linssen, V. M., et al. (2005). Increasing
the daily water intake for the prophylactic treatment of
headache: A pilot trial. European Journal of Neurology,
12, 715–718.
Strong, F. C., III. (2000). Why do some dietary migraine
patients claim they get headaches from placebos? Clinical and Experimental Allergy, 30, 739–743.
Sun-Edelstein, C., & Mauskop, A. (2009). Foods and sup-
175
plements in the management of migraine headaches.
Clinical Journal of Pain, 25, 446–452.
Taylor, F. R. (2011). Nutraceuticals and headache: the biological basis. Headache, 51, 484–501.
Tietjen, G. E., Herial, N. A., White, L., Utley, C., Kosmyna,
J. M., & Khuder, S. A. (2009). Migraine and biomarkers
of endothelial activation in young women. Stroke, 40,
2977–2982.
Udenfriend, S. Lovenberg, W., & Sjoerdsma, A. (1959). Physiologically active amines in common fruits and vegetables.
Archives of Biochemistry and Biophysics, 85, 487–490.
Vidaud, Z. E., Chaviano, J., Gonzales, E., & Garcia Roche,
M. O. (1987). Tyramine content of some Cuban cheeses.
Die Nahrung, 31(3), 221–224.
Wöber, C., Holzhammer, J., Zeitlhofer, J., Wessely, P., &
Wöber-Bingöl, C. (2006). Trigger factors of migraine
and tension-type headache: Experience and knowledge
of the patients. Journal of Headache Pain, 7, 188–195.
Wöber, C., & Wöber-Bingöl, C. (2010). Triggers of migraine
and tension-type headache. Handbook of Clinical Neurology, 97, 161–172.
Zencirci, B. (2010). Comparison of the effects of dietary
factors in the management and prophylaxis of migraine.
Journal of Pain Research, 3, 125–130.
C h a p t e r 11
Biofeedback and
Cognitive‑Behavioral Interventions
Reciprocal Contributions
Daniel Hamiel and Arnon Rolnick
We discuss in this chapter the integration of
cognitive-behavioral therapy (CBT) and biofeedback. Integrating these two major interventions
improves the potential of each and increases their
viability.
The concept of a common-law marriage can
figuratively describe the relationship between biofeedback and CBT. A common-law marriage is a
de facto relationship that is legally recognized as a
marriage even though no legal marriage ceremony
was performed. Similarly, CBT and biofeedback,
can “live together” de facto: Both are short-term
goal-directed interventions that aim to maintain
and reinforce their evidence-based status. Early in
their relationship, both approaches were related to
learning theory and were based partly on relaxation techniques. CBT handbooks today do not
include chapters about biofeedback (cf. Dobson,
2010; Bond & Dryden, 2004). The biofeedback literature, on the other hand, continues to recognize
the role of cognitive factors in biofeedback interventions. We are not aware of any clear therapeutic protocol that suggests how to integrate the two
methods.
We discuss in the first part of this chapter how
cognitive-behavioral concepts and techniques can
help practitioners using biofeedback. The second
part shows how cognitive-behavioral psychotherapists may use biofeedback techniques. We
conclude the chapter by considering how general
attitudes and presenting problems guide selection
of interventions.
Biofeedback intervention was originally based
on a learning theory model; however, it soon
became clear that thoughts, emotions, and interpersonal processes can affect the success of biofeedback interventions. Several authors contend
that from a theoretical point of view, the efficacy
of biofeedback is often partially due to cognitive
factors (Wickramasekera, 2002; Meichenbaum,
2007; Holroyd et al., 1984; Penzien & Holroyd,
2008).
How Biofeedback Practitioners Can
Integrate Cognitive‑Behavioral Elements
into Their Practice
The suggested model should take into account the
client’s thoughts during the session and the basic
assumptions (sometimes hidden) that impact the
client’s views about him- or herself and the ability to self-regulate. Here, a metaphoric screen,
the “cognitive–emotional screen” is introduced.
Clients share their thoughts with the therapist or
voice “what went on in their mind,” and therapist
and client both observe this process in relation to
the physiological data screen. Whereas working
with the physiological screen involves repeated
practice, working with the subjective experi176
11. Biofeedback and Cognitive-Behavioral Interventions
177
ence requires a significant amount of discussion.
Indeed, the focus within this model is on the relationship between subjective content (thoughts/
images) and objective measurement (physiology).
It is challenging to decide when and how to focus
on the physiological feedback versus engaging in
the verbal therapeutic discussion of the subjective
experiences.
One additional element of the integrated model
is the therapeutic relationship and how it influences the client’s ability to self-regulate. According to Taub and School (1978, p. 617), “perhaps
the most powerful factor influencing whether or
not thermal biofeedback learning will occur is
the quality of the interaction between the experimenter/therapist and the subject/patient, that is,
the ‘person factor.’ ” New directions in CBT, for
example, schema-focused therapy (Young, Klosko,
& Weishaar, 2003), also emphasizes the importance of the therapeutic relationship.
The proposed model introduces new elements
to classic biofeedback training by describing
ological demonstrations are crucial in establishing
the client’s motivation and trust in the intervention process.
Client motivation is also reinforced during the
psychophysiological intake interview, when the
practitioner gathers information pertaining to the
client’s reactivity and recovery rate. The practitioner shares this information with the client. The
process enables the therapeutic triangle of client,
practitioner, and biofeedback. This enables clients
to understand the cognitive–psychophysiological
relationships.
1. The macro level: How to use CBT elements in
the in the in the process of the training.
2. The micro level: How to use CBT elements
within each session.
The Macro Level: Stages
Acquaintance and Educational Stage
The acquaintance and educational stage is the
“psychophysiological profiling” stage, in which
the practitioner measures a client’s physiological responses to a variety of stimuli. We use the
term “acquaintance” for this stage, because the
practitioner becomes acquainted with the client’s
psychological and physiological responses, and the
client is introduced to the psychophysiological
view of his or her disorder for the first time.
The client experiences and learns the psychophysiological “dialogue” associated with their cognitions. The cognitive approach is based on the
biopsychosocial model, which emphasizes the crucial role played by the client’s thinking and active
involvement in treatment.
The acquaintance stage also involves client education and instruction in the intervention. Much
of the explanation is through psychophysiological
demonstrations. The practitioner shows the client how specific external stimuli (e.g., behavior,
thoughts, and imagination) influence physiological responses and symptoms. These psychophysi-
Behavioral Techniques Stage:
Acquiring Self‑Regulation
Relaxation techniques, represent the B, behavioral, in CBT. A CBT practitioner listens to the
client’s cognitions during the relaxation learning
process. What is the client’s internal dialogue?
Does the client believe that he or she is able to
train his or her physiology? Does the client feel
“betrayed” by his or her body? Failures in the selfregulation process can be used as an opportunity
to learn about possible cognitions that can be discussed and modified.
Cognitive Techniques Stage: Acquiring Self‑Regulation
In addition to biofeedback, cognitive techniques
can be very useful in teaching means to achieve
relaxation. For example, Meichenbaum's (2007)
stress inoculation technique (SIT) can facilitate
autogenic therapy. In SIT, subjects are taught to
substitute positive self-statements and expectations for negative self-statements about their ability to relax.
Clients often have difficulties reducing physiological arousal. This is a time to help the client
understand the automatic thoughts that can interfere with acquiring relaxation skills. According
to the ABC model, the source of an individual’s
emotional, physiological or behavioral Consequences is not the Actual events but the individual’s thoughts and Belief system. Using the ABC
model, we can help the client identify and modify
automatic thoughts and beliefs.
A “third generation” of cognitive interventions,
acceptance and commitment therapy (ACT;
Hayes, Luoma, Bond, Masuda, & Lillis, 2006), has
emerged in recent years. One of the main premises
of ACT is the understanding that to try too hard is
unproductive. Practitioners using biofeedback can
178
use this ACT principle to help clients learn to “let
go” and “not try too hard.”
Generalization Stage
Clients, then, must learn to implement self-regulation techniques outside the practitioner’s office,
thus generalizing them. Relaxing under less comfortable conditions is challenging, and altering the
conditions typically is done gradually. For example,
the client starts in a reclining chair, then a sitting
position in an office chair, and later standing up.
The client also learns to practice with more common mental state conditions. Training proceeds
with the client staying relaxed while maintaining
alertness, and being involved with the environment without being overwhelmed by its demands.
We ask the client what external and internal conditions help or thwart self-regulation. All of these
elements aid generalization.
Exposure and Desensitization Stage
Exposure is a major intervention in CBT. While
biofeedback procedures do not necessarily include
such intervention, the proposed model suggests
integrating various types of exposure into biofeedback practice.
The cardinal principle to remember during this
stage is to achieve a state of relaxation while simultaneously re-creating a real or imaginary stressful
scenario. For example, a client who is afraid of
speaking in front of an audience is asked to prepare
and deliver a speech while continuing to regulate
his or her physiology. An instruction to maintain
physiological readings at a low level or to lower
them as soon as they begin to rise, is provided. A
client who is anxious about taking tests provides
another example. The practitioner can create test
scenarios during the therapy session and train the
client to reduce stress during a test. At times it may
be necessary to re-create an actual test scenario,
but usually in vitro re-creation is sufficient.
We can therefore train clients to implement
and practice relaxation techniques during difficult
situations, whether in vivo or in vitro. Clients can
advance to the final stage only after they acquire
and learn to implement control abilities in stressful situations as well.
Weaning and Termination Stage
All forms of treatment are geared toward enabling
clients to separate from and leave the supportive
III. ADJUNCTIVE/COMPLEMENTARY INTERVENTIONS
environment. Clients who sought out biofeedback because they were not able to regulate their
tensions acquire the ability to do so without the
intervention. It is also assumed that some of the
client’s psychological problems will emerge during the transitional or separation phases. Therefore, working through this stage while the client
is still in treatment provides the opportunity to
re-create difficulties pertaining to this process. In
biofeedback therapy, termination involves “weaning” off the device as well, thus enabling the client
to gain control over his or her psychophysiological responses without direct feedback. During this
stage one of the client’s task is to link internal
physiological clues with psychophysiological readings, then construct a replacement in the form
of habitual internal feedback. Hence, this stage
deals with the ability to generate the relaxation
response even without the aid of the device. To
achieve this goal, the practitioner measures the
physiological variable without displaying the
feedback. The client is asked to “guess” his or her
physiological status, as well as attempt to estimate
whether the line of the graph is sloping upward
or downward (“discrimination training”; Andrasik
& Blanchard, 1983; Gainer, 1978). This stage is
marked by the client’s deliberations over whether
the time has come to terminate treatment, and
whether the treatment has achieved its goals.
Questions are raised pertaining to dependence
and independence. Moreover, the client must cope
with the loss of fantasies that “all my problems
will be solved” or “I will not get tense anymore.”
Indeed, the client must relinquish some of his or
her basic assumptions regarding constantly being
in control and feeling good. We expect that during this stage of treatment, the emphasis will be
on the verbal component of the psychotherapeutic
dialogue. (For a discussion of other ways to facilitate generalization and transfer of training, see
Lynn & Freedman, 1979.)
The Micro Level: Session Structure
As cognitive and emotional factors are introduced
into biofeedback training, the structure of the
sessions should include various phases that allow
identification and discussions of the client’s cognitions.
Beck (1995) suggests that CBT sessions should
be divided into a number of phases: brief update
and mood check, transition from previous session, setting agenda, homework review, discussion
of agenda issues, homework assignment, sum-
11. Biofeedback and Cognitive-Behavioral Interventions
179
mary, and feedback. The biofeedback literature is
less consistent regarding this issue. Some writers
describe the various sensors used to take measurements, the feedback screen, and other equipment in great detail; however, the session format
is often left unmentioned. This may explain why
so many laypeople believe that each biofeedback
session consists of a full hour with the client sitting in front of the screen trying to control some
line graphs. In contrast, Andrasik and Schwartz
(see Chapter 20, this volume) describe a possible
format for headache sessions: adaptation, baseline,
self-regulation, stimulation, biofeedback, reassessment of self-regulation (also see Andrasik &
Blanchard, 1983).
We divide each session into different phases;
some of them are mainly verbal phases in which
we discuss the client’s internal processes (emotions, thoughts, and cognitions). Other phases use
the physiological measures, with each phase using
biofeedback devices differently. So we obtain measurements and provide biofeedback several times
during a session. Each of these events is then
accompanied by a period dedicated to verbal discussion about the subjective and objective psychophysiological data. The verbal dialogue enhances
the physiological changes and vice versa. The full
description of these phases is available at www.
marksschwartzphd.com.
third generation of CBT. This new therapy trend
introduced the role of mindfulness (Hayes et al.,
2006) and controlling attention in therapy (Wells,
2000). In this section we show how biofeedback is
relevant to these developments, theoretically and
practically.
How Cognitive‑Behavioral Therapists Can
Integrate Biofeedback into Their Practice
In the past, physiology was neglected in psychotherapy, and to a certain extent this remains true
today. This trend began when Freud abandoned
his project aimed at combining neurology and
psychiatry (Gay, 1998) and continued when cognitive therapists ascribed an exclusive position in
therapy to thoughts (Beck, 1995). Work by Servan-Schreiber (2004) and LeDoux (1996) emphasize the special characteristics of the emotional
brain. This brain is responsible for emotions and
physiology but often functions separately from the
cortex and from cognitive influences. Hence, the
physiological component in therapy becomes very
important.
Under the assumption that physiology plays a
major role in mental health, we focus in this section on the role of biofeedback as a therapeutic
agent in a CBT setting. The role of biofeedback
and physiology in therapy has recently become
more natural with the reinforcing influence of the
Rationale: Biofeedback as Training
for Metacontrol
Individuals seek out therapy because they are
unable to stop or control their suffering, whether
physical or mental. Naturally, they believe that
regaining control will put an end to their suffering. Rather than directly trying to stop suffering
and get the individual back in control, CBT challenges and changes the concept of control. CBT
proposes the notion that a person's problem is not
with reality (i.e., not being able to stop the suffering) but with the way he or she perceives this
reality (the belief that the person should stop the
suffering to go on with life—the ABC model
discussed earlier). The goal of therapy is to persuade patients to give up their dysfunctional basic
assumption that life should go the way they want
or expect it to go. CBT seeks to provide a more
sophisticated concept of control, that is, the freedom to relate to one’s situations rather than trying
to control them completely (something that is not
always possible). Individuals should be liberated
from the (Western) addiction to achieving everything they want. According to the CBT concept,
“control” is the freedom to make a decision about
whether to try to change actual reality or whether
to take on the more challenging task of accepting
reality when necessary. This ability is referred to
as “metacontrol.”
Relaxation can serve as an important step
toward achieving acceptance and metacontrol.
Therapists like to help their patients choose
between holding on (remaining tense to make a
change) and letting go (relaxing and accepting).
Yet this choice might be threatening for a patient.
In contrast to relaxation, hypervigilance and tension serve a patient’s need to feel in control, which
is very hard to abandon. Biofeedback can provide
an exciting solution to this obstacle. Relaxation
using biofeedback involves experiencing control
(though it is actually an act of giving up) and is
therefore much less threatening. Indeed, biofeedback is unique in that the individual experiences
a sense of control while actually letting go. This is
precisely the meaning of “metacontrol,” as mentioned earlier.
180
In CBT, patients learn to give up their basic
dysfunctional assumptions related to their need
for control. The therapist persuades the patient to
distract him- or herself from things that used to
make him or her feel more responsible and more
in control (e.g., being preoccupied with worries).
Feeling in control based on continuous worrying is
an illusion, but letting go of the feeling is still not
easy. In therapy, patients are exposed to scenarios
that create the most threatening state of mind for
them—relinquishing the wish for complete confidence. To do so, patients must abandon the bodily
state of alertness that they use to maintain the illusion of being in control. By going through such a
process, the patient eventually gives up the need
for absolute control and the illusion of having it.
In return, the patient achieves the best type of
control possible in the actual situation. Biofeedback plays an important role in achieving this
“best possible control” experience, that is, “metacontrol.”
Conditioned Emotional Triggers: What CBT
Cannot Resolve
Dysfunctional basic assumptions can usually be
addressed by CBT. However, direct and increased
sensitivity to external triggers related to the emotional brain (discussed earlier) often cannot be
influenced by the cortex and should therefore be
treated by a system other than the cognitive one.
This sensitivity is conditioned and very much connected to physiology.
The integrated CBT and biofeedback (physiological) intervention relates to each individual
as a whole, including both cognitive and emotional–physiological dimensions. This proposed
intervention attempts to respond to the challenge
posed by direct, noncognitive, emotional, and
physiological sensitivity by adding a physiological
dimension to therapy. As in the first part of this
chapter, two different views of the intervention are
presented. First, the general role of biofeedback in
the different stages of CBT is discussed, then its
practical application in the sessions themselves is
considered. Each subsequent CBT stage is needed
if the previous one proves insufficient to solve the
patient's problem reasonably.
Stages in Biofeedback‑Assisted CBT Intervention
• Stage 1: Attempting to Stop the Suffering. At
this stage of therapy, the patient has already identified the catastrophic thoughts that accompany
III. ADJUNCTIVE/COMPLEMENTARY INTERVENTIONS
suffering, but they are not yet being effectively
challenged. The therapist assumes at this stage
that the patient does not really believe in the catastrophic thoughts, which serve other purposes,
and therefore has no need to challenge them. The
patient is aware of the poisoning effect of these
thoughts and has learned not to cooperate with
them. Biofeedback at this stage helps deflect the
patient’s focus from the thoughts and relieves suffering by reducing the bodily stress that accompanies these thoughts. This step reflects the normal
role of biofeedback as a tool for improved control over the body. By reducing bodily stress, the
patient can shift his or her focus from the “thinking about” mode to a more positive and active coping mode. This is sometimes sufficient to stop the
suffering.
• Stage 2: Attempting to Reduce Suffering by
Changing a Patient’s Way of Thinking to Change
the Resultant Destructive Behavior. Changing the
patient’s thinking is the hard core of most cognitive-behavioral interventions. At this stage, the
therapist tries to challenge the patient’s dysfunctional basic assumptions that are the source of
suffering, and to shift the focus of attention from
fixed suffering to functionality. The patient learns
to accept that direct efforts to stop the suffering
are insufficient. Biofeedback helps patients learn
and use the concept of “metacontrol” to achieve
this goal.
The therapist helps the patient understand
the reasons he or she is stuck because of negative
and catastrophic thoughts. Together they try to
discover the hidden reasons (“metacognitions”
according to Wells, 2000) that make the patient
focus on the negative thoughts. One major hidden
reason is the concomitant experience (or rather,
the illusion) of taking control over one’s situation
by constant thinking or worrying. The therapist
and the patient attempt to discover dysfunctional
basic assumptions that make the patient believe
in the catastrophic thoughts. For example, if a
patient believes that to feel safe he or she needs
a strong supportive figure, when alone he or she
might be afraid of “falling apart.”
The patient may try to avoid being alone, and
this effort may be expressed physiologically. CBT
that works on cognition alone will not be sufficient because of the strong negative conditioning
usually involved in such a case. Here, the role of
biofeedback is to release the physical tension preventing the patient from thinking realistically.
Bodily alertness and hypervigilance are related
181
11. Biofeedback and Cognitive-Behavioral Interventions
to narrowed attention and a polar way of thinking, which may be effective in emergencies but not
under normal circumstances.
Biofeedback and physiological techniques in
general help patients expand their attention and
open them up to more flexible ways of thinking
and behaving. They serve to release bodily tension
from emergency mode.
• Stage 3: Directly Experiencing Emotions and
Sensations. When an individual cannot control
his or her behavior due to the intensity of the negative emotions, a third stage is needed. Changing
cognitive assumptions, releasing physical tension,
and defocusing are not sufficient. Some forms of
suffering are resistant to disappearing.
At this stage, the patient is taught to overcome
judgmental thinking that feeds negative uncontrolled emotions and is related to the past. The
patient is directed to remain focused on the difficult emotions and sensations of bodily stress. The
patient learns how to enhance skills to remain in
the present (mindfulness) with the suffering (the
emotions and sensations) and to deflect the focus
away from negative judgmental thoughts (the
past) and anticipatory anxiety (the future). Bodily
sensations reflect the most pure experience of the
present. Hence, biofeedback, which helps connect
the patient to his or her body, plays an important
role in developing this “present-orientation skill.”
By teaching the patient to recognize bodily
changes, partly control them, and accept the fact
that control is only partial, biofeedback helps the
patient to keep from returning to a focus on the
negative thoughts (usually judgmental self-evaluation). Consequently, it helps to minimize the emotional and physiological impact on suffering and
behavior. The amount of suffering that remains
probably is necessary to illuminate some problematic issues the patient still needs to address.
Integrated biofeedback-assisted CBT interventions are recommended primarily for treating people with anxiety disorders that are marked by elements of hyperarousal and “fear of loss of control”
or states of uncertainty. These characteristics are
associated with many other psychiatric disorders,
such as eating, mood, sleep, and impulse control
disorders. The following is a clinical demonstration of the potential advantage of using biofeedback techniques in CBT. First, the use of biofeedback in the psychoeducational part of the therapy
session is described, followed by a demonstration
of the core of therapy.
The Use of Biofeedback Techniques in CBT
Experiential Psychoeducation
The psychoeducational part of CBT can be very
influential, and, for some patients, this accounts
for most of the change effected by the therapy. In
simple CBT, the therapist uses the cognition itself
to demonstrate some cognitive distortions and
relevant information. However, this remains on
the verbal level, which in turn is affected by the
patient’s cognitive distortions and subjective perceptions. On the other hand, integrated therapy
can help the patient directly experience various
effects of their thoughts and expectations. Here
are some examples:
• Demonstration of the fight-or-flight response.
Eliciting any surprise response from patients
enables them to experience the normality of such
a response, which they otherwise would have perceived as some form of personal weakness. In such
a response, the electrodermal response (EDR) line
graph ascends rapidly, then slowly descends. One
of the major problems in dealing with anxiety is
the patient’s response to his or her symptoms. The
patient becomes anxious and sometimes depressed
because of this anxiety. Normalizing this anxiety
by demonstrating the “fight-or-flight response”
and other mind–body connections can alleviate
some of the patient’s worries about his or her own
responses and symptoms.
• Demonstration of anticipatory anxiety. The
goal of this intervention is to challenge the patient’s
misconception that being ready in advance will
help. The therapist asks the patient to get ready for
the same kind of “surprise” generated by the “fightor-flight” response demonstration but this time to
try to suppress the response in advance. The therapist counts from 10 to 0, then presents the previously used stimulus, while assessing arousal level
throughout the entire process. The therapist then
shows the patient the high price of arousal that is
paid for being prepared. This overpreparedness is a
very common component of many psychopathologies. Anxious patients are constantly on the alert
in order to be ready for possible future disasters.
They are preoccupied with thoughts related to the
past and worries about future disasters. This intervention also demonstrates very nicely the concept
of “unnecessary suffering.” Relinquishing this tendency toward being prepared in advance will help
the patient avoid accumulating unnecessary stress
and worries.
182
• Demonstration of the thought–physiology–
emotion connection. According to CBT, an event
is not the reason for the consequences; rather, it is
due to the person’s thoughts or the belief system.
The following biofeedback demonstration can
illustrate this unequivocally. The therapist says:
“Try to make the line graph descend, faster . . .
faster. . . . I am disappointed in you.” The therapist can also “threaten” the patient with a difficult
math test.
This simple exercise can also demonstrate the
notion of “metacontrol.” Trying too hard will not
help the patient control his or her physiology. The
physiological recordings cannot be controlled
until the patient gives up or relinquishes the desire
to relax or achieve balance.
• Understanding psychosomatic phenomena.
This can be demonstrated by asking the patient
something that causes an internal conflict
between the wish to please the therapist and the
resistance to the request. For example, the therapist might say, “May I ask you a personal question?”
In a group, the therapist might say the following
to one of the participants: “Will you straighten up
the room at the end of the session?” The patient
will probably say “yes,” even though his or her body
keeps saying “no,” expressing all real feelings via
physiological changes, by increased arousal level,
increased muscle tension, and lowered peripheral
temperature. The therapist then explains that saying “no” too often will bring about a continuous
somatic response, or psychosomatic phenomena.
Such a demonstration, in contrast to a purely theoretical explanation, has the potential to give the
psychotherapy a real push by offering a rationale
for psychosomatic symptoms and the motivation
to look for the mental source of such symptoms.
• Comparison between practicing biofeedback
with eyes open and eyes closed. If a patient performs
better with his or her eyes open, this may signal
an added need for control (very often the case
with children). Problems in controlling biofeedback measures with eyes open may indicate performance anxiety and sometimes a coping model
of avoidance. In such cases, the therapist should
train the patient in biofeedback therapy with eyes
open while remaining aware of the patient’s performance. On the other hand, practicing biofeedback with eyes closed can provide exposure for a
patient who has difficulties performing under conditions of uncertainty. This kind of exposure can
be a particularly important experience for patients
with anxiety disorders.
III. ADJUNCTIVE/COMPLEMENTARY INTERVENTIONS
• Demonstration of the power of acceptance
and distraction. As described earlier, an important
element in CBT involves relinquishing the tendency to judge oneself through constant evaluations of success. Using biofeedback to distract
patients from this constant evaluation can help
them accept and evaluate results almost without
self-judgment. During biofeedback practice, the
patient is trained to distract him- or herself from
the direct goal of success. This usually improves
performance. The patient learns to give up on a
coping model that focuses on results and learns
to enjoy the process. The patient is trained to be
aware of his or her performance, to accept it even if
he or she is not satisfied, and to continue to do his
or her best in a balanced and enjoyable way. The
patient is taught the importance of distraction from
judgment and learns to focus on practical evaluation of performance, as well as to decide what
action to take. Both criticism and compliments of
the patient’s performance during biofeedback will
most probably harm performance. On the other
hand, acceptance of what cannot change for now
(unavoidable suffering in the present) will maximize performance.
• Understanding the activity of the autonomic
nervous system (ANS). Heart rate variability
(HRV) as a measure of ANS balance level is shown
by the ratio between the activity of the sympathetic and the parasympathetic nervous systems.
Biofeedback software settings that provide direct
feedback about the patient’s balance as reflected in
the ratio of these two systems help patients learn
that the goal is usually not relaxation but achieving a balanced state. The understanding that relaxation is not the goal is helpful to patients who fear
losing control, because balancing the ANS allows
for control, sometimes without the feature of relaxation. An increased belief in ability to control provides the patient with the needed confidence to
expose him- or herself to previously avoided “risky
situations.”
• Demonstrating the difference between a baseline state, which usually indicates low ANS balance, and the results of short biofeedback intervention
(biofeedback with slow breathing or guided imagery). Using this method, it is usually easy to generate a quick change. Low HRV usually co-occurs
with a coping model of alert hypervigilance or a
simplified conception of control, such as “It would
be awful for things to happen against my will, so I
have to remain alert.” Simple physiological intervention such as slow breathing with the help of
11. Biofeedback and Cognitive-Behavioral Interventions
183
biofeedback can be used to demonstrate how easy
it is to relinquish this attitude and increase HRV
and balance level, at least for the present time.
kinds of meditation or mindfulness techniques,
the patient attempts to minimize distractions,
reduce bodily responses to destructive ideas, and
decrease arousal level. These parameters can easily
be monitored with the EDR modality (and others as well), hence making it possible to check the
effectiveness of mindfulness and meditation techniques.
Paying attention to the body is an important
element in mindfulness techniques. Eastern philosophy and CBT share the notion that focusing
on the present releases individuals from their fears
and worries. An individual cannot become anxious unless past negative memories and worries
about the future are combined. Focusing on experiencing the present can prevent this. Using biofeedback to increase awareness of body sensations
can support this process. Awareness of bodily sensations helps distract the individual from thoughts
that are not related to the present and therefore
helps the person experience the present. This
training is critical in the third stage of therapy
described earlier.
Exposure is the main goal of each stage of CBT:
being exposed to life's tasks and to challenges the
patient is trying to avoid. Biofeedback can support
this process in several ways:
• The same can be demonstrated very nicely
using electromyography (EMG). Patients hooked up
to EMG sensors on their extensor or flexor muscles
usually exhibit high EMG levels, sometimes even
above 20 mV. For many patients, however, asking
them to relax their muscles and sit comfortably is
sufficient to monitor a dramatic reduction in muscle tension (below 1mV)! So after a short intervention, excessively high readings (usually a low
balance of ANS at baseline) can turn into much
more balanced ANS readings.
• The simultaneous use of two sensors (modalities) can reveal some very interesting information
about an individual’s coping model. Very often
we find a contradiction between measurements
in two different modalities. For instance, during
guided imagery, a child with encopresis describes
an “accident” while in class. As predicted, his
peripheral temperature decreases, but his EDR
also decreases! When asked what he is doing in
class during the “accident,” the child answers, “I
am pretending I am not there!” The child is dissociating, or in CBT terms, using a coping model
of avoidance. In this case, the child is not avoiding
the situation itself but rather is dealing with the
situation.
Our conclusion is that as long as an individual’s
physiological modalities are not working in the
expected direction, he or she is still paying a physiological price for some unfinished psychological
issues or a nonadaptive coping model. Biofeedback
can serve as a beacon, lighting the way for psychological intervention.
The Core of Therapy
In Stage 1 of the therapy, the task is to stop suffering. Here, the role of the biofeedback in the core
of therapy is clear. The therapist trains the patient
to better control physiological elements using different biofeedback modalities. Specifically, the
patient is taught to attempt to control symptoms
directly, such as headache and muscle tension,
breathing problems due to stress, extreme sweating, cold hands, panic attacks, and tremors.
In discussing the stages of CBT therapy earlier,
we described the importance of controlling attention. This skill is relevant in all stages of therapy. The easiest biofeedback modality to use for
enhancing this skill is the EDR. Using different
• Demonstrating the exposure principle. The
patient needs to learn that avoidance is not the
only way to reduce anxiety. Paradoxically, exposure does this much better, and the effect remains
for a longer time. The EDR and EMG line graphs
descend (and the temperature rises) with exposure (desensitization or even flooding type of
exposure), demonstrating habituation, while the
patient keeps a focus on the source of his or her
negative emotions. The therapist can use sounds,
pictures, or movies, or can create a scene related
to the patient's source of negative emotions.
This is appropriate for treating phobias, and can
also work very well with exposure to problematic
thoughts, as in obsessive–compulsive disorder. For
the patient, monitoring bodily sensations during
exposure is like looking inside oneself. The screen
presenting the new balance in the patients’ physiological responses convinces him or her that the
anxiety has been resolved even before it is noticed
personally.
• In vivo exposure. This can often be implemented in the clinic (e.g., when treating fears
related to blood tests, injections, blood pressure
measurement, or insects). The therapist prepares
184
the patients for an increase in stress and arousal
level at the beginning of exposure, and later habituation. Often, however, exposure occurs without
any anxiety emerging. The patient experiences the
paradoxical effect of the exposure from the first
moment. For example, asking the patient to focus
on negative thoughts leads to a decrease in arousal
level and anxiety, while asking the patient to avoid
negative thoughts causes an increase in arousal
level and anxiety. These phenomena, best demonstrated by the biofeedback graph, demonstrate the
importance of exposure. Also demonstrated is the
ineffectiveness of attempting to push thoughts out
of one’s mind and the effectiveness of accepting
thoughts.
• In vitro exposure. Guided imagery of the
actual problematic scenario the patient is trying
to avoid is used in preparing for in vivo exposure. If
the patient is hooked up to biofeedback sensors, it
is possible to identify the more problematic points
during in vitro exposure and better prepare for in
vivo exposure. More importantly, implementing
this imagined in vitro exposure many times makes
the real exposure (in vivo) much easier.
These are all examples of how biofeedback can
motivate, enhance, comment on, and demonstrate
a variety of cognitive techniques at different stages
of therapy.
III. ADJUNCTIVE/COMPLEMENTARY INTERVENTIONS
representing the second difference between the
models, depends on the individual’s characteristics, so it is important to explore the above characteristics in every client. The answer depends
also on the presenting problem, representing yet
another difference between the two models. If
extreme physiological responding is involved, it
makes sense to try to bring the patient some relief
and use biofeedback. However, this can still be
achieved by a psychotherapist familiar with biofeedback (second model or intervention).
Yet there is a clear difference between the two;
therefore, there are clear criteria for when to start
or to switch to psychotherapy. When there is a
major psychological problem (strong dysfunctional
basic assumptions), biofeedback intervention is
not sufficient. If this is known in advance, through
experience with the patient or questionnaires,
starting with the second intervention is advisable.
If a major psychological problem emerges during
biofeedback intervention, it is prudent to move
to psychotherapy as well. Otherwise, the main
criteria are personal characteristics, as described
earlier. Given that most psychotherapists still
refrain from the use of psychophysiological tools,
we can imagine an opposite scenario in the case
of extreme physiological symptoms or unbalanced
ANS that ordinary psychotherapy cannot effectively address. In such cases it makes sense to refer
the client to a biofeedback practitioner for a parallel intervention.
Conclusion
The first difference between these two models—
CBT-assisted biofeedback and biofeedback-assisted
CBT—is in who implements the intervention. In
the first model, this is a biofeedback practitioner,
while in the second, it is a psychotherapist. Naturally, the client is the first to select between the
two interventions by deciding whether to pursue
biofeedback intervention or psychotherapy. It is
assumed that individuals who select biofeedback
will have some, though not all, of the following
characteristics: pay more attention to the body,
become more concrete, look for control, be more
scientifically oriented, be less oriented to talking
about feelings, less able “to get into” thoughts and
feelings, less aware of possible psychological conflicts, and looking for a quick change (Wickramaskera, 2002). Those who prefer psychotherapy will
have more or less the opposite characteristics.
Still, if the decision is ours, whom should we
refer to which type of intervention? The answer,
References
Andrasik, F., & Blanchard, E. B. (1983). Applications of
biofeedback to therapy. In C. E. Walker (Ed.), Clinical psychology: Theory, research, and practice (Vol. 2,
pp. 1123–1164). Homewood, IL: Dow-Jones-Irwin.
Beck, J. (1995). Cognitive therapy basics and beyond. New
York: Guilford Press.
Bond, F. W., & Dryden, W. (2004). Handbook of brief cognitive behavioral therapy. New York: Wiley.
Dobson, K. S. (Ed). (2010). Handbook of cognitive-behavioral
therapy (3rd ed.). New York: Guilford Press.
Gainer, J. C. (1978). Temperature-discrimination training
in the biofeedback treatment of migraine headache.
Journal of Behavior Therapy and Experimental Psychiatry,
9, 185–188.
Gay, P. (1998). Freud: A life for our time. New York: Norton.
Hayes, S. C., Luoma, J., Bond, F., Masuda, A., & Lillis, J.
(2006). Acceptance and commitment therapy: Model,
processes, and outcomes. Behaviour Research and Therapy, 44, 1–25.
Holroyd, K. A., Penzien, D. B., Hursey, K. G., Tobin, D. L.,
Rogers, L., Holm, J. E., et al. (1984). Change mecha-
11. Biofeedback and Cognitive-Behavioral Interventions
185
nisms in EMG biofeedback training: Cognitive changes
underlying improvements in tension headache. Journal
of Consulting and Clinical Psychology, 52, 1039–1053.
LeDoux, J. E. (1996). The emotional brain: The mysterious underpinnings of emotional life. New York: Simon &
Schuster.
Lynn, S. J., & Freedman, R. R. (1979). Transfer and evaluation of biofeedback treatment. In A. P. Goldstein &
F. Kanfer (Eds.), Maximizing treatment gains: Transfer
enhancement in psychotherapy. New York: Academic Press.
Meichenbaum, D. (2007). Stress inoculation training: A
preventative and treatment approach. In P. M. Lehrer,
R. L. Woolfolk, & W. E. Sime (Eds.), Principles and practice of stress management (3rd ed.). New York: Guilford
Press.
Penzien, D. B., & Holroyd, K. A. (2008). Change mechanisms in EMG biofeedback training: Cognitive changes
underlying improvements in tension headache. Headache, 48(5), 736–737.
Servan-Schrieber, D. (2004). The instinct to heal: Curing
stress, anxiety and depression without drugs and without
talk therapy. London: Rodale International Ltd.
Taub, E., & School, P. (1978). Some methodological considerations in thermal biofeedback training. Behavior
Research Methods and Instrumentation, 10, 617–622.
Wells, A. (2000). Emotional disorders and metacognition:
Innovative cognitive therapy. Chichester, UK: Wiley.
Wickramasekera, I. (2002). the placebo effect and its use
in biofeedback therapy. In D. Moss, A. McGrady, T.
C. Davies, & I. Wickramasekera (Eds.), Handbook of
mind–body medicine for primary care. Thousand Oaks,
CA: Sage.
Young, J., Klosko, S., & Weishaar, M. E. (2003). Schema
therapy: A practitioner's guide. New York: Guilford Press.
Part IV
Relaxation Interventions
Chapter 12
Relaxation Today
Self‑Stressing and Psychological Relaxation Theory
Jonathan C. Smith
• A client has mastered several general strategies, including concentrative meditation, visual
imagery, autogenic imagery, PMR, and imagery. Should the client practice them one after
another, like a mechanical fitness routine? Is
there a way of more meaningfully combining or
sequencing these techniques?
• A client finds mindfulness and autogenics
exceptionally rewarding and is highly motivated
to extend the benefits of the training session to
work. How might a clinician address this wish?
• A client as successfully learned to control pain
or reduce physiological arousal. However, he
finds his routine mechanical and uninteresting
and is tempted to quit. How might you alter his
exercise to maintain interest and increase generalization?
• A client is a devout Catholic and devotes over
an hour a day to religious study and practice.
She finds biofeedback relaxation mechanical
and foreign. How might one incorporate her
relaxation skills into a meaningful spiritual
activity?
We live in a new world of relaxation. In decades
past, one might be proficient at a single approach,
say, autogenic training or progressive muscle relaxation (PMR), mastered in an extended and costly
program led by a master at a university, hospital,
training clinic, or distant retreat. Today, there is
a stunning profusion of freely accessible strategies
from around the planet. One can find a free MP3
or YouTube presentation of just about any technique imaginable, and attempt practice anywhere
the smartphone or tablet will boldly go.
Although this world may well seem brave and
new, it brings into clear focus a challenge that has
always faced serious relaxation clinicians and biofeedback practitioners—the mindless temptation
to oversimplify, to restrict oneself to one strategy,
to succumb to the comfortable myth that one size
fits all. Put clearly, I believe this is the core challenge: Lacking a road map, clinicians and clients
can readily miss important options and opportunities. We may fail to consider some untraditional
but important questions:
• A client appears to master PMR. At what point
might mindfulness meditation enhance the
effects of training?
• A client finds visual imagery too distracting.
When would preparatory breathing exercises,
or perhaps simple stretching, help? Which one
should be tried?
In this chapter I offer two road maps of selfrelaxation. I suggest a framework for making
sense of the hundreds of self-relaxation exercises
now available, and offer a way techniques can
be systematically explored and utilized. Over the
decades, I have offered various schemas. Else189
190
where I have also produced a comprehensive parallel model based on mindfulness and relaxation
(Smith, 2015; http://blogs.roosevelt.edu/jsmith).
Road Map 1: Self‑Stressing Theory
Many relaxation experts (e.g., Lehrer, Woolfolk, &
Sime, 2007) sort the myriad self-relaxation techniques into more or less six groups: yoga stretching, PMR, breathing exercises, autogenic training,
imagery/positive self-statements, and meditation/
mindfulness. I propose that this differentiation
is no accident but an inevitable consequence of
the very nature of psychological and physiological processes that underlie stress arousal (Smith,
2006, 2007a). According to self-stressing theory,
there are six ways we can trigger and sustain physiological stress arousal. Each form of self-stressing
suggests a corresponding family of relaxation technique (Smith, 2006, 2007a). Let me explain.
There are six ways that people trigger and
maintain their physiological “fight-or-fight” stress
response.
1. Stressed posture and position. When confronted
with stress, people often assume a variety of
defensive or aggressive postures or positions
(standing, crouching, bending over a desk) for
an extended time. This, combined with sustained immobility, can evoke skeletal muscle
tension, joint stress, reduced blood flow, and
pooling of blood, and contribute to tension,
fatigue, and decreased energy.
2. Stressed skeletal muscles. When threatened, one
clenches, grips, and tightens skeletal muscles
to prepare for attack or escape. When chronic,
such tension can contribute to pain and fatigue.
3. Stressed breathing. When stressed, one is more
likely to breathe in a way that is shallow,
uneven, and rapid, deploying greater use of the
intercostal (rib cage) and trapezius (shoulder)
muscles, and less use of the diaphragm.
4. Stressed body focus. Simply attending to and
evoking thoughts and images about a specific
body part or process can evoke related neurophysiologial changes. An individual facing a
threat may notice her rapidly beating heart or
churning stomach. Attending to and thinking
about these somatic reactions can aggravate
them.
5. Stressed emotion. We often motivate and energize ourselves for a stressful encounter with
affect-arousing cognitions. We entertain fanta-
IV. RELAXATION INTERVENTIONS
sies and repeat words and self-statements that
can evoke negative affects of anxiety, anger, or
depression.
6. Stressed attention. When dealing with a threat,
we actively and effortfully concentrate on
attacking, defending, or running. In addition,
we often direct our attention to multiple targets, including competing tasks (as in multitasking), a targeted task versus worried preoccupation, or self-stressing efforts (thinking how
one is breathing, maintaining a stressed posture or position, thinking about relaxed fantasies or negative emotions, etc.) rather than the
task at hand. Such attentional strain maintains
arousal.
Self-stressing theory proposes that the current
myriad popular self-relaxation techniques, used
alone and often in conjunction with biofeedback,
can be organized according to self-stressing. Thus,
we can think of six universal family groups of selfrelaxation techniques (see Table 12.1).1
Obviously, each family group does much more
than address one specific type of self-stressing. Selfstressing theory does not claim that one universal
TABLE 12.1. Self-Stressing and the Six
Universal Family Groups of Self-Relaxation
Family group
of self-relaxation
Self-stressing
Stretching exercises
Stressed posture and
position
Tense–let go exercises
Stressed skeletal muscles
Breathing exercises
Stressed breathing
Autogenic training
Stressed body focus
Imagery and positive selfstatements
Stressed emotion
Meditation and
mindfulness
Stressed attention
Note. Many popular approaches are blends of universal family
groups. Thus, a variant of “hatha yoga” may blend stretching,
breathing, and meditation. Jacobson’s original version of PMR
in fact incorporated minimal tense–let go exercises, an occasional stretch, and breathing. What is called “mindfulness”
is in fact often mindfulness and breathing with a little yoga
(and occasional autogenic suggestion) sprinkled in. Unfortunately, such blends make it difficult, if not impossible, to interpret relaxation. If a popular version of mindfulness works on
depression, was it the mindfulness itself, or the accompanying
breathing, stretching, or autogenic exercises? Yoga research is
most susceptible to such confounds.
12. Relaxation Today
family group works only for one type of symptom,
that PMR works only for skeletal muscle tension.
Instead, the type of self-stressing associated with a
type relaxation is (1) part of the training rationale,
(2) incorporated in a procedural self-unstressing
strategy, and (3) an initial exercise effect. So a
practitioner of PMR may read that PMR targets
the muscles through a tense–let go rebound effect
(a typical rationale), be instructed to “focus on
your shoulder muscles and let go” (a beginning
procedural instruction focusing on skeletal muscles), and immediately experience shoulder relaxation after shrugging and releasing tension (initial
effect). However, in time, many practitioners discover that all universal family groups of relaxation
can eventually address other components of selfstressing. The student of PMR may notice that she
is breathing deeper (relaxed breathing), engaging in pleasant fantasy (relaxing imagery), and
focusing more easily (meditation/mindfulness). A
skilled trainer may incorporate these additional
effects into training by adding techniques from
the universal family groups of breathing, imagery,
and meditation/mindfulness exercises.
Trainers often blend universal relaxation strategies. One might include breathing and imagery with PMR. Hatha yoga is often a mixture of
stretching, breathing, and meditation. Mindfulness, although often defiantly presented as a
pure approach, is usually a blend of mindfulness,
breathing, and often a touch of imagery. Perhaps
masters of relaxation intuitively recognize the
value of combining family groups. However, they
are usually clueless as to why various combinations work or which other combinations may work
better. To give an amusing example, a few years
ago I suggested that the nearly universal practice
of preceding mindfulness training with breathing
exercises is based on religious tradition rather than
empirical evidence. The preparatory value of other
approaches, perhaps autogenic training or PMR, is
a question worth exploring. World-famous mindfulness researchers reacted to my observation with
hostile derision (Smith, 2004). This occurred in
spite of the fact that at least one famous instructor
indeed unwittingly (mindlessly?) slips a few autogenic suggestions into his mindfulness training
(Kabat-Zinn, 1990).
Self-stressing theory serves two very important purposes. First, it helps the client simplify
a confusing universe of hundreds of relaxation
techniques. For example, there are thousands of
exercises described as “yoga.” However, they are
not the same. Hundreds are part of the universal
191
family group of stretching. To this family group
one might also include a wide range of Western
approaches not derived from yoga. Furthermore,
some exercises described as “yoga” actually belong
to the “breathing” or “imagery” family group. Second, the self-stressing map gives the practitioner a
powerful tool. Knowing the potential differences
among relaxation strategies, one can more clearly
consider how each fits relaxation goals and how
exercises might be combined. For example, a biofeedback client attempting to reduce pain associated with muscle tension might explore autogenic
warmth and heaviness exercises. If these happen
to have limited effect in reducing muscle tension,
our client may consider the map and add mindfulness meditation. Both autogenic and mindfulness
strategies might then be combined with peaceful
imagery.
Road Map 2: Psychological Relaxation Theory
It is difficult to travel the terrain of relaxation
without landmarks and road signs. These let us
know whether we are going in the desired direction and suggest changes we might make. In psychological relaxation theory, the landmarks and
signs of relaxation are psychological states of mind,
that is, client experiences. The idea is not new. For
biofeedback practitioners, client self-reports often
provide verbal feedback and confirmation of the
effect of feedback. For example, a client may report,
“My hands felt warm (self-report) the moment my
monitor changed color and the tone lowered (biofeedback signal).” However, the vocabulary of selfreport used by clinicians has been limited to a few
words (e.g., “Lift your finger if you feel ‘relaxed,’ ”
or “Let the words ‘warm and heavy hands’ float
through your mind”). I have long believed that
practitioners can tell us much much more.
For much of my professional life I have explored
the natural language that practitioners use to
report relaxation states. I began by searching
over 200 core instructional textbooks for diverse
traditions, East and West, old and new, spiritual
and secular. These included yoga, PMR, breathing, contemplation, imagery, prayer, meditation,
mindfulness, tai chi, imagery, self-hypnosis, and
autogenic training. My goal was to generate a
comprehensive lexicon of “relaxation,” psychological states associated with technique practice. My
initial relaxation dictionary contained about 400
relaxation words, which I later reduced to about
200. At this time, it became clear that identify-
192
ing the structure of the contents of this universe
would provide a psychological map of relaxation
experience.
Over the past 20 years, my colleagues, students,
and I have subjected variations of this lexicon to
nine factor-analytic studies involving 6,077 participants and over 40 relaxation techniques and
activities (Smith, 2006, 2007a). Patterns and
constellations of relaxation words have clearly
emerged from this vast linguistic expanse. Today
(Smith, 1999, 2001, 2015), I identify 19 relaxation
states, or “R-states” (or alternatively “M-States” for
“mindfulness states”). Twelve R-states emerged as
factors. To these I identified as R-states the words
that correlated uniquely and separately with various personality questionnaire variables including
those from the NEO Personality Inventory—
Revised (NEO PI-R; Costa & McCrae, 2012) and
the 16 Personality Factor Questionnaire (16 PF;
Cattell, Cattell, & Cattell, 1993). Also, if a set of
words differentiated the effects of two techniques
in a comparative pretest–posttest outcome study,
those words were identified as an R-state. Two
R-States (“Quick to Detect Mind Wandering” and
“Easy to Let Go” are based on relaxation theory
(Smith, 2015). Currently, (Smith, 2015) I group all
into five levels as follows:
Outline of Five Levels and 19 R‑State2
Word Groups
Level 1: Basic Relaxation
• R-State Disengaged (“Feeling distant, far away,
detached”)
• R-State Muscles Relaxed (“Body comfortable,
breathing easy”)
• R-State At Ease (“Peaceful, refreshed”)
Level 2: Basic Mindfulness
• R-State Aware (“Focused, clear”)
• R-State Centered (“Absorbed, grounded”)
• R-State Deepening (“Sense of ‘going deeper,’
‘things are changing’ ”)
• R-State Quiet (“Still, few thoughts”)
• R-State Accepting (“Accepting what I can't
have or change, ‘let it be,’ ‘it is what it is’ ”)
• R-State Quick to detect mind wandering (“Easy
to notice mind wandering or distraction. Catch
it early”)
• R-State Easy to let go and refocus (“Easy to let
go of mind wandering. Not stuck or caught up
in distraction”)
IV. RELAXATION INTERVENTIONS
Level 3: Mindful Flow and Change
• R-State Curious (“Interested. Things seem
new.”)
• R-State Savoring (“Enjoying each moment”)
• R-State One step at a time (“Each moment
comes and goes”)
Level 4: Positive Emotion
• R-State Happy, Optimistic, Trusting
• R-State Loving, Caring
• R-State Thankful, Grateful
Level 5 Mindful Transcendence
• R-State Awe and Wonder. Mystery.
• R-State Prayerful, Reverent
• R-State Timeless, Boundless, Infinite, At One
Clinical Applications
I find this map useful for comparing relaxation
techniques and assessing client progress. Research
has already identified several useful patterns
(Smith, 1999). For example, a client who reports
the R-states Disengaged and Muscles Relaxed
apparently has accessed a limited range of R-states,
a pattern I find typical for novice students of relaxation and those who benefit from PMR. Anxious
and depressed clients, or those under severe stress,
consistently report Disengaged as an R-state they
seek when practicing relaxation.
R-states are not restricted to one or two groups
of relaxation. One who reports the R-states Quiet
and Accepting (perhaps while practicing autogenic exercises) is describing states often reported
by practitioners of meditation or mindfulness.
Here one might augment autogenic training with
meditation or mindfulness exercises. Similarly, one
who reports feeling surprisingly Happy, or Loving,
is describing a cluster of R-states that are common among those who benefit from vivid visual
imagery, suggesting another strategy that might be
explored.
Schwartz (2003) provides a strong case for
home practice of relaxation using recorded exercise instructions. I use a variation of this in my
self-guided online relaxation program (Smith,
2015; http://sites.roosevevelt.edu/jsmith). Clients can
download 20-minute MP3 instructions for each
of the six universal family groups of relaxation
(stretching, PMR, breathing, autogenic exercises,
12. Relaxation Today
193
imagery, meditation and mindfulness). In several
ways these exercises differ from others that are
available. Most important, I present “pure” versions
so clients can identify R-states associated with each
strategy. For example, PMR exercises emphasize
“tense–release” cycles, with a minimum of breathing, autogenic suggestion (“Feel the warmth as your
hands relax”), or imagery. This way, one knows
whether any resulting R-states are the result of tensing and releasing rather than some other approach.
In recent books (Smith, 2006, 2015) I help
users explore the six family groups of exercises in
the context of mindfulness. My structured programs provide essential rationales and step-by-step
instructions. Short validated daily questionnaires
and diaries for assessing R-states are offered for
each approach. These sensitize clients to subtle
effects and provide an ongoing record whereby client and clinician can compare and contrast techniques.
• I try to present a relatively “pure” version of each
approach, so that clients can discover its unique
effects independently of other approaches. For
example, I present PMR tense–let go cycles,
minimizing breathing and imagery, so clients
can clearly note the relaxation effect of tensing
up and letting go.
• Once a client is trained in a variety of
approaches, we select those that work best and
construct an individualized script and tape.
Rather than present exercises as a mechanical
fitness routine (as one might find in a gym),
they are artfully integrated into a sequence with
internal structure and meaning that expresses a
client’s relaxation goals and aspirations.
• The goal of relaxation training goes beyond
the relaxation response of lowered arousal.
Additional objectives are cultivating appropriate R-states and acquiring beliefs and personal
philosophies conducive to deepening relaxation
and extending its rewards to all of life.
Relaxation Scripting
My approach is to craft an individualized relaxation recording based on a mutually developed,
verbatim script of exercise instructions. There are
several advantages to such script writing:
Over the last two decades, I have introduced
relaxation to thousands of individuals. They have
taught me one very unexpected lesson. I used to
think that most clients have very specific relaxation preferences and respond best to just one
general strategy, whether it be PMR, stretching,
meditation, or the like. I spent many years of
research attempting to identify “PMR,” “yoga,” or
“meditation” types of people. So far, I have found
few patterns. All this time, my clients were teaching me something very different; virtually all of
them preferred highly individualized mixtures of
many approaches. Very few people prefer just one
or two strategies alone. This discovery has led me
to develop a new approach to teaching relaxation:
scripting (Smith, 2007b).
I find scripting to be a powerful strategy for clients, although it is a bit time-consuming. Perhaps
more important, it is an excellent tool for teaching health professionals the diversity of relaxation
strategies, as well as the practical side of our two
road maps of relaxation and mindfulness. “Scripting” is based on the following ideas:
• Different approaches to relaxation have different effects and work differently for different
people.
• The best way to teach relaxation is not to
impose one or two approaches on everyone, but
to introduce a variety of approaches.
• Since the client is inventing his or her own
relaxation exercise, he or she is more likely to
take it seriously and practice it regularly. Indeed,
a client may well treasure his or her script as a
truly personal possession, and practice it very
seriously.
• Given that training is varied and changing,
interest and motivation are maintained, reducing premature quitting.
• Finally, relaxation can be used as a reminder of
personal philosophies conducive to living a life
of peace and calm.
Relaxation scripting takes at least 10 sessions
(although I offer variations that can be offered
in as few as one or two sessions in both individual and group formats—even online or over the
phone). The first seven sessions are devoted to an
orientation and training each of the six families
of relaxation. In sessions eight through —10, one
creates a script based on the exercise elements that
worked best. A script is not simply a hodgepodge
of exercises, or a workout sequence. Instead it is
an artfully crafted blend of exercises guided by
a unifying idea and targeted to client needs and
interests. Client-selected R-states are inserted to
enhance the effects of training (e.g., example,
194
using the words “warm and heavy” with PMR, or
“focused and clear” with meditation).
Care is taken to create a package of exercises
that flow, ranging from those that are complex,
active, and energized to those that are simple, passive, and calm. Thus, one might begin with active
PMR or stretching, move to breathing, and end
with autogenic suggestion and imagery, and eventually meditation. Similarly, the onset of a script
may incorporate R-state words such as “alive,”
“tingling,” and “release.” Mid-sequence words
may include “distant,” “far away,” and “limp.” Passive and focused concluding words might include
“focused,” “peaceful,” or even “reverent.”
Once client and trainer have agreed on a script,
both collaborate on an audio recording. This is a
unique feature of relaxation scripting, one that
departs from the common practice of using a
generic calm “relaxation voice” whenever speaking
relaxation instructions. Specifically, I pay particular attention to using a “relaxation voice” appropriate for each family group of techniques; that is,
each approach to relaxation has its own suggestive
voice tone and pattern. Briefly, the “tense up” components of PMR are spoken in a tone of voice that
is somewhat raised, like that of a gym coach. In
contrast, the following “let go” sequence of words
is voiced in a quiet and slow monotone. The voice
appropriate for yoga stretching is slow and somewhat energized, emphasizing the act of stretching
(as if one were “stretching” with one’s voice). The
breathing voice emphasizes the flow of breath, with
phrases paced with the slow and deep pace of the
trainer’s flow of breath. For example, as a trainer
says, “Slowly let the air flow out through your lips,”
he or she actually gently exhales while speaking the
instruction. The voice for imagery is somewhat colorful, a gentle “storytelling voice” as opposed to a
monotone. In contrast, a meditative voice is clear,
calm, slow, and simple. It has little inflection.
When teaching relaxation to clinicians, I devote
several weeks to cultivating the voice pattern most
appropriate for each family of relaxation. Often
trainers have to overcome the mistaken tendency
to use the same generic “soft relaxation voice” for
all relaxation techniques. Some voice patterns are
particularly difficult to master, for example, the
slow and colorless monotone I suggest for PMR
“let go” phrases, as well as autogenic training. For
more on suggestions for relaxation voicing see my
CD Relaxation Voice Training Program (http://
drsmith.deltalprinting.com). All of the free MP3
relaxation downloads I offer (http://blogs.roosevelt.
edu/jsmith) utilize my voice pattern suggestions.
IV. RELAXATION INTERVENTIONS
Conclusion
Standardized relaxation recordings, MP3 downloads, and YouTube presentations provide an exciting new world of opportunities for streamlining and
enhancing relaxation and biofeedback. However, I
think it is a mistake to give all clients the same
set of instructions and hope they practice until
it works. Until now, the alternative was often to
invite clients to wander untrained and unequipped
into uncharted lands of confusing claims and techniques. Self-stressing and psychological relaxation
theory suggest a new way. Give your client what
you truly believe is the best initial approach, say,
PMR. Feel free to use a CD, a DVD, a download, a
book, or a group workshop. Then provide the road
maps so your client can intelligently enhance and
extend what you have taught. Self-stressing theory
simplifies thousands of techniques into six universal family groups. For example, if you began with
PMR, your client might discover that the family
group meditation/mindfulness, or maybe imagery,
helps it work better and contributes to generalization to life at large. Psychological relaxation theory teaches that relaxation training can go beyond
the relaxation response. Clients may discover that
PMR can do more than relieve stress and physical
symptoms. Creatively enhanced, it can be a source
of energy and insight, a productivity tool, something creative and fun, and indeed even a form of
spiritual expression. Indeed, these additional goals
may reinforce the impact of technique on stress
and symptoms, as well as contribute to compliance
and generalization to life at large. In summary, I
suggest starting with what you do best. Instruction
does not end when your sessions are over. Properly
trained, your client is ready to step forth into a new
world.
Notes
1. In response to the inevitable question, “What about
biofeedback? Jogging? Massage? Listening to music? Aren’t
they all approaches to relaxation,” I refer the reader to
Smith (2006), where I explore the distinction between
self-relaxation, assisted relaxation, and casual relaxation.
“Self-relaxation” includes techniques one can do by oneself,
without outside mechanical, electrical, human, or animal
assistance (petting pets). I would include relaxation techniques that can be learned by reading a book or listening to
or watching recordings. “Assisted relaxation” requires the
initial or continued support of assists (biofeedback, massage, saunas, etc.). “Casual relaxation” includes everyday
activities that require neither assistance nor professional
12. Relaxation Today
training and may have relaxation as a side effect. These
include exercise, listening to music, taking walks, chanting, group prayer, having sex, or reading my books on the
paranormal (Smith, 2010).
2. Formal R-states are capitalized.
References
Cattell, R. B., Cattell, A. K., & Cattell, H. E. (1993). The
16PF Questionnaire (5th ed.). Champaign, IL: Institute
for Personality and Ability Testing.
Costa, P. T., & McCraw, R. R. (2012). NEO Personality
Inventory—Revised (NEO PI-R). Lutz, FL: Psychological
Assessment Resources.
Kabat-Zinn, J. (1990). Full catastrophe living. New York:
Delta.
Lehrer, P. M., Woolfolk, R. L., & Sime, W. E. (Eds). (2007).
Principles and practice of stress management (3rd ed.).
New York: Guilford Press.
Schwartz, M. S. (2003). The use of audiotapes for patient
education and relaxation. In M. S. Schwartz & F.
195
Andrasik (Eds.), Biofeedback: A practitioner’s guide (3rd
ed.). New York: Guilford Press.
Smith, J. C. (1999). ABC relaxation theory: An evidencebased approach. New York: Springer.
Smith, J. C. (2001). Advances in ABC Relaxation: Applications and inventories. New York: Springer.
Smith, J. C. (2004). Alternations in brain and immune
function produced by mindfulness meditation: Three
caveats. Psychosomatic Medicine, 66, 148–152.
Smith, J. C. (2006). Relaxation, meditation and mindfulness:
A guide for health professionals. New York: Springer.
Smith, J. C. (2007a). The psychology of relaxation. In P. M.
Lehrer, R. L. Woolfolk, & W. E. Sime (Eds.), Principles
and practice of stress management (3rd ed., pp. 38–52).
New York: Guilford Press.
Smith, J. C. (2007b). The relaxation, meditation and mindfulness essential self-training manual. Charlotte, NC: Lulu
Press.
Smith, J. C. (2010). Pseudoscience and extraordinary claims
of the paranormal: A critical thinker’s toolkit. New York:
Wiley-Blackwell.
Smith, J. C. (2015). Mindfulness reinvented and the M-Tracker
Method. Charleston, SC: Createspace.
CH
hA
aPT
tER
r 13
Cardiorespiratory Biofeedback
Richard N. Gevirtz, Paul M. Lehrer,
and Mark S. Schwartz
In this chapter we describe the biofeedback, or
applied psychophysiological, methods that have
been using cardiorespiratory signals. Although
this chapter is not intended as a literature review,
we cite research evidence as a guide.
breathing therapy with good patient education,
other relaxation procedures, and cognitive- or
mindfulness-based approaches. The specific need
for biofeedback-assisted breathing therapy (e.g.,
respiration rate, diaphragmatic breathing, CO2
feedback, and volumetric feedback) also remains
logical, although unsubstantiated.
Psychophysiological measurements and feedback are valuable at least for therapist information,
documentation, and information and for patient
motivation and confirmation.
RESPIRATORY FEEDBACK
Using feedback from various instruments, or just
instructing clients on breathing techniques, can
be helpful for treating a number of disorders.
Hyperventilation Syndrome
Functional Chest Pain and Functional
“Cardiac” Symptoms
There is a surprising paucity of studies using
breathing therapies alone for hyperventilation syndrome (HVS). The available reports and studies of
breathing therapy alone for HVS do show clinically
significant reductions of HVS symptoms (Fried &
Grimaldi, 1993; Grossman, De Swart, & Defares,
1985; Timmons & Ley, 1994). However, methodological problems and equivocal results of studies
detract from firm conclusions. Studies of cognitive
restructuring, relaxation, feedback for respiration
rate, and patient education indicate that these
approaches also lead to significant improvements
(Bass, 1994; Timmons & Ley, 1994).
Although using breathing therapies alone
for HVS remains a logical and sound approach,
practitioners probably will continue to combine
There are many patients with chest pain, without
positive cardiological findings. Often there is no
objective or probable organic cardiac pathology
explaining these symptoms. One correctly assumes
that psychophysiological factors play a role in these
symptoms for many patients (Clouse, 1992; Hegel,
Abel, Etscheidt, Cohen-Cole, & Wilmer, 1989).
However, chest pain associated with symptomatic hyperventilation is not “psychogenic” in the
sense of being purely cognitive in origin. There are
physical reasons, albeit often psychophysiological
ones, for the symptoms. Muscle tension, spasm,
and fatigue in the intercostal muscles constitute
one such mechanism (Bass, Gardner, & Jackson,
1994). Many physicians refer to this musculoskeletal explanation as “chest wall pain.”
196
13. Cardiorespiratory Biofeedback
Some physicians suggest that a distended stomach caused by aerophagia (air swallowing) places
excess pressure on the diaphragm and can cause
chest pain (Bass et al., 1994). Diaphragmatic
spasms may create chest symptoms. For some, dysfunctional esophageal involvement contributes to
these symptoms (Clouse, 1992). Here, too, psychophysiological factors often play a role, as noted by
Drossman et al. (1990). An international panel of
clinical investigators provided a preliminary consensus report of functional gastrointestinal disorders. Included among these disorders was “functional chest pain of presumed oesophageal origin”
(Drossman et al., 1990, p. 163; i.e., “midline chest
pain with or without dysphagia for at least three
months; and no evidence for oesophagitis, cardiac
or other disease to explain symptoms” [p. 163]).
Although esophageal disorders are common in
these “chest” symptoms, the authors recognized
the etiological potential for psychological factors.
In assessing chest pain, one also must consider reduced blood flow to the heart (Fried &
Grimaldi, 1993); hyperventilation (HV) can trigger paroxysmal vasospasms in the heart (and the
brain). The effects of HV on cardiac functioning
are not in question. They are real and are accepted
by almost all experts. In fact, HV and high arousal
contribute to many occurrences of cardiac symptoms of organic origin, such as angina pectoris and
infarction (Nixon, 1989; Fried & Grimaldi, 1993).
However, a question of clinical significance for
practitioners is whether the cardiac changes with
HV indicate an organic cardiac diagnosis. They
often do not; just as saying that breathing changes
and thoughts affect various parts of the brain does
not mean that a patient has an organic brain disorder.
Another question of clinical significance is
whether there are noncardiac reasons to explain
functional chest pain that may appear to patients
(and practitioners) as having a cardiac origin. The
answer is “yes.” There are musculoskeletal, diaphragm-, and esophagus-related causes of chest
pain that are often stimulated and provoked by
psychophysiological factors. Many people are
hypervigilant and focus on their bodily sensations
more keenly than is needed, or desired. These
people may catastrophize these sensations and
attribute dire causes to sensations and symptoms
that are not at all dangerous, or that are within
the range of normal physiological sensations and
events. Such attributions can result in cognitive
anxiety and worry, as well as physical musculoskeletal tension and autonomic nervous system (ANS)
197
arousal. These, in turn, can lead to, or accompany,
the physical changes that produce the chest symptoms.
Treatments of potential value for functional
chest pain include psychopharmacological therapy,
such as low-dose antidepressants (Clouse, 1992);
cognitive therapy (Salkovskis, 1992); behavioral
therapies similar to those for chronic pain (Bradley, Richter, Scarinci, Haile, & Schan, 1992); and
breathing therapies (DeGuire, Gevirtz, Kawahara,
& Maguire, 1992; DeGuire, Gevirtz, Hawkinson,
& Dixon, 1996; see also the review by Garssen, De
Ruiter, and van Dyck, 1992). To date, there are no
studies using the combination of these therapies,
and no studies comparing these therapies.
Here we focus on the study by DeGuire et al.
(1992), in which breathing therapies and patient
education led to significantly reduced symptoms
among patients with functional cardiac symptoms
who showed signs of HV. Patients received one of
three breathing therapies: (1) without physiological feedback, (2) with visual biofeedback from
thoracic and abdominal strain gauges, or (3) with
ETCO2 capnometer feedback. A control group
receiving no therapy was also studied. Therapy
that was common to all patients in the three
treated groups included the following:
1. Verbal patient education about respiratory
physiology and the hypothesized effect of HV
on functional cardiac symptoms.
2. Instructions for diaphragmatic breathing, and
the therapist’s demonstration of diaphragmatic
breathing.
3. Office practice and correction of errors.
4. Encouragement to persist and reassurance
about the expected decrease in patients’ often
reported discomfort caused by slow diaphragmatic breathing.
5. Encouragement to avoid increasing tidal volume, or amount of air inhaled, to compensate
for changes in rate of respiration.
6. Encouragement of a slow-paced rate of respiration (less than 14 beats per minute).
7. Encouragement to practice this slow pace during conversations and while visualizing situations in which patients were having problems
maintaining the new breathing.
In the two conditions that included feedback,
physiological monitoring and feedback typically
occurred in later sessions.
Several criteria for improvement included
the number of days with symptoms and the fre-
198
quency of cardiac symptoms. Patients completed
self-reported ratings of symptom frequency and
severity. There were six treatment sessions over
3 weeks. The study partly based improvement on
symptom changes between a 2-week baseline and
the 2 weeks after treatment. The three breathing
therapy groups reduced the days on which symptoms occurred from 8–to 10 of 14 days down to 4–5
of 14 days. In contrast, the control group showed
no drop from 10 of 14 days with symptoms.
The treated groups reduced their frequency of
cardiac symptoms from an average of 21 symptoms
(range 15–25) down to about 9 (range 4–15). All
three groups showed reduced symptoms compared
to the control group, which did not change. After
treatment, the three treatment groups decreased
reported symptoms from 23 to four (without
physiological feedback), from 15 to nine (with
visual biofeedback from thoracic and abdominal
strain gauges), and from 25 to 15 (with ETCO2
capnometer feedback), compared to the control
group, which only went from 27 to 26 symptoms.
The authors note that they did not measure duration of symptoms, and that this omission resulted
in underestimation of the effect. For example,
some patients recorded one episode of chest pain
all day; others reported only episodic mild pain.
Those particular patients thus appeared to show
increased symptoms rather than improvement,
which would be a more accurate interpretation.
ETCO2 increased significantly and respiratory rate
dropped (from 16.5 to 8.5 beats per minute) for
the three active treatment groups compared to the
control group, which showed no change from 15
beats per minute.
The breathing therapies led to ETCO2 increases.
They started with 34–38 torr and increased to 39
to 41 torr, showing a more normal level of CO2.
Reduction of cardiac symptoms and increased
ETCO2 occurred to a greater extent among those
with reduced breathing rate. The reduced respiration rate correlated with reduced frequency of
symptoms (r = .53, p < .001), number of days with
symptoms (r = .59, p < .001), and increased ETCO2
(r = .38, p = .018). DeGuire et al. (1992) noted
that with “the use of end-tidal CO2, abdominal
and thoracic strain gauge monitors did not add
significantly to either reducing the frequency of
cardiac symptoms or facilitating changes in physiology” (p. 676).
However, there was a trend for patients using
strain gauge monitors with computer-based visual
biofeedback to produce the largest effect. The
authors caution that the extent to which physi-
IV. RELAXATION INTERVENTIONS
ological feedback information added to the new
attributions about symptoms is unclear. In a follow-up study, 40% of the patients (representative
of the entire sample) were contacted 3 years later.
Reported symptom reductions were maintained, or
improved. Similarly, ETCO2 and respiration rate
measures remained in the normal range (DeGuire
et al., 1996).
Practitioners should consider including breathing therapies for patients diagnosed with functional chest pain and functional cardiac symptoms.
However, functional chest symptoms and HVS are
not synonymous (Bass, 1994). These authors also
remind us that HV only provokes chest pain in less
than half of the patients assessed. Even so, using
breathing therapy for these patients is logical,
considering the successful use of this therapy for
many patients with panic symptoms and the frequency of panic symptoms among these patients
(see below).
We can also note the potential benefit of relaxation therapies with breathing therapy during
cardiac rehabilitation for patients after myocardial
infarctions (van Dixhoorn, Duivenvoorden, Staal,
Pool, & Verhage, 1987; van Dixhoorn, Duivenvoorden, Staal, & Pool, 1989; Duivenvoorden &
van Dixhoorn, 1991). These authors report fewer
second coronary events, rehospitalizations, unstable angina episodes, and other serious cardiac
events. van Dixhoorn (2007) presented variations
of breathing therapy that emphasize attentional
states and total body involvement.
van Dixhoorn and Duivenvoorden (1999)
randomly assigned 156 patients with myocardial
infarction to a standard physical training or to
the same training with an additional component
of breathing awareness and retraining therapy (six
sessions). The breathing group failed physical testing less often, returned to work more often, lowered respiration rates, increased cardiac variability,
and had fewer cardiac events at a 2-year follow-up.
Seventeen of 76 patients had a significant reinfarction (five patients died) in the breathing group,
while 29 of 80 patients in the exercise-only group
experienced similar poor outcomes (seven findings
are consistent with other trials around the world
(Blumenthal et al., 1997; Patel et al., 1985).
Panic
Breathing therapy constitutes a basic part of current treatment for many patients with panic symptoms and panic disorder. The debate continues as
to whether HV causes panic or merely accompa-
13. Cardiorespiratory Biofeedback
nies it, in some patients with panic symptoms.
Ley (1993) is a prolific, tenacious, and persuasive
advocate of the role of HV in panic and the necessity of breathing therapy for these patients (see
Ley, 1991, 1992). Contrary views by Garssen et al.
(1992) and Clark and his colleagues (Salkovskis
& Clark, 1990) provide balance and perspective.
Ley’s (1993) proposal helps resolve some differences in opinion.
Some researchers report more benefit from
respiratory control with progressive exposure to
situations in which symptoms occur, compared
to only exposure, without breathing therapy.
However, breathing therapy often leads to mixed
results, whether or not one includes cognitive
components. This contrary view is that “recent
studies do not support the idea that HV is an
important causal mechanism in producing panic
attacks” (Garssen et al., 1992, p. 149). Rather,
according to this view, HV accompanies panic in
some patients with panic. Garssen et al. concluded
that methodology problems make it impossible to
derive interpretations and final conclusions on
the specific role of breathing therapy. Examples of
these problems include combining therapies, very
small samples, and lack of controls. The specificity
and mechanism of breathing therapy are unclear
and elusive.
The majority of the studies point to a therapeutic effect of breathing retraining and cognitive reattribution of physical symptoms to HV for
patients with HVS and the closely related panic
disorder, with or without agoraphobia. This conclusion seems warranted in that both help alleviate anxiety in patients with HVS or related disorders (Garssen et al., 1992).
However, Ley (1992, 1993) provides a studious
resolution to the disagreement about the relationship of HV and panic. He has proposed three types
of panic attacks (PAs; Ley, 1992). He calls the classic PA “Type I” or “classic PA.” He distinguishes
this from “anticipatory PA (Type II)” and “cognitive PA (Type III).” The classic Type I PA has distinctive and objective physiological features, especially compared to cognitive PA, Type III. These
features include (1) sharp drops in pCO2 (> 10 mm
Hg), (2) sharp increases in respiration rate and/or
tidal volume, (3) sharp increases in heart rate (>
10 beats per minute), (4) sharp increases in electrodermal activity, and (5) low finger temperature
(< 80°F).
Two more studies support Ley’s position. Biber
and Alkin (1999) divided 51 patients diagnosed
with panic disorder into two groups: those with
199
predominantly respiratory symptoms and others. The respiratory group had more sensitivity to
inhaled CO2, scored higher on panic and anxiety
scales, and had longer duration of illness. Moynihan and Gevirtz (2001) divided patients with
panic disorder by symptoms as described earlier
and compared the respiratory group with the cognitive group on a number of respiratory parameters
during various conditions. As expected, the respiratory group had lower ETCO2 (especially during
a stressor) and more rapid respiratory rates, both of
which are risk factors for panic.
Ley (1992, 1993) suggests that practitioners
should expect the most benefit from breathing
therapy for patients with Type I symptoms and
expect the most benefit from cognitive therapies
for Type III. One should read at least these two
references to appreciate his position.
More recently, a group of researchers that originated at Stanford University produced studies that
support the importance of dysregulated breathing
in panic and the use of breathing therapies in treatment (Roth, Wilhelm, & Trabert, 1998; Meuret,
Wilhelm, Ritz, & Roth, 2003, 2008; Meuret, Ritz,
Wilhelm, & Roth, 2005; Meuret, Rosenfield, Hofmann, Suvak, & Roth, 2009; Meuret, Rosenfield,
Seidel, Bhaskara, & Hofmann, 2010).
Meuret et al. (2009) completed a study comparing five sessions of biofeedback in a 1-month
period, using an ambulatory capnometer, to a
delayed treatment wait-list group. Twenty participants were assigned to the biofeedback treatment
and 17 to the delayed treatment.
The treatment had five major components: (a) educating patients about the role of breathing in the
etiology and maintenance of PD [panic disorder],
(b) directing their attention to potentially problematic respiratory patterns, particularly those observed
during the extended physiological monitoring, (c)
having them perform different breathing maneuvers with capnometer feedback to experience how
changes in breathing affect physiology, symptoms,
and mood, (d) teaching them ways to simultaneously
control pCO2 level and RR [respiration rate], (e) and
having them practice breathing exercises daily. The
weekly sessions were aimed at reviewing changes in
pCO2 and RR along with changes in symptoms and
emotions. Individual training exercises, to be performed twice-daily for 17-min, at home or elsewhere,
consisted of three parts: (a) a baseline period (baseline), during which patients sat quietly with their eyes
closed for 2-min, (b) a 10-min paced breathing period
(paced) during which patients breathed in synchrony
with tones while occasionally checking their pCO2
and RR on a feedback device, and (c) a 5-min
200
breathing period without pacing tones during which
patients were to maintain their previously paced RR
and pCO2 level using the feedback device (transfer).
The paced breathing was used as a guiding tool to
gradually shape slower breathing across treatment
weeks. Patients were instructed to gradually adjust
their breathing patterns (RR, rhythm, and depth)
to reach or maintain pCO2 in a normocapnic range
(pCO2 > 35 mmHg; (Oakes, 1996). In the first two
weeks the emphasis was on stabilization of breathing patterns (RR and rhythm), while in the last two
weeks the emphasis shifted to normalizing pCO2.
For the minority of patients (n = 11) with normocapnic pCO2 levels, treatment focused on regularity
of breathing (regular rate, avoiding of sigh breaths).
(Meuret et al., 2008, p. 4, emphasis in original)
Compliance and attendance were excellent
(91% of homework completed), with only a few
dropouts at the follow-ups. Results suggested that
treated participants could raise pCO2 levels (32–37
mmHg) and dramatically reduce panic severity
scores (PDSS, 2–4 on a 4-point scale), depression,
and anxiety sensitivity. Effect sizes ranged up to d
= 2.2, a large effect compared to cognitive-behavioral therapy (CBT) effects (Hofmann & Smits,
2008).
In a subsequent study, Meuret, Hofmann, and
Rosenfield (2010) and Meuret and Ritz (2010)
indicated that respiratory factors mediate symptom improvement as well as (if not better than)
cognitive changes in therapy. Taken as a whole,
this literature justifies the use of biofeedback in
panic treatment.
In subsequent studies, Meuret and colleagues
have replicated the efficacy of the previous results
(Meuret et al., 2008) and showed that decrease
in panic symptoms were mediated by respiratory
changes to a greater degree than changes in dysfunctional cognitions (Meuret, Rosenfield, et al.,
2010).
Heart Rate Variability Biofeedback
The Importance of ANS Parameters in Health,
Illness, and Performance
Note: Excellent papers describing the phenomenon of HRV are available (see for example: Bernardi et al., 1994; Berntson et al., 1997; Berntson,
Cacioppo, & Quigley, 1993; Berntson, Cacioppo,
Quigley, & Fabro, 1994; Grossman & Taylor,
2007). A theoretical perspective is provided by
Stephen Porges (2011).
IV. RELAXATION INTERVENTIONS
Heart rate variability (HRV) feedback has
received a lot of attention in recent years because
of the emerging realization that many disorders
found in Western health care settings may be
mediated by autonomic processes. Table 13.1 presents some of these disorders and lists the probable
or suspected autonomic mediator. Since psychological processes (stress, anger, depression, worry,
rumination, etc.) are known to affect autonomic
processes, psychological etiology is often implied
for many of these disorders (usually ignoring the
psychophysiological pathways; Gevirtz, 2007).
Table 13.1 summarizes some of these relationships.
As can be seen, a number of prominent disorders
that are treated by biofeedback practitioners have
autonomic mediators. For this reason, HRV analysis can be useful in assessing the psychophysiological profile of the patients. Low HRV can be seen in
both “time domain” and “frequency domain” measures, based on the time intervals between R-spikes
in the electrocardiogram. Each of these measures is
calculated exclusively from “normal” cardiac interbeat intervals (those produced by sinus rhythms,
under neural control, but excluding particularly
long or short intervals caused by valve malfunctions, abnormal cardiac arrhythmias, etc.). Time
domain measures include standard deviation of
normal to normal R-wave duration (SDNN) or,
for 5-minute intervals, the standard deviation of
the average NN intervals (SDANN), the percent
of successive normal interbeat intervals differing
by 50 milliseconds or more (pNN50), or the root
mean square of successive differences (RMSSD).
Frequency domain measures are calculated from
spectral analysis of the electrocardiogram signal
and include high-frequency (HF) variability, 0.15–
0.4 Hz, reflecting respiratory influences on heart
rate and vagal (parasympathetic) control, low-frequency (LF) activity, 0.05–0.15 Hz, reflecting control of blood pressure by the baroreflex and probably influenced by both the parasympathetic and
sympathetic systems, and the LF/HF ratio, which is
often interpreted as reflecting autonomic balance.
Note that various measures of HRV reflect two
separate processes in the body, and, although the
effects of these may overlap, they probably are not
one and the same. One process is autonomic balance. Thus, abnormal processes associated with
either sympathetic or parasympathetic hyperarousal or hyperreactivity tend to be reflected
in increases or decreases in various HRV parameters. The other is the activity of various regulatory reflexes. Two of these are respiratory sinus
201
13. Cardiorespiratory Biofeedback
TABLE 13.1. Explaining “Unexplained” Medical Symptoms
Symptom cluster
Typical diagnoses
Potential mediators
Level of evidence
Local muscle pain
Lower back pain (LBP),
cervical strain, tension
headache, repetitive
strain injury (RSI), etc.
Sympathetically modulated
trigger points
Herbs, Gevirtz, & Jacobs
(1994)
Abdominal
pain, diarrhea,
constipation, bloating
Irritable bowel syndrome
(IBS), recurrent
abdominal pain (RAP)
Autonomic imbalance with
excessive sympathetic tone and
prolonged vagal withdrawal
Sowder, Gevirtz, Shapiro,
& Ebert (2010); Mayer et
al. (2001)
Posttrauma: arousal,
reexperiencing
dissociation
Posttraumatic stress
disorder (PTSD)
Cortical overload, vegetative
vagus, limbic reshaping
van der Kolk (2001, 2006);
Lanius et al. (2002, 2006);
Zucker et al. (2009)
Constant worry,
inability to relax
Generalized anxiety
disorder (GAD)
Low vagal tone, weak
inhibitory circuits
Thayer, Friedman, &
Borkovec (1996)
Nonrestorative sleep,
allodynia, fatigue
Fibromyalgia (FM),
chronic fatigue
syndrome (CFS)
Dopamine depletion in
hippocampal limbic circuits,
substance P in dorsal horn
Wood et al. (2007);
Russell (2000)
Labile blood pressure
readings
Labile or white-coat
hypertension
Weak baroreflex
Joseph et al. (2005);
Schein et al. (2001); Parati
& Steptoe (2004)
Asthma symptoms
Asthma
Pulmonary smooth
muscle hyperreactivity,
parasympathetic overreactivity,
airway inflammation
Lehrer et al. (2000a, 2004)
arrhythmia and the heart rate component in the
baroreflex system. Respiratory sinus arrhythmia
(RSA) refers to the rhythms in heart rate that
correspond with breathing, such that increases
in heart rate correspond with inhalation and
decreases with exhalation. RSA reflects processes
controlling respiratory and gas exchange control
(Hayano, Yasuma, Okada, Mukai, & Fujinami,
1996; Yasuma & Hayano, 2004), and affects many
other regulatory processes throughout the body
(Eckberg, 2003). The baroreflex modulates blood
pressure variability through changes in heart rate.
When blood pressure rises, heart rate falls; when
blood pressure falls, heart rate rises. The changes
in blood flow associated with heart rate produce
mechanical changes in blood pressure.
As we will see in the following sections early
research on HRV biofeedback showed that the
technique alters autonomic function, can restore
homeostatic autonomic balance, increases homeostatic regulation (ability to recover to normal
values after a perturbation), and improves various emotional and somatic symptoms affected
by the autonomic nervous system. Indirect effect
may influence inflammatory and emotional processes, through vagal-inflammatory pathways and
stimulation of brainstem structures involved in
emotional control, and directly affecting respiratory gas exchange efficiency by controlling phase
relationships between heart rate and breathing. If
these data hold up over time, this type HRV feedback will prove valuable in treating a variety of disorders (Gevirtz, 2006; Mayer, Naliboff, & Chang,
2001a; Mayer, Nalibott, Chang, & Coutinho,
2001b; van der Kolk, 2001, 2006; Frewen, Pain,
Dozois, & Lanius, 2006; Hopper, Frewen, van der
Kolk, & Lanius, 2007; Lanius et al., 2002; Lanius,
Bluhm, Lanius, & Pain, 2006; Thayer, Friedman,
& Borkovec, 1996; Thayer & Siegle, 2002; Thayer
& Friedman, 2002; Wood, 2006; Wood et al., 2007;
Russell, 2000; Parati & Steptoe, 2004; Schein et
al., 2001; Joseph et al., 2005; Lehrer et al., 2004;
Hassett et al., 2007; Karavidas et al., 2007; Zucker,
Samuelson, Muench, Greenberg, & Gevirtz, 2009;
Siepmann, Aykac, Unterdorfer, Petrowski, &
Mueck-Weymann, 2008; Siepmann et al., 2014;
Grossman, van Beck, & Wientjes, 1990).
202
Resonance Frequency Stimulation
as the Mechanism of HRV Biofeedback
The mechanism by which HRV biofeedback stimulates very high-amplitude oscillations in heart
rate is the interaction between two important control reflexes: respiratory sinus arrhythmia and the
baroreflex. The particular patterns by which these
reflexes are stimulated influence both respiratory
gas exchange (Hayano et al., 1996; Ito et al., 2006)
and the baroreflex (Lehrer et al., 2003; Vaschillo,
Lehrer, Rishe, & Konstantinov, 2002), which
modulates blood pressure (Fisher, Kim, Young, &
Fadel,2010; Gisolf, Imminck, van Lieshout, Stok,
& Karemaker, 2005; Liu et al., 2002). The baroreflex induces a specific rhythm in HRV, averaging about 5.5 times/minute (Vaschillo, Vaschillo,
& Lehrer, 2006). When blood pressure rises, the
baroreflex causes heart rate to slow. The subsequent reduction in heart rate then mechanically
causes a decrease in blood pressure, because less
blood flows through the vessels. When blood pressure decreases, the baroreflex then causes heart
rate to rise again. Because of plasticity in the
blood vessels and inertia in blood flow through
the system, the change in blood flow (and pressure) caused by heart rate changes is delayed by
about 4–5 seconds. This causes a low-frequency
rhythm in heart rate, at about 5.5 times/minute,
on average. Because respiration can be slowed
to approximately this rate, an interaction can
occur between respiratory sinus arrhythmia and
the baroreflex. In each person, there is a specific
oscillation frequency at which heart rate varies
perfectly in phase (0° phase relationship) with
breathing (Vaschillo, Vaschillo, & Lehrer, 2004),
and perfectly out of phase (180o phase relationship) with blood pressure (Vaschillo et al., 2002).
This frequency happens to coincide with the frequency in which the baroreflex affects heart rate.
Thus, when people breathe at this frequency, the
increases in heart rate accompanying inhalation
compound the increases in heart rate caused by the
baroreflex (induced by decreases in blood pressure
at that same time), causing simultaneous effects of
(1) increased baroreflex gain, (2) increased respiratory sinus arrhythmia, and 3) improved respiratory gas exchange (Yasuma & Havano, 2004). The
baroreflex is an important mechanism by which
the body modulates blood pressure changes.
Actually, the rhythms produced by the baroreflex create resonance characteristics in the cardiovascular system, such that stimulation (by breathing) at baroreflex frequency causes effects that
IV. RELAXATION INTERVENTIONS
are characteristic of resonance: an endogamous
rhythm in heart rate that is always present and
easily stimulated, and, when this rhythm is stimulated, very high-amplitude oscillations at a the
resonance frequency that recruit and obliterate
other sources of variation. Analogous examples of
resonance effects can be seen when a microphone
is placed in front of a speaker, or when a child is
pushed in a swing, in rhythm with every oscillation. An example of resonance effects in heart
rate during this process is shown in Figure 13.1,
in which the smaller oscillations represent resting
HRV, and the high-amplitude oscillations reflect
heart rate during resonance-frequency breathing.
Conversely, resonance frequency breathing can be
shaped and reinforced through HRV biofeedback.
When people try to produce maximal increases
in heart rate during inhalation and maximal
decreases during exhalation, they must be breathing at the resonance frequency produced by the
baroreflex system. Lehrer et al. (2003) have shown
that daily practice of this “resonance frequency
training” increases the total amount of HRV,
increases baroreflex gain, and improves pulmonary function, even among healthy people.
The overlapping of respiratory and baroreflex effects on HRV during slow respiratory rates
presents some treatment opportunities. Based on
experimental work by Vaschillo et al. (2002), Lehrer, Vaschillo, and Vaschillo (2000b) have theorized that practicing HRV biofeedback stimulates
resonance characteristics in the cardiorespiratory
system caused by the baroreflex system. They have
shown that daily practice of this “resonance frequency training” increases the total amount of
HRV, with almost all of the oscillations occurring
at a single frequency, thus increasing baroreflex
FIGURE 13.1. Example of resonance.
203
13. Cardiorespiratory Biofeedback
gain and, at the same time, improving pulmonary function and increasing cardiac vagal tone,
as reflected in RSA (Lehrer et al., 2003), even
among healthy people. Profound changes in these
measures are found both immediately, while practicing the biofeedback technique, and, over time,
in resting measures, when the individual is not
practicing biofeedback. Practice of the technique
improves both autonomic balance and autonomic
regulation. This has been found even when autonomic function (both sympathetic and parasympathetic) has been almost completely blocked by
experimental exposure to an inflammatory drug
(Lehrer et al., 2010). Repeated stimulation of the
baroreflex “exercises” is hypothesized to increase
the efficiency of the reflex (Lehrer et al., 2003),
thus leading to improved overall control of blood
pressure, autonomic reactivity, and through baroreflex-associated brainstem activity (centered at
the nucleus tractus solitarius) that impacts on
the limbic system’s control of emotional reactivity
(Gray et al., 2009).
HRV biofeedback may also affect central nervous
system functioning by way of the vagal afferents.
The vagal afferent system has recently emerged as
an area of interest due to the use of vagal nerve
stimulation for (1) severe depression (George et al.,
1994, 2000; Goodnick, Rush, George, Marangell,
& Sackheim, 2001; Marangell et al., 2002; Mu et
al., 2004; Rush et al., 2000; Sackeim et al., 2001),
(2) seizures (Hsiang, Wong, Kay, & Poon, 1998;
Sahin, Ilbay, Imal, Bozdogan, & Ates, 2009; Shahwan, Bailey, Maxiner, & Harvey, 2009; Sherman
et al., 2008), (3) congestive heart failure (Sabbah
et al., 2011; Zhang, Ilsar, Sabbah, Ben David, &
Magzgalev, 2009), and (4) autoimmune disorders
(Zitnik, 2011). Brown and Gerbarg (2005) have
speculated that slow breathing techniques (similar
to those used in HRV biofeedback) may be stimulating vagal afferents in a way that is similar to
the invasive medical devices. This would open up
many other applications for this technique.
Application Areas
Consistent with its important physiological effects,
many groups around the world have been reporting that “RSA” biofeedback is a viable feedback
modality. Wheat and Larkin (2010), reviewed the
literature that existed until about 2009 and concluded:
Results revealed that HRV biofeedback consistently
effectuates acute improvements during biofeedback
practice, whereas the presence of short-term and
long-term carry-over effects is less clear. Some evidence suggests HRV biofeedback may result in longterm carry-over effects on baroreflex gain, which
is an area most promising for future investigations.
(p. 229)
Practicing HRV biofeedback has produced dramatic improvements in asthma, including apparent 100% effectiveness in eliminating asthma
exacerbations, while improving pulmonary function, decreasing symptomatology, and allowing
decreases in consumption of asthma medication
(Lehrer et al., 2004). Herbs, Gevirtz, and Jacobs
(1994) have shown that HRV biofeedback reduces
blood pressure in hypertensives. Anecdotal reports
from Russia suggest that it may be helpful in treating a variety of psychosomatic and stress-related
physical disorders (Chernigovskaya, Vaschillo,
Petrash, & Rusanovsky, 1990). A similar technique
involving biofeedback training in slow breathing (without assessing resonance frequency, but
reducing breathing to the range where resonance
effects should be expected) has been deemed by
the U.S. Food and Drug Administration (FDA) to
be an effective adjunctive method for controlling
hypertension (Schein et al., 2001). This is the only
biofeedback method that has received an FDA
indication for treating any disease. Tables 13.2 and
13.3 list the various clinical conditions for which
there is evidence that HRV biofeedback has beneficial effects.
Procedure for Performing Resonance Frequency
Biofeedback Training
In the HRV biofeedback technique, the client is
instructed to maximize the peak–valley amplitude
based on a cardiotachometer line graph. Over
time, almost all participants achieve this by slowing and deepening breath, sometimes also by practicing a “mindful” mental state and using other
relaxation techniques. Although the method is
still too new for a codified procedure, a suggested
procedural method has been outlined by Lehrer
et al. (2000b, 2007). The first clinical session is
usually devoted to determining the trainee’s resonance frequency. This is accomplished by having
the individual breathe at various frequencies near
0.1 Hz, and finding the frequency that yields the
highest amplitude of heart rate oscillations. The
trainee is then advised to practice breathing at
this frequency daily until the next session (approximately a week later). In subsequent sessions, bio-
204
HRVB + oxymeter feedback
Chronic obstructive
pulmonary disease
HRVB
Recurrent abdominal pain
HRVB
HRVB
HRVB + CBT
HRVB + stress management
Congestive heart failure
Coronary artery disease
Coronary artery disease
Congestive heart failure
Cardiac rehabilitation
HRVB
HRVB
Cyclic vomiting
Fibromyalgia
HRVB integrated into other
therapies
HRVB
IBS
Recurrent abdominal pain
HRVB
Recurrent abdominal pain
Slow breathing + finger
temperature feedback
HRVB
Asthma
Functional gastrointestinal
disorders
Recurrent abdominal pain
Intervention
Disorder
Case studies
HRVB + CBT vs. stress
management
Vs. TAU
Vs. sham EEG
Vs. TAU
TAU
Case study
Case study
Vs. hypnosis
Vs. control
Vs. TAU
Vs. TAU
Vs. sham EEG
Design (control)
Harvested heart tissue
viability
HRV measures +
adjustment scales
HRV measures (SDNN)
6-minute walk
Standard fibromyalgia scales
IBS symptom measures
Vomiting frequency
Symptom log
IBS Symptom Severity
Scale, HADS
Symptom ratings and HRV
measures
Parent and child symptom
ratings
6-minute walk
Symptoms, lung function,
medication
Measures
Training group equal to LVAD
HRVB + CBT > stress
management
HRVB > TAU
HRVB > sham EEG if LVEF > 31
HRVB > TAU
HRVB > TAU
Greatly improved
Greatly improved
Both groups improved equally
(HRVB slightly better)
Symptom improvement
associated with SDNN gains
Breathing > control
HRVB > TAU
HRVB > control
Results
Moravec (2008); Moravec
& McKee (2013)
Nolan et al. (2005)
Del Pozo, Gevirtz, Scher,
& Guarneri (2004)
Swanson, Gevirtz, Spira,
& Guarneri (2006)
Hassett et al. (2007)
Ebert (2013)
Slutsker, Konichezky, &
Gothelf (2010)
Masters (2006)
Dobbin, Dobbin, Ross,
Graham, & Ford (2013)
Sowder, Gevirtz, Shapiro,
& Ebert (2010)
Humphreys & Gevirtz
(2000)
Giardino, Chan, &
Borson (2004)
Lehrer et al. (2000a,
2004)
Reference
TABLE 13.2. Disorders Treated with HRV Biofeedback That Are Hypothesized to Have Restoration of Autonomic Function as the Primary Mediator
205
Breathing and temperature
HRVB (stress eraser)
PIH
Postpartum depression
Vs. TAU (but random
assignment)
Vs. activity management
vs. TAU
Vs. matched case histories
Edinburgh Postnatal
Depression Scale
BP levels logged daily
BP, birthweight, gestation
length
Preterm stress, preterm
delivery
Measures of pain and
function
Measures of pain, vitality
and social functioning
Trigger point pain
Pain and function measures
BP, HRV
BP, HRV, BRS
Medication adjustment
and BP
HRVB had less anxiety, sleep
disturbance at 1 month than
controls
Biofeedback group halted rising
BP levels vs. other groups
HRVB > controls for birthweight
and gestation length
HRVB > control for stress; 13%
vs. 33% preterm delivery (n.s.)
HRVB > TAU
HRVB > TAU
HRVB combined with physical
release relieves pain
Combination superior to other
interventions
Slow breathing = EMG feedback
> slow breathing alone
HRVB > either control, improved
on BP, HRV, and BRS measures
HRVB maintained BP with fewer
medications
Kudo, Shinohara, &
Kodama (2014)
Sommers, Gevirtz, Jasin,
& Chin (1989)
Cullin et al. (2013)
Siepmann et al. (2014)
Berry et al. (2013)
Hallman, Olsson, von
Scheele, Melin, & Lyskov
(2003)
Gevirtz (2006)
Gordon & Gevirtz
(2006); Vagades et al.
(2011)
Wang et al. (2009)
Lin et al. (2012)
Reinke, Gevirtz, &
Mussgay (2007)
HRVB, heart rate variability biofeedback; CBT, cognitive-behavioral therapy; EMG, electromyogram; EEG, electroencephalogram; TAU, treatment as usual; IBS, irritable bowel syndrome; HADS,
Hamilton Anxiety and Depression Scale; SDNN, standard deviation of normal-to-normal heartbeats; BP, blood pressure; BRS, baroreflex sensitivity; LVEF, left ventricular ejection fraction; LVAD, left
ventricular assist device; 6-minute walk, maximum distance walked in 6 minutes; PIH, pregnancy-induced hypertension. From Gevirtz (2013). Copyright by the Association for Applied Psychophysiology
and Biofeedback. Reprinted by permission.
HRVB (stress eraser)
PIH
Vs. TAU
HRVB
Vs. control sessions
Vs. TAU
HRVB
HRVB
Case studies
HRVB
Preterm labor
OB/Gyn
Four groups: stabilization
exercises, HRVB alone,
myofascial release alone,
or combination
Vs. slow breathing alone
Vs. slow breathing and
control, 3-month follow-up
Vs. sham EEG
HRVB and myofascial
release
Slow abdominal breathing
+ EMG biofeedback
Prehypertensives
Chronic muscle pain
HRVB
HRVB
Pre-hypertensives
Hypertension
206
Depressed patients
vs. healthy controls
Vs. TAU after
cardiac surgery
Vs. relaxation
HRVB
HRVB
HRVB
HRVB
Vs. delayed treatment
Anxiety
HRVB
Anxiety and mood
Vs. matched
controls
Case example
Case study
HRVB + TAU
HRVB
Vs. relaxation
Vs. control
Vs. TAU
HRVB + DBT
Phobia
PTSD
HRVB
Anxiety disorders
PTSD
HRVB
Vs. Zoloft
Vs. DBT +
relaxation
HRVB with stress
eraser + DBT
HRVB + DBT +
Zoloft
No control, singlegroup trial
HRVB
Depression
Design (control)
Intervention
Disorder
HRVB > control
Somatic symptoms
Approach phobic object
PCL
PCL
Information processing
CAPS, Trauma Symptom
Checklist
BDI and HDS
BDI and HDS
CES-D
BDI
BDI and HDS
BDI and HDS
Measures
Henriques, Keffer, Abrahamson, &
Horst (2011)
HRVB using heartmath > control
Improved phobic avoidance
PCL improved 21%
HRVB = relaxation
HRVB > information processing
HRVB > TAU
HRVB > Zoloft alone
HRVB > relaxation
HRVB > TAU
Depressed patients reduced on BDI, no
changes in controls
HRVB group superior
Depression reduced markedly
Results
Nada (2009)
Prigatano (1973)
Reyes (2014)
Zucker, Samuelson, Muench,
Greenberg, & Gevirtz (2009)
Ginsberg, Berry, & Powell (2010)
Tan, Dao, Farmer, Sutherland, &
Gevirtz (2011)
Rene et al. (2011)
Rene, Gevirtz, Muench, & Birkhead
(2011)
Patron et al. (2013)
Siepmann, Aykac, Unterdorfer,
Petrowski, & Mueck-Weymann (2008)
Zucker, Samuelson, Muench,
Greenberg, & Gevirtz (2009)
Karavidas et al. (2007)
References
TABLE 13.3. Disorders Treated with HRV Biofeedback That Are Hypothesized to Involve Central Nervous System Mediators
207
HRVB
HRVB
HRVB
HRVB
HRVB
HRVB (Em wave)
Sleep
Performance
Baseball
Golf
Dance
Dance
Music
Vs. control
Vs. neurofeedback
vs. control
Vs. neurofeedback
vs. control
Case study
Vs. sports
psychology control
Vs. autogenics
(AT), vs. control
Vs. control
Performance anxiety
measures
Refereed dance ratings
Refereed dance ratings
Golf performance
Hitting performance
HF amplitude during sleep
Sleep disturbance scale +
actigraphy
Sleep log
Anxiety measures
Cholesterol, glucose, heart
rate, BP, positive outlook, and
overall psychological distress
HRVB > control
No effect on dance, HRVB reduced
anxiety
HRVB and neurofeedback > control
Reduced anxiety, improved performance
HRVB > controls
HRVB > AT > control
HRVB > controls
Insomnia improvement, long-term
maintenance
Improvement
HRVB > control on all measues,
projected cost savings
Thurber (2006)
Gruzelier, Thompson, Brandt, &
Steffert (in press)
Raymond, Sajid, Parkinson, & Gruzelier
(2005)
Lagos, Vaschillo, Vaschillo, Lehrer, &
Bates (2008)
Strack & Gevirtz (2011)
Sakakibara, Hayano, Oikawa,
Katsamanis, & Lehrer (2013)
Ebben, Kurbatov, & Pollak (2009)
McLay & Spira (2009)
Reiner (2008)
McCraty, Atkinson, Lipsenthal, &
Arguelles (2009)
DBT, dialectical behavior therapy; PTSD, posttraumatic stress disorder; AT, autogenic training; BDI, Beck Depression Inventory; HDS, Hamilton Depression Scale; CES-D, Center for Epidemiologic
Studies Depression Scale; CAPS, Clinician-Administered PTSD Scale; PCL, PTSD Checklist. Other abbreviations as in Table 13.2. From Gevirtz (2013). Copyright by the Association for Applied
Psychophysiology and Biofeedback. Reprinted by permission.
HRVB
Case report
Single-group study
HRVB + therapy
HRVB (stress eraser)
Vs. control
HRVB + stress
management
Sleep lab
insomnia
Sleep
Sleep
Stress
208
feedback is provided in the form of a cardiotachometer display or an online Fourier analysis of
heart rate, updated every few seconds. The trainee
is instructed to maximize the amplitude of heart
rate variability at his or her resonance frequency,
and to fine-tune the estimation of his or her resonance frequency by observing the respiration rate
that continues to yield the highest amplitude of
heart rate variability. To enhance the effect, the
individual also is sometimes instructed to breathe
abdominally and to exhale through pursed lips.
The trainee is also advised to breathe shallowly
in order to avoid hyperventilating during practice
of this method. Although most people are able to
produce high amplitudes of HRV within just a few
minutes of training, it usually takes several sessions to learn to stabilize HRV at the resonance
frequency. The number of sessions required to
maximize effects is not yet known. Data from our
laboratories suggest the curve representing physiological effects has not yet leveled off after 10 sessions of training.
Conclusion
HRV biofeedback has become a popular modality in the professional community and has been
widely disseminated to the public. Thanks to
products such as the emWave (heartmath.com), the
Stress Eraser (stresseraser.com), Journey to the Wild
Divine (wilddivine.com), MyCalmBeat (mybrainsolutions.com), Resp-e-Rate (resperate.com), and
others, a worldwide audience is now using versions of this biofeedback technique. It remains to
be determined how the actual biofeedback can be
combined with other therapies or meditative disciplines. Additionally, the actual mechanisms of
change need to be identified (Giardino, Lehrer, &
Feldman, 2000). Lehrer and Gevirtz (2014) have
recently reviewed possible mechanisms.
Glossary
Oscillators. Devices or physiological systems that pro-
duces to and fro rhythmic activity.
R-wave. Electrocardiographic wave represented by the
peak during electrical stimulation of the ventricles.
It is used in biofeedback settings to trigger a timer
to measure “interbeat interval” (the time between
R-wave peaks).
Spectral analysis. A mathematical and graphical tech-
nique used to decompose complex wave forms into
their constituent components (frequency bins). The
IV. RELAXATION INTERVENTIONS
mathematical formulae to do this were first worked
out by Fourier; therefore, the analysis is often called a
fast Fourier transform.
References
Bass, C. (1994). Management of patients with hyperventilation-related disorders. In B. H. Timmons & R. Ley
(Eds.), Behavioral and psychological approaches to breathing disorders. New York: Plenum Press.
Bass, C., Gardner, W. N., & Jackson, G. (1994). Psychiatric
and respiratory aspects of functional cardiovascular syndromes. In B. H. Timmons & R. Ley (Eds.), Behavioral
and psychological approaches to breathing disorders (pp.
125–138). New York: Springer.
Bernardi, L., Leuzzi, S., Radaelli, A., Passino, C., Johnston,
J. A., & Sleight, P. (1994). Low-frequency spontaneous
fluctuations of R-R interval and blood pressure in conscious humans: A baroreceptor or central phenomenon?
Clinical Science, 87(6), 649–654.
Berntson, G. G., Bigger, J. T., Jr., Eckberg, D. L., Grossman,
P., Kaufmann, P. G., Malik, M., et al. (1997). Heart rate
variability: Origins, methods, and interpretative caveats. Psychophysiology, 34, 623–648.
Berntson, G. G., Cacioppo, J. T., & Quigley, K. S. (1993).
Cardiac psychophysiology and autonomic space in
humans: Empirical perspectives and conceptual implications. Psychological Bulletin, 114(2), 296–322.
Berntson, G. G., Cacioppo, J. T., Quigley, K. S., & Fabro,
V. T. (1994). Autonomic space and psychophysiological
response. Psychophysiology, 31(1), 44–61.
Berry, M. E., Chapple, I. T., Ginsberg, J. P., Gleichauf, K. J.,
Meyer, J. A., & Nagpal, M. L. (2013) Non-pharmacological intervention for chronic pain in veterans: A pilot
study of heart rate variability biofeedback. Global Advice
Health Medicine, 3(2), 28–33.
Biber, B., & Alkin, T. (1999). Panic disorder subtypes: Differential responses to CO2 challenge. American Journal
of Psychiatry, 156(5), 739–744.
Blumenthal, J. A., Jiang, W., Babyak, M. A., Krantz, D. S.,
Frid, D. J., Coleman, R. E., et al. (1997). Stress management and exercise training in cardiac patients with myocardial ischemia: Effects on prognosis and evaluation
of mechanisms. Archives of Internal Medicine, 157(19),
2213–2223.
Bradley, L. A., Richter, J. E., Scarinci, I. C., Haile, J. M.,
& Schan, C. A. (1992). Psychosocial and psychophysical assessments of patients with unexplained chest pain.
American Journal of Medicine, 92(5A), 65S–73S.
Brown, R. P., & Gerbarg, P. L. (2005). Sudarshan Kriya
yogic breathing in the treatment of stress, anxiety, and
depression: Part II. Clinical applications and guidelines.
Journal of Alternative and Complementary Medicine,
11(4), 711–717.
Chernigovskaya, N. V., Vaschillo, E. G., Petrash, V. V., &
Rusanovsky, V. V. (1990). Voluntary regulation of the
heart rate as a method of functional condition correction in neurotics. Human Physiology, 16, 58–64.
13. Cardiorespiratory Biofeedback
Clouse, R. E. (1992). Psychopharmacologic approaches to
therapy for chest pain of presumed esophageal origin.
American Journal of Medicine, 92(Suppl 5A), 106–113.
Cullins, S. W., Gevirtz, R. N., Poeltler, D. M., Cousins,
L. M., Edward Harpin, R., & Muench, F. (2013). An
exploratory analysis of the utility of adding cardiorespiratory biofeedback in the standard care of pregnancyinduced hypertension. Applied Psychophysiology and Biofeedback, 38(3), 161–170.
DeGuire, S., Gevirtz, R., Hawkinson, D., & Dixon, K.
(1996). Breathing retraining: A three-year follow-up
study of treatment for hyperventilation syndrome and
associated functional cardiac symptoms. Biofeedback and
Self-Regulation, 21(2), 191–198.
DeGuire, S., Gevirtz, R., Kawahara, Y., & Maguire, W.
(1992). Hyperventilation syndrome and the assessment
of treatment for functional cardiac symptoms. American
Journal of Cardiology, 70(6), 673–677.
Del Pozo, J. M., Gevirtz, R. N., Scher, B., & Guarneri, E.
(2004). Biofeedback treatment increases heart rate variability in patients with known coronary artery disease.
American Heart Journal, 147(3), E11.
Dobbin, A., Dobbin, J., Ross, S. C., Graham, C., & Ford,
M. J. (2013). Randomised controlled trial of brief intervention with biofeedback and hypnotherapy in patients
with refractory irritable bowel syndrome. Journal of
the Royal Colleege of Physicians of Edinburgh, 43(1),
15–23.
Duivenvoorden, H. J., & van dixhoorn, J. (1991). Predictability of psychic outcome for exercise training and
exercise training including relaation therapy after myocardial infarction. Journal of Psychosomatic Research, 35,
569–578.
Ebben, M. R., Kurbatov, V., & Pollak, C. P. (2009). Moderating laboratory adaptation with the use of a heart-rate
variability biofeedback device (StressEraser). Applied
Psychophysiology and Biofeedback, 34(4), 245–249.
Ebert, C. (2013). The use of HRVB for the treatment of FGID
in children and adolescents. Paper presented at the AAPB,
Portland, OR.
Eckberg, D. L. (2003). The human respiratory gate. Journal
of Physiology, 548(Pt. 2), 339–352.
Fisher, J. P., Kim, A., Young, C. N., & Fadel, P. J. (2010).
Carotid baroreflex control of arterial blood pressure
at rest and during dynamic exercise in aging humans.
American Journal of Physiology: Regulatory, Integrative
and Comparative Physiology, 299(5), R1241–R1247.
Frewen, P. A., Pain, C., Dozois, D. J., & Lanius, R. A.
(2006). Alexithymia in PTSD: Psychometric and fMRI
studies. Annals of the New York Academy of Sciences,
1071, 397–400.
Fried, R., & Grimaldi, J. (1993). The psychology and physiology of breathing. New York: Springer.
Garssen, B., De Ruiter, C., & van Dyck, R. (1992). Breathing retraining: A rationale placebo. Clinical Psychology
Review, 12, 141–153.
George, M. S., Sackeim, H. A., Marangell, L. B., Husain,
M. M., Nahas, Z., Lisanby, S. H., et al. (2000). Vagus
nerve stimulation: A potential therapy for resistant
209
depression? Psychiatric Clinics of North America, 23(4),
757–783.
George, R., Salinsky, M., Kuzniecky, R., Rosenfeld, W., Bergen, D., Tarver, W. B., et al. (1994). Vagus nerve stimulation for treatment of partial seizures: 3. Long-term
follow-up on first 67 patients exiting a controlled study:
First International Vagus Nerve Stimulation Study
Group. Epilepsia, 35(3), 637–643.
Gevirtz, R. (2006). The muscle spindle trigger point model
of chronic pain. Biofeedback, 34(2), 53–57.
Gevirtz, R. N. (2006). The muscle spindle trigger point
model of chronic pain. Biofeedback, 34(2), 53–57.
Gevirtz, R. N. (2007). Psychophysiological perspectives on
stress related and anxiety disorders. In P. Lehrer & W.
Sime (Eds.), Stress management: Principles and practice
(2nd ed., pp. 209–226). New York: Guilford Press.
Giardino, N. D., Chan, L., & Borson, S. (2004). Combined
heart rate variability and pulse oximetry biofeedback
for chronic obstructive pulmonary disease: Preliminary findings. Applied Psychophysiology and Biofeedback,
29(2), 121–133.
Giardino, N. D., Lehrer, P., & Feldman, J. M. (2000). The
role of oscillations in self-regulation: A revision of the
classical model of homeostasis. In J. G. Carlson, D.
Kenny, F. J. McGuigan, & J. L. Sheppard (Eds.), Stress
and health: Research and clinical applications (pp. 27–52).
Amsterdam, The Netherlands: Harwood.
Ginsberg, J. P., Berry, M. E., & Powell, D. A. (2010). Cardiac coherence and posttraumatic stress disorder in
combat veterans. Alternative Therapies in Health and
Medicine, 16(4), 52–60.
Gisolf, J., Immink, R. V., van Lieshout, J. J., Stok, W. J.,
& Karemaker, J. M. (2005). Orthostatic blood pressure
control before and after spaceflight, determined by timedomain baroreflex method. Journal of Applied Physiology,
98(5), 1682–1690.
Goodnick, P. J., Rush, A. J., George, M. S., Marangell, L.
B., & Sackeim, H. A. (2001). Vagus nerve stimulation
in depression. Expert Opinion in Pharmacotherapy, 2(7),
1061–1063.
Gordon, C., & Gevirtz, R. (2006). Trigger point pain
theory and practice: Recent developments [Abstract].
Applied Psychophysiology and Biofeedback, 29(4), 299.
Gray, M. A., Rylander, K., Harrison, N. A., Wallin, B., Gunnar, B., & Critchley, H. D. (2009). Following one’s heart:
Cardiac rhythms gate central initiation of sympathetic
reflexes. Journal of Neuroscience, 29(6), 1817–1825.
Grossman, P., de Swart, J. C., & Defares, P. B. (1985). A
controlled study of a breathing therapy for treatment
of hyperventilation syndrome. Journal of Psychosomatic
Research, 29(1), 49–58.
Grossman, P., & Taylor, E. W. (2007). Toward understanding respiratory sinus arrhythmia: Relations to cardiac
vagal tone, evolution and biobehavioral functions. Biological Psychology, 74(2), 263–285.
Grossman, P., van Beek, J., & Wientjes, C. (1990). A comparison of three quantification methods for estimation
of respiratory sinus arrhythmia. Psychophysiology, 27(6),
702–714.
210
Gruzelier, J., Thompson, T., Brandt, R., & Steffert, T. (in
press). Application of alpha/theta neurofeedback and
heart rate variabilty training to contemporary dancers:
State anxiety and creativity. International Journal of Psychophysiology.
Hallman, D. M., Olsson, E. M., von Scheele, B., Melin,
L., & Lyskov, E. (2003). Effects of heart rate variability
biofeedback in subjects with stress-related chronic neck
pain: A pilot study. Applied Psychophysiology and Biofeedback, 36(2), 71–80.
Hassett, A. L., Radvanski, D. C., Vaschillo, E. G., Vaschillo, B., Sigal, L. H., Karavidas, M. K., et al. (2007). A
pilot study of the efficacy of heart rate variability (HRV)
biofeedback in patients with fibromyalgia. Applied Psychophysiology and Biofeedback, 32(1), 1–10.
Hayano, J., Yasuma, F., Okada, A., Mukai, S., & Fujinami,
T. (1996). Respiratory sinus arrhythmia: A phenomenon
improving pulmonary gas exchange and circulatory efficiency. Circulation, 94(4), 842–847.
Hegel, M. T., Abel, G. G., Etscheidt, M., Cohen-Cole, S.,
& Wilmer, C. I. (1989). Behavioral treatment of anginalike chest pain in patients with hyperventilation syndrome. Journal of Behavior Therapy and Experimental
Psychiatry, 20(1), 31–39.
Henriques, G., Keffer, S., Abrahamson, C., & Horst, S. J.
(2011). Exploring the effectiveness of a computer-based
heart rate variability biofeedback program in reducing
anxiety in college students. Applied Psychophysiology and
Biofeedback, 36(2), 101–112.
Herbs, D., Gevirtz, R. N., & Jacobs, D. (1994). The effect
of heart rate pattern biofeedback for the treatment of
essential hypertension [Abstract]. Biofeedback and SelfRegulation, 19(3), 281.
Hofmann, S. G., & Smits, J. A. (2008). Cognitive-behavioral therapy for adult anxiety disorders: A meta-analysis of randomized placebo-controlled trials. Journal of
Clinical Psychiatry, 69(4), 621–632.
Hopper, J. W., Frewen, P. A., van der Kolk, B. A., & Lanius,
R. A. (2007). Neural correlates of reexperiencing, avoidance, and dissociation in PTSD: Symptom dimensions
and emotion dysregulation in responses to script-driven
trauma imagery. Journal of Traumatic Stress, 20(5), 713–
725.
Hsiang, J. N., Wong, L. K., Kay, R., & Poon, W. S. (1998).
Vagus nerve stimulation for seizure control: Local
­experience. Journal of Clinical Neuroscience, 5(3), 294–
297.
Humphreys, P. A., & Gevirtz, R. N. (2000). Treatment of
recurrent abdominal pain: Components analysis of four
treatment protocols. Journal of Pediatric Gastroenterology
and Nutrition, 31(1), 47–51.
Ito, S., Sasano, H., Sasano, N., Hayano, J., Fisher, J. A., &
Katsuya, H. (2006). Vagal nerve activity contributes to
improve the efficiency of pulmonary gas exchange in
hypoxic humans. Experimental Physiology, 91(5), 935–
941.
Joseph, C. N., Porta, C., Casucci, G., Casiraghi, N., Maffeis,
M., Rossi, M., et al. (2005). Slow breathing improves
arterial baroreflex sensitivity and decreases blood pres-
IV. RELAXATION INTERVENTIONS
sure in essential hypertension. Hypertension, 46(4),
714–718.
Karavidas, M. K., Lehrer, P. M., Vaschillo, E., Vaschillo, B.,
Marin, H., Buyske, S., et al. (2007). Preliminary results
of an open label study of heart rate variability biofeedback for the treatment of major depression. Applied Psychophysiology and Biofeedback, 32(1), 19–30.
Kudo, N., Shinohara, H., & Kodama, H. (2014). Heart rate
variability biofeedback intervention for reduction of
psychological stress during the early postpartum period.
Applied Psychophysiology and Biofeedback, 39(3–4), 203–
211.
Lagos, L., Vaschillo, B., Vaschillo, E., Lehrer, P., & Bates,
M. (2008). Heart rate variability biofeedback as a strategy for dealing with competitive anxiety: A case study.
Biofeedback.36(3), 109–115.
Lanius, R. A., Bluhm, R., Lanius, U., & Pain, C. (2006). A
review of neuroimaging studies in PTSD: Heterogeneity
of response to symptom provocation. Journal of Psychiatric Research, 40(8), 709–729.
Lanius, R. A., Williamson, P. C., Boksman, K., Densmore,
M., Gupta, M., Neufeld, R. W., et al. (2002). Brain activation during script-driven imagery induced dissociative
responses in PTSD: A functional magnetic resonance
imaging investigation. Biological Psychiatry, 52(4), 305–
311.
Lehrer, P. M. (2007). Biofeedback training to increase heart
rate variability. In P. M. Lehrer, R. L. Woolfolk, & W. E.
Sime (Eds.), Principles and practices of stress management
(pp. 227–248). New York: Guilford Press.
Lehrer, P. M., & Gevirtz, R. N. (2014, July). Heart rate variability biofeedback: How and why does it work? Frontiers
of Psychology.
Lehrer, P. M., Karavidas, M. K., Lu, S. E., Coyle, S. M.,
Oikawa, L. O., Macor, M., et al. (2010). Voluntarily produced increases in heart rate variability modulate autonomic effects of endotoxin induced systemic inflammation: An exploratory study. Applied Psychophysiology and
Biofeedback, 35(4), 303–315.
Lehrer, P. M., Smetankin, A., & Potapova, T. (2000a).
Respiratory sinus arrhythmia biofeedback therapy for
asthma: A report of 20 unmedicated pediatric cases
using the Smetankin method. Applied Psychophysiology
and Biofeedback, 25, 193–200.
Lehrer, P. M., Vaschillo, E., & Vaschillo, B. (2000b). Resonant frequency biofeedback training to increase cardiac
variability: Rationale and manual for training. Applied
Psychophysiology and Biofeedback, 25(3), 177–191.
Lehrer, P. M., Vaschillo, E., Vaschillo, B., Edelberg, R., Lu,
S. E., Eckberg, D., et al. (2003). Heart rate variability
biofeedback increases baroreflex gain and peak expiratory flow. Psychosomatic Medicine, 65(5), 796–805.
Lehrer, P. M., Vaschillo, E., Vaschillo, B., Lu, S. E.,
Scardella, A., Siddique, M., et al. (2004). Biofeedback
treatment for asthma. Chest, 126(2), 352–361.
Ley, R. (1991). The efficacy of breathing retraining and the
centrality of hyperventilation in panic disorder: A reinterpretation of experimental findings [comment] [see
comments]. Behavior Research Therapy, 29(3), 301–304.
13. Cardiorespiratory Biofeedback
Ley, R. (1992). The many faces of Pan: Psychological and
physiological differences among three types of panic
attacks. Behavior Research Therapy, 30(4), 347–357.
Ley, R. (1993). Breathing retraining in the treatment of
hyperventilatory complaints and panic disorder: A reply
to Garssen, de Ruiter, & van Dyck. Clinical Psychology
Review, 13(5), 393–408.
Lin, G., Xiang, Q., Fu, X., Wang, S., Chen, S., Shao, L., et
al. (2012). Heart rate variability biofeedback decreases
blood pressure in prehypertensive subjects by improving
autonomic function and baroreflex. Journal of Alternative and Complementary Medicine, 18(2), 143–152.
Liu, H.-K., Guild, S.-J., Ringwood, J. V., Barrett, C. J., Leonard, B. L., Nguang, S.-K., et al. (2002). Dynamic baroreflex control of blood pressure: Influence of the heart
vs. peripheral resistance. American Journal of Physiology: Regulatory, Integrative and Comparative Physiology,
283(2), R533–R542.
Marangell, L. B., Rush, A. J., George, M. S., Sackeim, H.
A., Johnson, C. R., Husain, M. M., et al. (2002). Vagus
nerve stimulation (VNS) for major depressive episodes:
One year outcomes. Biological Psychiatry, 51(4), 280–287.
Masters, K. S. (2006). Recurrent abdominal pain, medical
intervention, and biofeedback: What happened to the
biopsychosocial model? Applied Psychophysiology and
Biofeedback, 31(2), 155–165.
Mayer, E. A., Naliboff, B. D., & Chang, L. (2001a). Basic
pathophysiologic mechanisms in irritable bowel syndrome. Digestive Diseases, 19(3), 212–218.
Mayer, E. A., Naliboff, B. D., Chang, L., & Coutinho, S. V.
(2001b). V. Stress and irritable bowel syndrome. American Journal of Physiology. Gastrointestinal and Liver Physiology, 280(4), G519–G524.
McCraty, R., Atkinson, M., Lipsenthal, L., & Arguelles, L.
(2009). New hope for correctional officers: An innovative program for reducing stress and health risks. Applied
Psychophysiology and Biofeedback, 34(4), 251–272.
McLay, R. N., & Spira, J. L. (2009). Use of a portable biofeedback device to improve insomnia in a combat zone:
A case report. Applied Psychophysiology and Biofeedback,
34(4), 319–321.
Meuret, A. E., Hofmann, S. G., & Rosenfield, D. (2010).
Catastrophic appraisal and perceived control as moderators of treatment response in panic disorder. International Journal of Cognitive Therapy, 3, 262–277.
Meuret, A. E., & Ritz, T. (2010). Hyperventilation in panic
disorder and asthma: Empirical evidence and clinical strategies. International Journal of Psychophysiology,
78(1), 68–79.
Meuret, A. E., Ritz, T., Wilhelm, F. H., & Roth, W. T.
(2005). Voluntary hyperventilation in the treatment
of panic disorder—functions of hyperventilation, their
implications for breathing training, and recommendations for standardization. Clinical Psychology Review,
25(3), 285–306.
Meuret, A. E., Rosenfield, D., Hofmann, S. G., Suvak, M.
K., & Roth, W. T. (2009). Changes in respiration mediate changes in fear of bodily sensations in panic disorder. Journal of Psychiatric Research, 43(6), 634–641.
211
Meuret, A. E., Rosenfield, D., Seidel, A., Bhaskara, L., &
Hofmann, S. G. (2010). Respiratory and cognitive mediators of treatment change in panic disorder: Evidence
for intervention specificity. Journal of Consulting and
Clinical Psychology, 78(5), 691–704.
Meuret, A. E., Wilhelm, F. H., Ritz, T., & Roth, W. T.
(2003). Breathing training for treating panic disorder:
Useful intervention or impediment? Behavior Modification, 27(5), 731–754.
Meuret, A. E., Wilhelm, F. H., Ritz, T., & Roth, W. T.
(2008). Feedback of end-tidal pCO2 as a therapeutic approach for panic disorder. Journal of Psychiatric
Research, 42(7), 560–568.
Moravec, C. S. (2008). Biofeedback therapy in cardiovascular disease: Rationale and research overview. Cleveland Clinical Journal of Medicine, 75(Suppl. 2), S35–S38.
Moravec, C. S., & McKee, M. G. (2008). Biofeedback in
the treatment of heart disease. Cleveland Clinic Journal
of Medicine, 78(Suppl. 1), S20–S23.
Moravec, C. S., & McKee, M. G. (2013). Psychophysiologic
remodeling of the failing heart. Biofeedback, 41(1), 7–12.
Moynihan, J. E., & Gevirtz, R. N. (2001). Respiratory and
cognitive subtypes of panic. Preliminary validation of
Ley's model. Behavior Modification, 25(4), 555–583.
Mu, Q., Bohning, D. E., Nahas, Z., Walker, J., Anderson, B.,
Johnson, K. A., et al. (2004). Acute vagus nerve stimulation using different pulse widths produces varying brain
effects. Biological Psychiatry, 55(8), 816–825.
Nada, P. J. (2009). Heart rate variability in the assessment
and biofeedback training of common mental health
problems in children. Medicinski Archives, 63(5), 244–
248.
Nixon, P. G. (1989). Human functions and the heart. In D.
S. A. Cribb (Ed.), Changing ideas in health care. Chichester, UK: Wiley.
Nolan, R. P., Kamath, M. V., Floras, J. S., Stanley, J., Pang,
C., Picton, P., et al. (2005). Heart rate variability biofeedback as a behavioral neurocardiac intervention to
enhance vagal heart rate control. American Heart Journal, 149(6), 1137.
Oakes, D. F. (1996). Clinical practitioner's guide to respiratory
care (4th ed.). Old Town, MN: Health Educator.
Parati, G., & Steptoe, A. (2004). Stress reduction and
blood pressure control in hypertension: A role for transcendental meditation? Journal of Hypertension, 22(11),
2057–2060.
Patel, C., Marmot, M. G., Terry, D. J., Carruthers, M., Hunt,
B., & Patel, M. (1985). Trial of relaxation in reducing
coronary risk: Four year follow up. British Medical Journal
(Clin Res Ed), 290(6475), 1103–1106.
Patron, E., Messerotti Benvenuti, S., Favretto, G., Valfre,
C., Bonfa, C., Gasparotto, R., et al. (2013). Biofeedback assisted control of respiratory sinus arrhythmia as
a biobehavioral intervention for depressive symptoms
in patients after cardiac surgery: A preliminary study.
Applied Psychophysiology and Biofeedback, 38(1), 1–9.
Porges, S. (2011). The polyvagal theory. New York: Norton.
Prigatano, G. P. (1973). Spider phobia: Autonomic reactions
and biofeedback control of heart rate variability. Unpub-
212
lished doctoral dissertation, Bowling Green State University, Bowling Green, KY.
Raymond, J., Sajid, I., Parkinson, L. A., & Gruzelier, J. H.
(2005). Biofeedback and dance performance: A preliminary investigation. Applied Psychophysiology and Biofeedback, 30(1), 64–73.
Reiner, R. (2008). Integrating a portable biofeedback
device into clinical practice for patients with anxiety
disorders: Results of a pilot study. Applied Psychophysiology and Biofeedback, 33(1), 55–61.
Reinke, A., Gevirtz, R., & Mussgay, L. (2007). Effects of
heart rate variability feedback in reducing blood pressure. Applied Psychophysiology and Biofeedback, 32, 134.
Rene, R., Gevirtz, R., Muench, F., & Birkhead, E. (2011).
The efficacy of a portable heart rate variability (HRV)
biofeedback device in conjunction with mental health treatment of clients with major depressive disorder enrolled in
a county welfare-to-work program. Unpublished dissertation, Alliant University, San Diego, CA.
Reyes, F. J. (2014). Implementing heart rate variability biofeedback groups for veterans with PTSD. Biofeedback,
42(4), 137–142.
Roth, W. T., Wilhelm, F. H., & Trabert, W. (1998). Autonomic instability during relaxation in panic disorder.
Psychiatry Research, 80(2), 155–164.
Rush, A. J., George, M. S., Sackeim, H. A., Marangell, L.
B., Husain, M. M., Giller, C., et al. (2000). Vagus nerve
stimulation (VNS) for treatment-resistant depressions:
A multicenter study. Biological Psychiatry, 47(4), 276–
286.
Russell, I. J. (2000). Fibromyalgia syndrome. In S. Mense,
D. Simons, & I. J. Russell (Eds.), Muscle pain: Understanding its nature, diagnosis and treatment (pp. 289–337).
New York: Lippincott, Williams & Wilkins.
Sabbah, H. N., Ilsar, I., Zaretsky, A., Rastogi, S., Wang,
M., & Gupta, R. C. (2011). Vagus nerve stimulation in
experimental heart failure. Heart Failure Reviews, 16(2),
171–178.
Sackeim, H. A., Rush, A. J., George, M. S., Marangell, L.
B., Husain, M. M., Nahas, Z., et al. (2001). Vagus nerve
stimulation (VNS) for treatment-resistant depression:
Efficacy, side effects, and predictors of outcome. Neuropsychopharmacology, 25(5), 713–728.
Sahin, D., Ilbay, G., Imal, M., Bozdogan, O., & Ates, N.
(2009). Vagus nerve stimulation suppresses generalized
seizure activity and seizure-triggered postictal cardiac
rhythm changes in rats. Physiology Research, 58(3),
345–350.
Sakakibara, M., Hayano, J., Oikawa, L. O., Katsamanis,
M., & Lehrer, P. (2013). Heart rate variability biofeedback improves cardiorespiratory resting function during
sleep. Applied Psychophysiology and Biofeedback, 38(4),
265–271.
Salkovskis, P. M. (1992). Psychological treatment of noncardiac chest pain: The cognitive approach. American
Journal of Medicine, 92(5A), 114S–121S.
Salkovskis, P. M., & Clark, D. M. (1990). Affective
responses to hyperventilation: A test of the cognitive
model of panic. Behavior Research Therapy, 28(1), 51–61.
IV. RELAXATION INTERVENTIONS
Schein, M. H., Gavish, B., Herz, M., Rosner-Kahana, D.,
Naveh, P., Knishkowy, B., et al. (2001). Treating hypertension with a device that slows and regularises breathing: A randomised, double-blind controlled study. Journal of Human Hypertension, 15(4), 271–278.
Shahwan, A., Bailey, C., Maxiner, W., & Harvey, A. S.
(2009). Vagus nerve stimulation for refractory epilepsy
in children: More to VNS than seizure frequency reduction. Epilepsia, 50(5), 1220–1228.
Sherman, E. M., Connolly, M. B., Slick, D. J., Eyrl, K. L.,
Steinbok, P., & Farrell, K. (2008). Quality of life and
seizure outcome after vagus nerve stimulation in children with intractable epilepsy. Journal of Child Neurology, 23(9), 991–998.
Siepmann, M., Aykac, V., Unterdorfer, J., Petrowski, K.,
& Mueck-Weymann, M. (2008). A pilot study on the
effects of heart rate variability biofeedback in patients
with depression and in healthy subjects. Applied Psychophysiology and Biofeedback, 33(4), 195–201.
Siepmann, M., Hennig, U., Siepmann, T., Nitzsche, K.,
Muck-Weymann, M., Petrowski, K., et al. (2014). The
effects of heart rate variability biofeedback in patients
with preterm labour. Applied Psychophysiology and Biofeedback, 39(1), 27–35.
Slutsker, B., Konichezky, A., & Gothelf, D. (2010). Breaking the cycle: Cognitive behavioral therapy and biofeedback training in a case of cyclic vomiting syndrome.
Psychology, Health, and Medicine, 15(6), 625–631.
Sommers, P., Gevirtz, R., Jasin, S., & Chin, H. (1989). The
efficacy of biobehavioral and compliance enhancement
interventions in the adjunctive treatment of mild pregnancy induced hypertension. Biofeedback and Self Regulation, 14(4), 309–317.
Sowder, E., Gevirtz, R., Shapiro, W., & Ebert, C. (2010).
Restoration of vagal tone: A possible mechanism for
functional abdominal pain. Applied Psychophysiology and
Biofeedback, 35(3), 199–206.
Strack, B., & Gevirtz, R. (2011). Getting to the heart of the
matter: Heart rate variability biofeedback for enhanced
performance In B. Strack, M. Linden & V. Wilson
(Eds.), Biofeedback and neurofeedback applications in sport
psychology (pp. 145–174). Wheat Ridge, CO: AAPB.
Swanson, K., Gevirtz, R., Spira, J., & Guarneri, E. (2006).
Does biofeedback increase heart rate variability in patents with known New York Heart Association class I-III
heart failure [Abstract]. Applied Psychophysiology and
Biofeedback, 2, 71–91.
Tan, G., Dao, T. K., Farmer, L., Sutherland, R. J., & Gevirtz,
R. (2011). Heart rate variability (HRV) and posttraumatic stress disorder (PTSD): A pilot study. Applied Psychophysiology and Biofeedback, 36(1), 27–35.
Thayer, J. F., & Friedman, B. H. (2002). Stop that!: Inhibition, sensitization, and their neurovisceral concomitants. Scandinavian Journal of Psychology, 43(2), 123–
130.
Thayer, J. F., Friedman, B. H., & Borkovec, T. D. (1996).
Autonomic characteristics of generalized anxiety disorder and worry. Biological Psychiatry, 39(4), 255–266.
Thayer, J. F., & Siegle, G. J. (2002). Neurovisceral integra-
13. Cardiorespiratory Biofeedback
tion in cardiac and emotional regulation. IEEE Engineering in Medicine and Biology Magazine, 21(4), 24–29.
Thurber, M. R. (2006). Effects of heart-rate variability biofeedback training and emotional regulation on music performance anxiety in university students. Denton: University
of North Texas Press.
Timmons, B. H., & Ley, R. (Eds.). (1994). Behavioral and
psychological approaches to breathing disorders. New York:
Plenum Press.
Vagades, J., Gordon, C., Schwaemmle, M., Andrasik, F.,
Gevirtz, R. N., Hautzinger, M., et al. (2011). Does deep
breathing training improve myofascial release in comination
with trigger point therapy for patients with low back pain? .
Paper presented at the 15th annual meeting of the Biofeedback Foundation of Europe.
van der Kolk, B. A. (2001). The psychobiology and psychopharmacology of PTSD. Human Psychopharmacology,
16(Suppl. 1), S49–S64.
van der Kolk, B. A. (2006). Clinical implications of neuroscience research in PTSD. Annals of the New York Academy of Sciences, 1071, 277–293.
van Dixhoorn, J. (2007). Whole body breathing. In P. M.
Lehrer, R. L. Woolfolk, & W. E. Sime (Eds.), Principles
and practices of stress management (pp. 291–332). New
York: Guilford Press.
van Dixhoorn, J. J., & Duivenvoorden, H. J. (1999). Effect
of relaxation therapy on cardiac events after myocardial
infarction: A 5-year follow-up study. Journal of Cardiopulmonary Rehabilitation, 19(3), 178–185.
van Dixhoorn, J., Duivenvoorden, H. J., Staal, H. A., &
Pool, J. (1989). Physical training and relaxation therapy
in cardiac rehabilitation assessed through a composite
criterion for training outcome. American Heart Journal,
118(3), 545–552.
van Dixhoorn, J., Duivenvoorden, H. J., Staal, J. A., Pool,
J., & Verhage, F. (1987). Cardiac events after myocardial
infarction: Possible effect of relaxation therapy. European Heart Journal, 8(11), 1210–1214.
Vaschillo, E. G. (1984). Dynamics of slow-wave cardiac
rhythm structure as an index of the functional state of an
operant. Doctoral dissertation, Leningrad State University, Leningrad, Russia.
Vaschillo, E., Lehrer, P., Rishe, N., & Konstantinov, M.
213
(2002). Heart rate variability biofeedback as a method
for assessing baroreflex function: A preliminary study of
resonance in the cardiovascular system. Applied Psychophysiology and Biofeedback, 27(1), 1–27.
Vaschillo, E., Vaschillo, B., & Lehrer, P. (2004). Heartbeat
synchronizes with respiratory rhythm only under specific circumstances. Chest, 126(4), 1385–1386; author
reply 1386–1387.
Vaschillo, E. G., Vaschillo, B., & Lehrer, P. M. (2006).
Characteristics of resonance in heart rate variability
stimulated by biofeedback. Applied Psychophysiology and
Biofeedback, 31(2), 129–142.
Wang, S. Z., Li, S., Xu, X. Y., Lin, G. P., Shao, L., Zhao, Y.,
et al. (2009). Effect of slow abdominal breathing combined with biofeedback on blood pressure and heart rate
variability in prehypertension. Journal of Alternative and
Complementary Medicine, 16(10), 1039–1045.
Wheat, A. L., & Larkin, K. T. (2010). Biofeedback of heart
rate variability and related physiology: A critical review.
Applied Psychophysiological Biofeedback, 35(3), 229–242.
Wood, P. B. (2006). Mesolimbic dopaminergic mechanisms
and pain control. Pain, 120(3), 230–234.
Wood, P. B., Schweinhardt, P., Jaeger, E., Dagher, A., Hakyemez, H., Rabiner, E. A., et al. (2007). Fibromyalgia
patients show an abnormal dopamine response to pain.
European Journal of Neuroscience, 25(12), 3576–3582.
Yasuma, F., & Hayano, J. (2004). Respiratory sinus arrhythmia: Why does the heartbeat synchronize with respiratory rhythm? Chest, 125(2), 683–690.
Zhang, Y., Ilsar, I., Sabbah, H. N., Ben David, T., & Mazgalev, T. N. (2009). Relationship between right cervical
vagus nerve stimulation and atrial fibrillation inducibility: Therapeutic intensities do not increase arrhythmogenesis. Heart Rhythm, 6(2), 244–250.
Zitnik, R. J. (2011). Treatment of chronic inflammatory
diseases with implantable medical devices. Annals of the
Rheumatic Diseases, 70(Suppl. 1), i67–i70.
Zucker, T. L., Samuelson, K. W., Muench, F., Greenberg, M.
A., & Gevirtz, R. N. (2009). The effects of respiratory
sinus arrhythmia biofeedback on heart rate variability
and posttraumatic stress disorder symptoms: Aa pilot
study. Applied Psychophysiological Biofeedback, 34(2),
135–143.
Part V
Practice Issues
Chap ter 14
Intake and Preparation for Intervention
Mark S. Schwartz
This chapter is about intake decisions and the
preparation of persons1 for interventions.2 It
includes considerations and guidelines for making
decisions about selection of persons, planning of
appropriate interventions, discussion of tailoring
intervention goals and cognitive preparations of
individuals, and sections that focus on interviewing, history taking, and self-report measures. One
basic intake decision in the therapeutic setting is
whether to use biofeedback3 with a specific person.
This chapter includes discussion of many factors in
this decision-making process.
It is my intention in this chapter that the content, although not exhaustive, will be sufficient for
most of the situations practitioners encounter, and
that practitioners will benefit from this information and guidelines. Readers who want more information about intake and considerations about
specific disorders are referred to other chapters in
this volume.
titioners first consider the published literature and
current practice.
Published Literature and Current Practice
Good research is one cornerstone of practice
and a basis for deciding which symptoms, conditions, and disorders to consider, and the persons for whom biofeedback-related interventions
are indicated. However, research often does not
capture the essence of applied applications, and
high-quality research is lacking for many conditions and behaviors. Therefore, practitioners need
not wait for well-controlled research to support
all procedures and applications before using biofeedback and associated therapies. However, they
do need to base decisions on sound logical and
responsible criteria. Other considerations become
more important when adequate research is lacking.
Prudent professionals know their limitations.
They also recognize the limits of published
research and the limits of other practitioners.
As practitioners, we preferably guard against the
problem of “not knowing what we do not know.”
We ideally places ourselves in the role of the person with whom we are providing interventions:
“Suppose I were that person” or “Suppose someone asked me to pay for the intervention for this
person. What questions would I ask? What compromises and accommodations would I ask for and
Conditions for Which Biofeedback
and Related Interventions Are Appropriate
When considering the conditions4 for which biofeedback and other applied psychophysiological
treatments are appropriate, one considers the individual for whom the interventions are intended,
the correctness of the diagnosis, and the specific
features and stages of the condition. Prudent prac217
218
regard as proper?” Prudent practitioners consider
several sources of data.
Conditions and Disorders
Agreement on a list of conditions and disorders for
which biofeedback is appropriate presents a challenge considering the many issues, options, and
diversity of providers. Practitioners consider many
sources when selecting conditions for which to
recommend using biofeedback, and for which persons. Unanimous or near-unanimous agreement
is unlikely. Whether a condition is on an indication list depends on the selection criteria, the
individual persons for whom biofeedback is being
considered, and the degree of caution preferred or
adopted by the involved parties developing the list.
In prior editions of this volume, I was tempted to
avoid lists, because lists are subjective, somewhat
arbitrary, subject to criticism, and may quickly
become obsolete. However, lists were included in
this chapter in prior editions. The three lists conveyed my various degrees of confidence in the literature and practice. The lists were merely guidelines. Furthermore, I offered additional cautions.
“Mindful practitioners never automatically accept
or reject for treatment all or most patients with a
disorder on any list.”
I decided to forgo the A, B, and C lists of the
prior editions and instead refer readers to efficacy
documents from the Association for Applied Psychophysiology and Biofeedback (AAPB; Yucha &
Montgomery, 2008), which review the efficacy for
41 types of disorders (a summary of these disorders
may be found at my website, www.marksschwartzphd.
com), in addition to the chapters in this book, of
course, and especially the literature found in the
journal Applied Psychophysiology and Biofeedback,
the AAPB “magazine” Biofeedback, and the International Society for Neurofeedback and Research
(ISNR) Journal of Neurotherapy. However, this
chapter provides some reminders, guidelines, caveats, and cautions to use when reviewing the conclusions and reviews from “official” sources.
• Even consideration of symptoms, conditions,
and disorders that are considered to have the
“best” support does not imply that biofeedback
alone is always the treatment of choice or the
sole or primary treatment. One exception is
nocturnal enuresis, for which forms of biofeedback can be the primary or sole treatment.
• Even for some symptoms, conditions, and disorders for which there is less than ideal support for
V. PRACTICE ISSUES
biofeedback used alone or as a major intervention component, it is still reasonable to consider
biofeedback as a legitimate part of an intervention plan. In many cases, other interventions
often play a larger role.
• For other symptoms, conditions, and disorders
for which there is weak support (e.g., case studies), one can consider including biofeedback
within a stepped-care approach, assuming that
the practitioner can clearly justify and document the rationale.
• All symptoms, conditions, and diagnoses require
careful patient selection and tailoring of treatment combinations to the individual. Inclusion
of a disorder in any listing is not intended to
imply that biofeedback is suitable for all or most
patients with this diagnosis.
One also can include symptoms, conditions, and
disorders for which relaxation and other applied
psychophysiological techniques are effective and
one uses biofeedback instruments to obtain a
more complete assessment. Biofeedback can help
some of these persons change their beliefs about
themselves, including their self-confidence about
making changes. Biofeedback helps some persons
to improve their self-regulation. Instrumentation
also allows practitioners to assess and document
progress.
One might also include other symptoms and
conditions from the field of physical medicine and
rehabilitation
Cautions and Contraindications
Prudent practitioners consider various cautions
and contraindications before using biofeedback
and other applied psychophysiological interventions. Experts generally agree about many of these;
however, there is no single, agreed-upon document
listing or discussing all cautions and contraindications. Adler and Adler (1984, 1989a, 1989b)
offered sage opinions regarding limitations of biofeedback, and I still recommend reading of those
chapters.
Consider the following disorders and conditions
as outright contraindications to biofeedback, or
at least as indicating the need for much caution.
These include severe depression; acute agitation;
acute or fragile schizophrenia (or a strong potential
for psychotic decompensation); mania; paranoid
disorders with delusions of influence; severe obsessive–compulsive disorder (OCD); delirium; acute
medical decompensation; or a strong potential for
219
14. Intake and Preparation for Intervention
a dissociative reaction, fugue state, or depersonalization. However, there is very little or no literature
on biofeedback or other applied psychophysiological interventions for patients with these disorders,
because logic has precluded such interventions
with these patients.
In the rare cases when a practitioner can justify
using relaxation and biofeedback with a patient
who has one of these diagnoses or conditions,
prudent standards of practice dictate using special
assessment and treatment procedures. For example, one might treat tension or migraine headaches
in a patient with OCD.
Caution must also be employed in using some
forms of biofeedback and relaxation therapies for
patients with certain other conditions. These are
not contraindications; however, providers must
be very knowledgeable and experienced with
these conditions, and well versed in using special
approaches. These conditions include moderate
to severely impaired attention or memory (as in
dementia and various forms of intellectual disability), as well as some seizure disorders. One also
needs to be cautious with patients with significant
“secondary gain” from symptoms, as would be true
for any psychological-based intervention.
Practitioners should inform patients taking
medications for certain medical disorders (e.g.,
diabetes mellitis, hypothyroidism, seizure disorders, hypertension, glaucoma, and asthma) that
relaxation therapies might result in a need for
reduced dosage, and should discuss this possibility
with the patients’ physicians. However, it should
be noted that reports documenting adverse effects
or altered medication requirements associated with
relaxation therapies and biofeedback are very rare.
have already ruled out medical causes and diagnoses, and who will reevaluate patients as indicated.
The scope of this chapter does not permit a
detailed discussion of this subject, which has
been well treated by others elsewhere. Readers are
referred to Chuang & Forman (2006), Maldonado
(2009), Taylor (2007), and to my website (www.
marksschwartzphd.com) for excellent, detailed
information.
A variety of types of medical disorders produce
psychological/psychiatric symptoms and signs, as
noted in Table 14.1. It is in the early stages of many
of these disorders that persons present with the
symptoms and practitioners are more likely to be
misled.
Note examples of classes of prescription medications that can induce psychological/psychiatric-like symptoms: chemotherapy, immunosuppressants (e.g., cyclosporine), antivirals (e.g.,
interferons), antiparkinsonian, cardiovascular,
thyroid, anticholinergic, corticosteroids, psychostimulants, sympathomimetics, sedative and central nervous system (CNS) depressants (e.g., barbiturates, benzodiazepines), opioids, hormones,
diuretics, antihistamines, vitamin B complex,
and nicotinic acid (Hall, 1980; Maldonado, 2009;
Othmer & Othmer, 2002).
Examples of substances that can induce psychological/psychiatric-like symptoms include alcohol,
cocaine, marijuana, phencyclidine (PCP), lysergic
acid diethylamine (LSD), heroin, amphetamines,
Jimson weed (street names are thornapple, stinkweed, locoweed), gamma-hydroxybutyric acid
(GHB; e.g., medical applications as prescription
medication Xyrem for cataplexy and excessive daytime sleepiness associated with narcolepsy).
Medical Conditions Masquerading
as Psychological Symptoms
Recommendations
All practitioners need to know and understand
that many symptoms “masquerade” as psychological or functional symptoms but actually are caused
by an organic medical disorder requiring different
interventions (Rosse, Deutsch, & Deutsch, 2000;
Othmer & Othmer, 2002; Hall, 1980). Practitioners need to be familiar with the symptoms and
other manifestations of diagnosed medical conditions, and other possible diagnoses and conditions.
Nonmedical practitioners particularly need to
exercise extra caution and consult medical specialists. Prudent practice standards dictate working
closely with competent physicians and other medical professionals in appropriate specialties who
The following recommendations are intended to
help practitioners avoid important mistakes:
• Be familiar with the special features of the many
medical conditions/disorders and the symptoms/signs for disorders known or suspected
among the practitioner’s patients/clients/subjects (Maldonado, 2009; Taylor, 2007).
• Be familiar with the features of psychological/
psychiatric conditions/disorders for those persons with one or more of these conditions/disorders.
• Make sure that patients/clients/subjects will
soon have or recently have had a competent
medical examination.
220
V. PRACTICE ISSUES
TABLE 14.1. Selected Symptoms and Signs from Selected Medical Disorders That May Masquerade
as Psychological/Psychiatric Disorders
Medical disorders
Symptoms/signsa may masquerade as . . .
Hypothyroidism
Tiredness, depression, myalgia, hypertension, bradycardia, cold periphery,
anxiety, mania, dementia, schizophrenia
Hyperthyroidism (Grave’s disease)
Irritability, restlessness, malaise, muscle weakness, tremor, breathlessness,
palpitations, anxiety (e.g., panic disorder or generalized anxiety disorder),
mood disorders, attention-deficit/hyperactivity disorder in children
Hypoadrenalism
Depression, nausea/vomiting, abdominal pain, constipation, joint or back
pain, fatigue, apathy, anorexia
Hyperadrenalism (Cushing’s disease)
Fatigue, mood lability, decreased mood, sleep disturbance
Pheochomocytoma
Anxiety resembling panic attacks
Hypoparathyroidism
Anxiety and panic
Pancreatic cancer
Anxiety, depression, dementia
Oat-cell lung cancer
Anxiety and panic
Systemic lupus erythematosus
Anxiety, depression, multiple symptoms in many body areasb
Wilson’s disease, multiple sclerosis,
cerebral ischemia
Anxiety and/or depression
Supraventricular tachycardia
Multiple panic-like symptoms (e.g., rapid heartbeats, often above 140 to
about 200 beats per minute)
aThe basis for selecting these few of the possible symptoms was the types of symptoms presented by persons most commonly seeking,
or referred to practitioners for, biofeedback and applied psychophysiology.
bDevelops over a long time, often many years.
• Maintain a close relationship with patients’/
clients’ physicians and/or other physicians who
can help.
• Recommend to patients/clients/subjects that
they maintain regular or periodic contact with
their physicians.
• Encourage patients/clients/subjects to report
any new symptoms, new and unusual behaviors,
and changes in existing symptoms.
• Know the drugs (prescription and over the
counter [OTC]) and substances your patients/
clients/subjects take to understand possible
symptoms and side effects.
Other Considerations in Choosing Treatments
Stepped Care
The stepped-care approach usually involves first
using less complicated, less expensive, and sometimes less time-consuming treatments. There is
a strong precedent for this approach in medicine
and other health care fields. For example, practitioners consider less potent medications or lower
doses before choosing more potent or higher doses,
and changes in diet and exercise before medication. The stepped-care model is very consistent
with the cost containment zeitgeist.
There are situations in which an even more
conservative approach than biofeedback is appropriate and preferable to try first. Patients and referral sources will usually be grateful and respectful
when practitioners are conservative and successful
with “less” rather than “more” treatment.
One can consider dietary changes (e.g., cessation of caffeine use) before relaxation and biofeedback. All patients need not stop caffeine use
before starting biofeedback and related therapies.
However, consider this in particular with patients
with headaches, anxiety, Raynaud’s disease, irritable bowel syndrome (IBS), psychophysiological
insomnia, and essential hypertension. Consider
221
14. Intake and Preparation for Intervention
stopping or changing other substances, and/or
changing medications that might be contributing
to symptoms (in consultation with the treating
physician).
At a minimum, discuss with such patients
the rationale for eliminating caffeine to reduce
interference with relaxation and biofeedback. If
stopping caffeine before other treatments leads
to improvement, there will be a credible demonstration of the effects of caffeine. However, even
when eliminating caffeine does not result in
symptom relief, it still allows for a more meaningful symptom baseline. Patients also may be more
motivated to comply with later recommendations,
and better able to do so, if they have first made
indicated dietary changes (see Block, Gylenhaal,
& Schwartz, Chapter 10, this volume).
Stepped care also includes other factors that are
often simple to change. For example, to stop gum
chewing is sometimes proper before relaxation and
biofeedback for tension-type headaches, temporomandibular (TMD) symptoms, or tinnitus. Consider the patient who chews gum a few hours per
day and has headaches in the temporalis muscles
or facial pain from probable daytime bruxism. This
simple change is a necessary first step that may be
sufficient.
Another disorder that may lend itself to this
approach is psychophysiological insomnia.
Tried and/or Available Alternative Treatments
Among the important initial factors to consider
are the prior therapies that have been attempted
and their outcomes. If relaxation or biofeedback
interventions were previously unsuccessful, consider the following:
• Find out exactly which interventions were used,
to avoid providing the same or similar interventions.
• Determine the patient’s/client’s understanding
of the rationale and procedures, and his or her
compliance with these.
• Determine the patient’s comfort level with the
prior practitioner.
• Ascertain whether the prior practitioner was
present or absent during sessions.
• Find out the instrumentation and modalities
used, and the placement sites of electrodes and
sensors.
• Find out the types of relaxation instructions
and body positions assumed during relaxation
and biofeedback.
• Determine the patient’s or client’s understanding and have him or her demonstrate desired
breathing techniques.
• Determine the degree of the patient’s past generalization and transfer of training procedures.
A therapist can usually obtain answers to these
and related questions within a few minutes. A
prior unsuccessful trial need not prevent another
trial with biofeedback, assuming an adequate
rationale, competent provider, and the presence
of other necessary criteria that are noted and
discussed in this chapter and others in this volume. Of course, there need to be indications that
another trial could improve upon important therapy procedures.
An adequate trial of proper medication might
offer a less expensive yet effective option that is
acceptable to many persons. Furthermore, biofeedback and associated therapies are often more easily
justified after an unsuccessful trial of other proper
therapies.
Severity and/or Seriousness
of Symptoms or Disorders
The practitioner also needs to consider the seriousness or severity of the patient’s symptoms and
disorder when deciding whether to offer biofeedback. Consider physiological self-regulatory therapies, including biofeedback, for some persons with
serious or severe symptoms even when there is
insufficient research and/or clinical experience to
support its use, especially when options are nonexistent, more risky, or far more expensive. Again,
this assumes an adequate rationale for biofeedback, a competent provider, and the presence of
other necessary criteria for such therapies.
Consider for example, a patient of mine who
presented with hyperemesis gravidarum, a serious
disorder involving unrelenting nausea and vomiting during pregnancy. Antiemetic medications
were either contraindicated because of her pregnancy or were no longer available in the United
States. Research studies on the use of biofeedback
and relaxation therapies for this disorder did not
exist. However, the clinical rationale was sound.
Practitioners and research support the successful use of such therapies for patients with functional emesis not associated with pregnancy (e.g.,
patients with anticipatory emesis associated with
chemotherapy for cancer). Alternative therapies
for this patient were far less practical. The only
other option was hospitalization and intravenous
222
nutritional therapy for the remaining 6 months of
her pregnancy or until the hyperemesis stopped. A
brief and intensive therapy program led to a substantial reduction in the emesis.
A practitioner might reasonably consider avoiding or deferring treatment if a patient’s symptoms
are mild or infrequent enough that he or she can
live comfortably with minimal or no treatment.
This category, for example, might include patients
with vascular headaches once every 2–3 months
that last a few hours and respond well to a safe
medication; patients with very mild tension-type
headaches two or three times a month, each lasting a few hours; or patients with bruxism without
pain or damage to the teeth or other oral structures
or tissues. Some practitioners encourage such people to live with their symptoms unless a brief and
inexpensive therapy program has a good chance of
success, or they may devote a session or two teaching them procedures for successful coping.
Geographical (or Other) Distance
between Patients and Treatment Facilities
Some patients live beyond a reasonable driving
distance from treatment facilities. Often a suitable
referral in a patient’s home area is not available.
Other patients prefer treatment away from their
home area, because they have had bad experiences there. Still others prefer or need to maintain strict confidentiality away from their home
area. Some prefer the type of therapy, professional
care, and credibility of a specific facility. In some
large medical centers, a patient can only stay for
a short time.
When there are good indications for biofeedback and related therapies, and more regularly
spaced treatment options are impractical, inappropriate, or nonexistent, a practitioner might consider a “massed-practice” therapy program. Such a
program involves one or two daily office sessions
for a few or several days. There are advantages to
this schedule, and sometimes it is the only or the
best option available. It can also serve to encourage the patient to continue therapy when he or she
returns home.
A limitation of massed-practice therapy away
from the home environment and usual routine is
that the patient might be experiencing less stress,
and there are no opportunities to practice techniques in real-world settings and to troubleshoot
with the therapist, and so forth. There also may
be self-imposed or implied pressure to accomplish
more than is reasonable in a short period of time.
V. PRACTICE ISSUES
The Professional’s Confidence and Competence
with Biofeedback
The practitioner’s confidence and competence
in using necessary instrumentation during evaluation and therapy are obviously important.
Whether other professionals believe they can
attain the same results without instrumentation
may be a moot point. If a professional uses instrumentation competently and prefers to do so, then
doing so is acceptable, regardless of whether noninstrumentation-based procedures might produce
similar results.
There are considerable precedents for this philosophy of practice in medicine and psychology.
For example, some psychologists and psychiatrists
prefer to rely usually or solely on detailed interviews and direct observations to make diagnoses
and recommendations. This is a common and
acceptable approach. In contrast, many professionals prefer to add and often rely on psychological assessments and testing, which can add significant costs to evaluations. However, the additional
assessments are common practice, and many
believe that they add to the quality of the diagnostic and therapy plan.
Similarly, many neurologists and other physicians request neuropsychological assessments to
help confirm a diagnosis or add information about
a patient’s functioning. Sometimes these merely
confirm practitioners’ clinical impressions. This is
acceptable and common clinical practice. Other
neurologists and physicians believe such assessments are unnecessary. This, too, is acceptable
clinical practice.
Alternative therapies and procedures may be
equally effective. Some studies show one treatment to be better, whereas other studies show that
a second treatment is better than the first; still
other studies may show them to be equal. Practitioners have the right to choose among evaluative
and therapeutic approaches that are consistent
with their interest and confidence given that their
interest and confidence are crucial to the therapeutic outcome. Of course, practitioners must also
keep in mind risks, costs, and efficacy.
Initial Physiological Evaluation
and Baseline Session(s)
The results of the psychophysiological baseline
assessment constitute an important source of
information for making therapy decisions. For
example, suppose that a practitioner observes con-
14. Intake and Preparation for Intervention
sistently low levels of muscle activity from multiple
muscle areas during rest and stressor segments.
In this situation, the practitioner should consider
deferring biofeedback, particularly surface electromyography (SEMG) biofeedback, or omitting biofeedback from the treatment plan. In such a case,
the instrumentation-based monitoring serves a
very useful purpose. It reveals that the patient can
relax within a therapeutic range. The practitioner
then clearly explains the meaning of that finding and instructs, encourages, and supervises the
patient in the use of relaxation. This includes using
relaxation frequently enough, long enough, and at
the right times for it to be of therapeutic value.
It may be that selected muscular tension is
occurring in the patient’s daily life but not in the
practitioner’s office. Without demonstration of
tension in the office, it may still be proper to use
non-instrumentation-based physiological self-regulatory procedures in the patient’s daily activities.
Another question that arises in this context
concerns the criteria for being sufficiently relaxed.
There are no hard and fast rules, and professionals disagree except at the extremes. The level and
duration of physiological activity needed for a positive therapeutic effect differ among patients. Practitioners often do not agree on necessary physiological criteria for relaxation to result in positive
results for the symptoms and disorders treated.
Reaching clear or ideal criteria during biofeedback-assisted relaxation office sessions is probably
often unnecessary. Patients often improve despite
somewhat tense levels in early and later baselines.
Practitioners consider recommending avoidance
of excessive tension, especially for sustained periods. A person might benefit from reproducing the
reduced tension observed in the baselines, even if
he or she has not reached the ideally relaxed range.
Indeed, the therapist should suggest this and help
the patient learn to do it frequently, rapidly, for
various durations, and at the right times.
Baselines need to be long enough to observe
increasing or decreasing physiological activity
over several minutes. Baselines that are too short
or those integrated over long periods can obscure
such trends. Also consider varying the conditions
under which baselines are conducted. Baselines
achieved with the person’s eyes closed can be very
misleading and inadequate, because many people
show much higher tension with their eyes open.
In addition, monitoring only during resting conditions without stressors provides unrealistic results.
Such recordings often show little or no tension
or arousal, whereas with anticipation of or dur-
223
ing stressful stimuli, tension and arousal are often
greater.
It should also be noted that psychophysiological
measurements can be unreliable across sessions.
Thus, the activity in one session does not reliably
occur in other sessions. In the first session, the person may show more tension because of the novelty
of the situation. This is one reason for sometimes
considering additional baseline segments in making decisions about the need for biofeedback.
Furthermore, some persons relax adequately at
home and in their real-life situations; however,
they have difficulty relaxing in a professional’s
office. No matter how a practitioner may present
it, there is an implied evaluative atmosphere in an
office that some people find difficult to overcome.
This is sometimes called “office hypertension” or
“white coat hypertension.” The practitioner who
suspects this should consider evaluating how the
person views the session and how he or she feels
during office sessions. The practitioner should
then attempt to check the physiological activity
during practice periods outside the office.
Other useful data include the physiological
responses during and after feedback segments. The
questions to be asked include the following:
• Is there significantly lower tension and arousal
with the feedback?
• Is the lowered arousal and tension maintained
after the feedback?
• Does the feedback lead to increased arousal and
tension?
• Do stressors increase the arousal or precipitate
arousal?
There are several questions and types of information on which one may base the decision to provide more biofeedback. Here are a few examples:
• Is there much excess tension and arousal during
resting baseline and self-regulation segments
without feedback?
• Is much tension and arousal precipitated or
worsened with office stressors?
• Does feedback result in significantly lowered
tension and arousal?
• Does the person return to baseline tension and
arousal after feedback?
It is instructive to remember that such logical
questions and criteria have not yet been clearly
shown to predict the necessity of biofeedbackassisted therapies for achieving positive therapeu-
224
tic outcomes. Until research supports such evidence, practitioners need to remain cautious and
conservative in making such decisions and recommendations.
Symptom Changes in the First Weeks
Symptom changes during the first weeks of therapy are important for deciding whether to continue sessions and determining which therapy procedures to pursue. Such improvements often occur
early, before reaching ideal physiological mastery.
If the initial intervention results in a clinically significant reduction of symptoms, practitioners need
to justify more office-based intervention.
The primary goal is to decrease or stop symptoms, not just to reduce microvolts, increase hand
temperatures, or change the electroencephalographic (EEG) frequency or pattern. When there
are clinically significant changes in the first
few weeks, some prudent practitioners consider
deferring more biofeedback sessions, even if they
assume that some persons need specific physiological mastery for reliable therapeutic changes. In
addition to cost considerations, another reason for
this determination is credibility.
Consider preparing the patient for the possibility that more office interventions may be needed
later. As an example, here is a statement that
might be made to a person who has shown clinically significant improvement of symptoms in the
early weeks of therapy.
“I am happy for your rapid improvement with
significantly reduced hours of severe headache,
increased hours without headaches. You also are
showing improved ability to reduce muscle tension. However, the muscle tension in your head
is still there when we measure it in the office.
It might be better to relax these muscles deeper
and faster, and maintain the lower levels longer. More biofeedback sessions could speed up
improvement, but I cannot predict that with
certainty. Let’s consider you continuing to apply
the therapy and keep records of your symptoms
for another few weeks and then review the situation.”
Patient Characteristics
Even when studies indicate that person characteristics significantly differentiate more successful
from less successful subjects, there is usually much
overlap between groups. There are many variables
V. PRACTICE ISSUES
that influence the relationships. These studies
need replications in a variety of clinical settings
and with potential “moderator variables” before
we conclude that some persons should not receive
an intervention.
A question to ask here is what to do with persons with different characteristics: Do we practitioners need to provide improved education? Will
more frequent or more closely spaced sessions work
better? Will procedural variations help?
Patient’s Motivation and Compliance:
Enhancement with Biofeedback
Maintaining motivation for practice and application of physiological self-regulatory therapies are
needed and include persons who can benefit from
non-instrumentation-based intervention (e.g.,
relaxation only). For many persons, biofeedback is
confirming and encouraging. It helps the patient
maintain or gain confidence in the intervention
and abilities. Some question and dispute the idea
that they are physically tense or sympathetically
aroused. Others doubt that their thoughts affect
their physiological tension and arousal. Still others
doubt that they can control their physiology. They
all need concrete and credible evidence. Using
instrumentation with physiological self-regulation
procedures is not a rejection of the value of relaxation alone. It is not an either–or choice, as some
portray it.
The Patient’s Choice and Cost of Therapy
Respect each person’s needs and treatment
choices, and realistically discuss their options,
time needed, desired results, prognosis, and costs.
As practitioners, we want to see much improvement. We often measure success against research
criteria such as symptom reductions of 50, 70, or
90%, along with reduced medications when indicated. Indeed, we strive for these ideal goals. However, such goals may not match the patient’s goals,
and they are sometimes unrealistic. Some people
welcome an improvement of 20–50%, especially
after years of very little change. Other people
have learned to be less distressed by the symptoms, although the intensity and duration have
not decreased. For example, patients with chronic
pain sometimes report improvement in “affective/
reactive” aspects rather than “sensory/intensity”
features.
A person may decide that additional improvement is not worth the additional financial and
225
14. Intake and Preparation for Intervention
time investments, and time away from other priorities. Also consider the costs and inconvenience
of transportation and child care. Allow people to
participate in choosing how much benefit they
desire or will accept.
An Introduction to Patient Education
Rationale
Some people are skeptical, critical, and resistant to
accepting the potential benefits of relaxation, biofeedback, and other applied psychophysiological
interventions. That should not lead us to become
defensive or dismiss them as unsuitable candidates. Many have already seen other practitioners
who were optimistic about interventions, yet the
results were unsuccessful. We are now asking them
to accept a different approach, often perceived as
the last one.
People are very often unfamiliar with behavioral and self-regulatory strategies. They often
need information they can understand and accept
about the rationale for these interventions. Furthermore, some people are skeptical about some
practitioner specialties (e.g., nonmedical health
care professionals).
Relaxation and biofeedback interventions can
appear to be rather simplistic. In contrast, explanations can appear very complex (e.g., some patients
may think to themselves something like the following): “You mean I have had these symptoms
for years, went to several good doctors, took lots of
medication, and continue to suffer. Now you tell
me that this relaxation (and/or biofeedback) is all
that I need? I would like to believe that . . . but
convince me!”
Do not assume that others understand and
accept the rationale for intervention, that explanations are sufficient, or that patients spontaneously ask questions or directly tell us their concerns. They usually do not!
Furthermore, we should not assume that people
accept and remember explanations and recommendations. Thus, seriously consider well-planned
and well-executed presentations to increase attention, understanding, recall, confidence, satisfaction, and compliance. Devote adequate time early
in the relationship and thereafter to prepare and
teach adequately. The values of good education are
often underestimated or neglected.
Good education and information can also
reduce anxiety, increase the credibility of the
professional and the intervention procedures,
and facilitate positive expectations. As Shaw and
Blanchard (1983, p. 564) concluded,
giving participants a high initial expectation of therapeutic benefit . . . has significant benefit in terms
of self report of change and reduced physiological
reactivity, and . . . these improvements are mediated
at least in part by increased compliance with home
practice instructions. . . . The procedures, per se, are
not especially powerful without the appropriate set.
They also noted that “a certain degree of salesmanship and trainer enthusiasm certainly can
make a difference in outcome” (p. 564).
Good educational presentations can improve
patients’ knowledge and attitudes about the causes
of their functioning and about intervention. It can
enhance their perceptions of a practitioner (e.g., as
being credible and trustworthy). Adherence and
intervention effectiveness partly depend on such
knowledge, beliefs, perceptions, and a positive alliance.
Metaphors
Tailoring educational presentations to a specific
person depends on that person’s intelligence, education, reading ability, sophistication, and psychological mindedness. Professionals commonly use
metaphors to communicate and educate. One of
the major reasons for using metaphors is to help
simplify information, concepts, and procedures.
Metaphors are excellent for presenting the rationale, concepts, and ideas that people need to
understand. This can make the ideas easier to
accept and use (Combs & Freedman, 1990). One
definition of “metaphor” is that it “is a way of . . .
describing something in terms of something else”
(Morris & Morris, 1985, p. 387).
Metaphors are much more than simply analogies (Black, 1962; Richards, 1936). Muran and
DiGiuseppe (1990) and Boyd (1979) describe metaphors as “cognitive instruments” by which similarities are created that previously were not known
to exist. Metaphorical communication is highly
persuasive as a means of conveying and altering
thought and is a vehicle of change (Petrie, 1979).
Siegelman (1990) starts her book by stating, “Most
of us, and our patients . . . find ourselves cleaving
to metaphor to communicate experience that is
hard to convey in any other way” (p. 1).
Practitioners of cognitive-behavioral therapies
also advocate analogies and metaphors (Stott,
Mansell, Salkoskis, Lavender, Cartwright-Hatton,
226
2010). McMullin (1986) describes many examples of perceptual shift techniques for cognitive
restructuring therapy. Although professionals use
many metaphors, further discussion or examples is
beyond the scope of this chapter.
Cautions in the Use of Metaphor
Many linguists and psychologists warn about the
potential misuse and misleading potential of metaphors. For example, the use of metaphor can foster
careless thought “by acting as a substitute for the
hard, analytic work of determining precisely what
to say, a point previously raised by Aristotle . . .
when he warned of the ambiguity and obscurity
inherent in metaphor” (Muran & DiGiuseppe,
1990, p. 72).
Practitioners need to be careful about why they
are using a metaphor, with whom, and in what context.
One premise of this section is that metaphors
can improve cognitive preparation and education
of people, and that improved cognitive preparation improves adherence for effective intervention
results (Levy, 1987).
Evaluation/Assessment: Interviewing, History
Taking, and Self‑Report Measures
Psychological Evaluation
Deciding where to begin history taking, and
whether (and, if so, how soon) to include psychological inquiries, depends on practitioner judgment, circumstances, and the patient/client. With
many medical patients, mental health practitioners are often wise to begin with a history of physical symptoms. However, exceptions abound. Often
all or most of the symptom history information is
available in the recorded history.
With medical patients, there are often practical constraints for psychological evaluations (e.g.,
schedules, resistance due to limited psychological
mindedness). However, even a brief psychological
evaluation can be helpful. Asking a few psychological questions can help build rapport and assess
receptiveness or resistance to this type of question
and intervention. It also can help with deciding
whether or not more evaluation is needed.
The following list of psychological factors is
based on one by Adler and Adler (1987), who
provide an erudite, insightful, refined, and skilled
commentary on history taking. One must read
their original text to appreciate their style and
V. PRACTICE ISSUES
clinical wisdom. Although it was written as a
guide for interviewing people with headaches,
their list and discussion are useful for other disorders. The Adlers suggest considering evaluation of
many factors:
• Patients’ expectations of themselves.
• Perceived expectations by others.
• Existence of past or present family conflicts.
• Sensitivity to criticism and to emotional expressions.
• Comfort with and skills in being assertive.
• Illnesses and hospitalizations.
• Past or present grief, or anticipated grief.
• Medication misuse.
• Perceptions of health care professionals.
• Perceived emotional triggers or factors increasing the risk of a symptom.
Do patients’ personality features and psychopathology worsen or maintain current symptoms,
or are they the effects of chronic symptoms? How
necessary is it for practitioners to assess and treat
psychopathology to reduce current physical symptoms significantly? Also, do life stressors in the
past cause or contribute to current symptoms? For
example, what is the role of past sexual abuse or
grief in current symptoms? How necessary is it for
practitioners to assess and treat these factors to
reduce current physical symptoms significantly?
History Taking and Interviewing
There are many resources for history taking and
interviewing (Brannon, 2011; Hersen & Turner,
1985; Othmer & Othmer, 2002; Sommers-Flanagan & Sommers-Flanagan, 2008). Practitioners
often use interview outlines as guides for specific
conditions (e.g., consider Lacks [1987] and Morin
and Espie [2003] for insomnia). The topics and
specific items covered and the time invested for
each depend on many factors: professional setting, professional specialty, referral source, referral
information available, whether there will be continuing care by another professional, results from
screening measures, stepped-care considerations,
cost consideration for the patient, time available
by the patient and the practitioner, and purpose of
the consultation or evaluation.5
Practitioners responsible for assessment and
treatment are wise to obtain at least some of their
own history information rather than to rely exclusively on information from others. This is true
even when the other sources are competent profes-
14. Intake and Preparation for Intervention
sionals. A practitioner needing specific information often must obtain it directly from the patient.
The practitioner can review the prior reports aloud
with the patient for his or her confirmation and
elaboration. Even seemingly clear information,
such as onset, location, frequency, and duration,
can differ when one asks the questions and listens
carefully to the answers.
Even competent and experienced professionals
can overlook potentially important items. This
does not mean that they are careless or incompetent. Patients give different professionals different
information and provide different answers to the
same types of questions (e.g., Blanchard, O”Keefe,
Neff, Jurish, & Andrasik, 1981). Practitioners may
also misunderstand patients’ statements. Furthermore, practitioners, including physicians, sometimes obtain only the information needed for the
purpose of their consultation—which may be to
make a diagnosis; rule out serious organic pathology; prescribe medication; and/or make referrals
for psychological evaluations, biofeedback, physical therapy, or other treatments.
Physicians and other practitioners with special
interests and expertise in specific symptoms and
disorders often collect more detailed information
than do other professionals. For example, for headaches, these areas of information include dietary
factors, gum-chewing habits, use of bed pillows,
sleep habits, stress, work postures and other ergonomic factors, driving habits, beliefs, and sexual
and physical abuse. The practitioner who observes
discrepancies between the recorded history and
the information he or she now receives from the
patient should address these discrepancies tactfully.
Sexual Abuse
Research reveals that sexual abuse is often part of
the history of patients presenting with multiple
somatic disorders (e.g., functional6 gastrointestinal disorders, nonspecific chronic pain, psychogenic seizures, chronic pelvic pain, fibromyalgia
(when sexual abuse was defined as rape); Paras et
al., 2009). There is separate support for the relationship between emotional, physical, and sexual
abuse in persons with fibromyalgia syndrome
(Häuser, Kosseva, Üceyler, Klose, & Sommer,
2011). Sexual abuse history also is associated with
later risks for multiple psychiatric/psychological
disorders (i.e., anxiety disorder, depression, eating
disorders, posttraumatic stress disorder [PTSD],
sleep disorders, and suicide attempts; Chen et al.,
227
2010). These results are independent of gender or
age when the abuse happened.
One conclusion and implication is that practitioners need to inquire about a history of possible
sexual and physical abuse at least among persons
presenting with the types of symptoms and disorders noted here.
The interview questions from Drossman et al.
(1990) comprise one set of possible questions.
Practitioners should review other sources and
guidelines for self-report and interview questions,
and consult with specialists in this field before
deciding how to assess this complex and delicate
topic.
Evaluating this topic of abuse and considering
proper intervention are of potential value to biofeedback practitioners. One potential advantage of
knowing about abuse is the opportunity to consider its possible influence on the development
and/or maintenance of medical and psychophysiological problems.
There are now substantial legal concerns
for health care professionals when questioning
patients about a history of sexual abuse, and when
providing psychotherapy for patients based on the
history (Cannell, Hudson, & Pope, 2001; Scheflin & Spiegel, 1998). This concern is particularly
important if a professional is inquiring about
and/or suspects a history of chronic sexual abuse
(CSA) that the patient is not reporting. Prudent
health care professionals guard against overzealousness, are well informed about the topic of CSA
and so-called “repressed memories,” the needs
for careful informed consent, and the other legal
issues and recommendations (see www.apa.org/
topics/trauma/memories.aspx).
Clinicians are well-advised to practice defensively in cases involving memory or dissociative
disorders, especially by keeping extracautious
notes and more frequent use of informed consent
forms. Books on legal risk management are important preventive guides that could help avoid costly
and unnecessary lawsuits. (Scheflin & Spiegel,
1998).
However, the history of abuse in some patients
may sometimes be very significant. “Failing to
inquire about a history of trauma, and therefore
assuming that it did not occur or if it did is unimportant, can be as damaging as insisting that a
trauma history must lurk behind any symptom”
(Scheflin & Spiegel, 1998, p. 861). Does it perhaps influence tolerance for symptoms? Does it
contribute to those factors that motivate people
with these symptoms to seek medical help? Does
228
V. PRACTICE ISSUES
it perhaps contribute to the factors that motivate
people to avoid treatment? Self-blame, poor selfimage, control issues, shame, trust, vulnerability,
dependency conflicts, sexual dysfunction, and
suppressed anger could affect all of these.
Interview Outline
The following outline or checklist offers a guideline to consider and from which to glean ideas for
interviewing and other intake procedures. Most of
these questions and items are useful in the general clinical practice of applied psychophysiology,
and especially in the treatment of headache and
anxiety disorders. Some items and questions do
not apply to other disorders, such as incontinence.
(See Andrasik & Schwartz, Chapter 20, for a more
detailed discussion of taking a headache history.)
1. Symptom(s). (Note: A patient’s highest-pri-
ority symptoms are not always the reason for
referral.)
a. Description. “What are the symptoms
like?” [Offer choices.]
b. Location. “Where does it begin? Show me.
Does it move around?”
c. Frequency. “How often does it occur?
When does it increase–decrease?”
d. Timing. “When do the symptoms occur?
What time of day do they occur? Do they
always or usually occur?”
e. Duration. “How long do the symptoms
last? Do they last for . . . ? What are the
shortest, longest, and usual durations?”
f. Intensity. [Consider rating scales.] “Are
the symptoms slight, mild, moderate,
severe, or very severe?”
g. Origin. “When did the symptoms originally begin?”
h. Development. “Has it changed over weeks,
months, or years?”
i. Course/progression. “Does it change over
minutes/hours after it starts?”
j. Precipitants/antecedents.
[Look
for
dietary, environmental, postural, hormonal, emotional, work/family stress, and
time factors.] “What do you think causes
or starts the symptoms? Do you suspect
that anything might be triggering it? Is
there anything that often seems to precede it?”
k. Aggravating/worsening factors. “Does
anything increase the severity? What
makes it worse?”
l. Alleviating/helping factors. “Does any-
thing decrease the severity? What makes
it better? What do you do that reduces the
symptoms?”
m. Medications. (Note when and why medications are taken. Are medications taken
when the patient is anticipating situations? Does the patient take medications
with minimal symptoms?) “When do you
take medications? How soon do you get
relief after taking . . . ?”
n. Nonrelief. “What has not worked for you?
When was it taken?”
o. Periods of remission. “Are you ever totally
free of symptoms for days, weeks, longer?”
p. Associated symptoms. “What other symptoms do you get with the main one? Do
you get . . . ?” (Ask specific questions about
specific symptoms.)
q. Reactions of others. “What does your family do when you have symptoms?”
r. Behaviors before, during, and after onset,
including behaviors and attitudes on days
without the symptoms. “On days when you
are feeling much better or have no symptoms, do you try and catch up with house/
yard work, and other activities? Do you
typically have worsening or resumption of
symptoms soon after or the next day? Do
you feel you need to fulfill your responsibilities on good days?”
s. Limitations in life due to symptoms.
t. Family members with similar or the same
symptoms (optional).
2. Prior treatments.
a. Prior psychological treatments. (Ask
when, where, with whom; number of sessions, duration, results; patient’s reactions
and views.)
b. Prior experience with relaxation therapies. (Ask about prior relaxation therapies,
what was done, how long this was used,
what is still done, when it is done, what
seems to work, what does not work. Ask
for demonstration, especially breathing.)
c. Prior experience with biofeedback. (Ask
who provided it, where on body sensors
were placed, whether eyes were open or
closed, what body positions were used,
what was done during sessions, and
whether patient was alone or with therapist. Ask about perceptions and attitudes
about this treatment.)
d. Other treatments. (Ask about other thera-
14. Intake and Preparation for Intervention
pies, including other applied psychophysiological therapies; when, number, and
what helped; and patient’s attitudes about
these treatments, including either desperation or open-mindedness.)
3. Current treatments. (Psychological and medical. Obtain names and addresses of professionals seen, and ask about attitudes, comfort,
expectancies, content, preferences, questions,
and plans.)
4. Attitudes about health care professionals,
treatments, and symptoms.
a. Symptoms. “What do you think is causing your symptoms? Do you think anything has been overlooked?” [This is when
one learns of a patient’s beliefs and fears
about a cause not yet found.] “What are
your thoughts when symptoms start and
worsen?” [Cognitive factors.] “How would
your life be different without these symptoms or with greatly reduced symptoms?”
b. Treatments. “What do you expect from treatments? What have you heard or read
about this treatment? What did your physician tell you?”
c. Professionals. “What do you think/feel
about coming to a [e.g.,
psychologist]?”
5. Reasons for seeking treatment now. “Why
have you come for treatment now? Are your
symptoms worse? Have new features? Is your
depression worse, or is your job or marriage at
risk?” (Is the patient planning life changes,
such as pregnancy, that entail a need to
stop medications? Is the patient returning
to school, changing jobs, getting married, or
making other major changes calling for better
treatment for the symptoms? Is there another
agenda, such as secondary gain, or seeking
help as a socially acceptable means of access
to the practitioner?)
6. Stressors. (Check past and current stressors
in interview and/or questionnaires. Stressor
areas include interpersonal, work, schedule
overload, perfectionism, procrastination,
disorganization, inefficient time use, lack of
goals and priorities, family, financial, health,
sexual, living conditions, legal, existential.)
7. Emotions. (Observe, ask, and consider measures for depression, anxiety, anger.)
8. Neurocognitive factors. (Observe, review
records, ask, and consider assessing for longterm limitations or acquired impairments in
memory, attention/concentration, intellec-
229
tual, language, academic achievement. Check
for head/brain injuries and surgeries with
residual effects.)
9. Physical factors. (Observe, check records. Ask
about hearing, vision, and physical limitations.)
10. Dietary and chemical intake. (Check records
and ask for past and current use of caffeine,
tobacco, alcohol, other vasoactive substances
[e.g., tyramine and monosodium glutamate,
other stimulants and depressants, gum chewing, other foods and dietary substances, and
so-called “street” or “recreational” substances/
drugs; see Block et al., Chapter 10, this volume].)
11. Medications. (Check records. Ask about all
prescription and OTC medications, results,
and side effects.)
12. Health-promoting behaviors. (Check records.
Ask about exercise, time use management,
dietary, vacation.)
13. Social support systems. (Check on family and
friends, church/synagogue activities, volunteer and other organizational activities. Ask
where children and other family members
live, their relationships with the patient, frequency of visits with them.)
14. Education and work history (recent and current).
15. Sleep. (Check for at least basics, such as bedtime and awake time, sleep-onset latency,
sleep interruptions and durations, sleeping
partner’s observations [e.g., snoring, breath
stopping, teeth grinding]. Check for feelings
after morning awakening and daytime sleepiness. Consider Epworth Daytime Sleepiness
questions [Johns, 1991; http://epworthsleepinessscale.com].)
16. Abuse. (Check for physical and sexual abuse
in childhood, adulthood. If recorded history
or other sources provide insufficient information, consider asking. The intake interview is
usually not the time for detailed discussion of
this topic or probing unless the patient wants
to talk about it then. Gaining this information is a very delicate matter fraught with
complex subtleties. Factors guiding whether
to obtain this information and how much
include presenting symptoms; purposes of the
interview; time available; likelihood of seeing the patient again; and your own experience, skills, and comfort. Consider asking, at
least, “Did you experience any abuse, sexual
or physical, as a child or adolescent?” or “Is it
230
possible that you experienced anything as a
child or adolescent that one might consider
sexual or physical abuse?” A patient’s equivocal response, such as “Not that I remember,” is
a cue to a history of possible abuse. Consider
inquiring further, or wait for or create chances
for further inquiry in later sessions—for example, “When I asked you about abuse, you said
you didn’t think so as far as you could remember. Is there anything you want to tell me or
do you have any questions?” [If this leads to
the possibility, probability, or confirmation of
abuse, consider a consultation with or referral
to specially trained and experienced professionals].)
17. Recommendations: Considerations and discussions. (Further evaluation with interview,
inventories, self-report measures, or neurocognitive assessment. Referral, psychophysiological assessment, multiple types of relaxation therapies and demonstration, symptom
log, biofeedback-assisted therapies. Consider
“prudent limited office treatment” (PLOT),
stepped-care options, and so forth. Discussion/patient education on varied topics tailored to patient.)
Self‑Report Measures as Part of Intake
Rationale, Uses, and Issues
The usefulness of self-report measures in clinical
practice is well established among most practitioners (e.g., Turk & Melzack, 2010). Their use has
many advantages, summarized in the list below.
These measures can provide information about
topics not obtained during interviews and observations. They can shed light on unclear behaviors
and provide hypotheses to explain these. They
provide quantification and documentation of
many variables of interest and are often necessary
for reports to other professionals and third-party
payers. The usual issues are selection of measures,
when and how to use them, interpretation, and
costs.
Some professionals argue persuasively that there
are situations in which it is prudent to administer
sets of such measures routinely. Bradley, McDonald-Haile, and Jaworski (1992) state that in their
inpatient program, they educate, prepare, and
reassure patients before the evaluation that
the psychological evaluation [is] part of the medical diagnostic process. In order to reduce patients’
concerns that their symptoms are not viewed as
V. PRACTICE ISSUES
legitimate . . . they are informed that the psychological assessment is mandatory for all patients . . . performed . . . prior to completion of the medical diagnostic procedures. . . . required to identify interactions
between pathophysiological and psychologic[al] processes that affect patients’ physical symptoms, disabilities, and social and familial activities . . . [and]
also may suggest interventions that might help to
reduce the patients’ suffering.” (p. 194)
However, there are clinical situations in which
such measures are unnecessary and not costefficient. They sometimes do not add enough to
clinical decision making and treatment plans to
justify the required time and expense. They also
sometimes can interfere with desired rapport and
the therapeutic alliance between practitioner and
patient. Many practitioners are skilled interviewers and highly experienced clinicians; self-report
paper-and-pencil measures often do not provide
much more information than such practitioners
can gain in a good interview.
Let us consider, for example, a consultation to
decide the appropriateness of biofeedback for tension-type headaches for a probable work-posturerelated tension myalgia. Let us further assume that
this is a consultation with a patient who is resistant to seeing a psychologist. Now consider the
potential perceptions and reaction of this patient
to a series of self-report mood and personality measures.
The decision to proceed or not to proceed with
biofeedback and related therapies will be the same,
regardless of information gained from paper-andpencil self-report measures. Skilled practitioners
can often base such decisions on prior recorded
information and an interview. This sounds like an
argument against the use of the measures. That is
not my point at all; however, I maintain that one
must use them prudently and not routinely in all
clinical situations. The merits of self-report measures include their ability to do the following:
• Document symptoms, personality, beliefs, and
behaviors.
• Document changes or lack of changes.
• Direct practitioners to areas needing more time
and effort.
• Increase patients’ awareness of their beliefs,
behaviors, and personality factors.
• Provide a basis for feedback to patients about
attending to their beliefs, behaviors, and personality factors.
• Correct some self-misperceptions by patients.
• Provide cautions for practitioners.
231
14. Intake and Preparation for Intervention
• Confirm or disconfirm impressions from interviews.
• Correct practitioners’ misperceptions of some
patients.
• Generate hypotheses about possible problems
and treatments.
• Assist less experienced practitioners.
• Potentially save interviewing and treatment
time.
• Raise topics, beliefs, and behaviors for discussion.
• Select patients needing special attention.
The selection of measures is the prerogative
of individual practitioners and depends on many
factors. A detailed discussion of these factors is
beyond the scope of this chapter; a brief list will
suffice:
• Availability of the measures.
• A patient’s motivation and availability for the
time needed.
• A practitioner’s experience with the measures.
• Brevity of the measures and ease of administration and scoring.
• Reading level of the measures and reading ability of the patient.
• Useful and/or important clinical and treatment
plan questions needing information obtainable
from the measures.
Conclusion
This chapter has covered topics and guidelines for
selecting whom to treat with biofeedback or other
applied psychophysiological interventions (including other physiological self-regulation therapies).
It includes topics and guidelines for intake interviewing and patient education.
Notes
1. Terms such as “persons” and “individuals” are often
used in this chapter rather than relying solely on the terms
“patients’ and “clients.” (See Note 4.)
2. The term “intervention” is used in this chapter
rather than “therapy,” which was used in prior versions of
this chapter in previous editions of this volume. “Intervention” is used commonly in multiple fields (e.g., interventional cardiology), so it is appropriate for those practitioners
working in medical and mental health fields, and it works
adequately for those working with sports and performing
artists, as well as other applications.
3. In this chapter, the term “biofeedback” subsumes
both biofeedback and neurofeedback or EEG biofeedback.
This convention was adopted in part for practical reasons
(i.e., space and readability).
4. Conditions include diagnoses, behavioral habits and
patterns of athletes and performing artists, subjects in studies, students, and others. In prior editions of this volume,
this chapter referred to “diagnoses” rather than “conditions” because I worked in a medical setting and the vast
majority of applications involved diagnoses of patients/
clients, and the applications for sports and performing artists were not so far along in their development. This is no
longer the case.
5. Practitioners’ evaluations of patients for therapy with
other professionals will differ from their evaluations of
patients they themselves intend to see for therapy.
6. “Functional” here signifies the absence of a structural, infectious, or metabolic cause (Berkowitz, 1998).
References
Adler, C. S., & Adler, S. M. (1984). Biofeedback. In T. B.
Karasu (Ed.), The psychiatric therapies: The American
Psychiatric Association Commission on Psychiatric Therapies. Washington, DC: American Psychiatric Association.
Adler, C. S., & Adler, S. M. (1987). Evaluating the psychological factors in headache. In C. S. Adler, S. M. Adler,
& R. C. Packard (Eds.), Psychiatric aspects of headache.
Baltimore, MD: Williams & Wilkins.
Adler, C. S., & Adler, S. M. (1989a). Biofeedback and
psychosomatic disorders. In J. V. Basmajian (Ed.), Biofeedback: Principles and practice for clinicians (3rd ed.).
Baltimore, MD: Williams & Wilkins.
Adler, C. S., & Adler, S. M. (1989b). Strategies in general
psychiatry. In J. V. Basmajian (Ed.), Biofeedback: Principles and practice for clinicians (3rd ed.). Baltimore, MD:
Williams & Wilkins.
Berkowitz, C. D. (1998). Medical consequences of child
sexual abuse. Child Abuse and Neglect, 22(6), 541–550.
Black, M. (1962). Models and metaphor. Ithaca, NY: Cornell
University Press.
Blanchard, E. B., O’Keefe, D., Neff, D., Jurish, S., & Andrasik, F. (1981). Inter-disciplinary agreement in the diagnosis of headache types. Journal of Behavioral Assessment, 3, 5–9.
Boyd, R. (1979). Metaphor and theory change: What is
metaphor for? In A. Ortony (Ed.), Metaphor and thought.
New York: Cambridge University Press.
Bradley, L. A., McDonald-Haile, J., & Jaworski, T. M.
(1992). Assessment of psychological status using interviews and self-report instruments. In D. C. Turk & R.
Melzack (Eds.), Handbook of pain assessment. New York:
Guilford Press.
Brannon, G. (2011). History and Mental Status Examination. Retrieved from http://emedicine.medscape.com/
article/293402-overview#aw2aab6b3.
232
Cahill, C., Llewelyn, S. P., & Pearson, C. (1991). Treatment
of sexual abuse which occurred in childhood: A review.
British Journal of Clinical Psychology, 30, 1–12.
Cannell, J., Hudson, J. I., & Pope, H. G., Jr. (2001). Standards for informed consent in recovered memory therapy. Journal of the American Academy of Psychiatry and
the Law, 29, 138–147.
Chen, L. P., Hassan, M., Paras, M. L., Colbenson, K. M.,
Sattler, A. L., Goranson, E. N., et al. (2010). Sexual
abuse and lifetime diagnosis of psychiatric disorders:
Systematic review and meta-analysis. Mayo Clinic Proceedings, 85(7), 618–629.
Chuang, L., & Forman, N. (2006, April 13). Psychiatric presentation of medical illness: Mental disorders secondary to
general medical conditions. Retrieved May 11, 2015, from
http://www.nepsychotherapy.com/id3.html.
Combs, G., & Freedman, J. (1990). Symbol, story, and ceremony: Using metaphors in individual and family therapy.
New York: Norton.
Drossman, D., Lagerman, J., Nachman, G., Li, Z., Gluck,
H., Toomey T., et al. (1990). Sexual and physical abuse
among women with functional and organic gastrointestinal disorders. Annals of Behavioral Medicine, 113,
828–833.
Hall, R. C. W. (Ed.). (1980). Psychiatric presentation of medical illness. New York: Spectrum.
Häuser, W., Kosseva, M., Üceyler, N., Klose, P., & Sommer, C. (2011). Emotional, physical and sexual abuse in
fibromyalgia syndrome—a systematic review with metaanalysis. Arthritis Care and Research, 63(6), 808–820.
Hersen, M., & Turner, S. M. (1985). Diagnostic interviewing.
New York: Plenum Press.
Johns, M. W. (1991). A new method for measuring daytime
sleepiness: The Epworth Sleepiness Scale. Sleep, 14(6),
540–545.
Lacks, P. (1987). Behavioral treatment for persistent insomnia.
New York: Pergamon Press.
Levy, R. L. (1987). Compliance and clinical practice. In
J. A. Blumenthal & D. C. McKee (Eds.), Application
in behavioral medicine and health psychology: A clinician’s source book. Sarasota, FL: Professional Resource
Exchange.
Maldonado, J. R. (2009). Neuropsychiatric masquerades:
Medical and neurological disorders that present with psychiatric symptoms: Part 1. Presented at 22nd Annual U.S.
Psychiatric and Mental Health Congress, Las Vegas, NV.
McMullin, R. E. (1986). Handbook of cognitive therapy techniques. New York: Norton.
Morin, C. H., & Espie, C. A. (2003). Insomnia: A clinician's
guide to assessment and treatment. New York: Springer.
V. PRACTICE ISSUES
Morris, W., & Morris, M. (1985). Harper dictionary of contemporary usage (2nd ed.). New York: Harper & Row.
Muran, J. C., & DiGiuseppe, R. A. (1990). Towards a cognitive formulation of metaphor use in psychotherapy.
Clinical Psychology Reviews, 10, 69–85.
Othmer, E., & Othmer, S. C. (2002). The clinical interview
using DSM-IV-TR (Vol. 1). Washington, DC: American
Psychiatric Press.
Paras, M. L., Murad, M. H., Chen, L. P., Goranson, E. N.,
Sattler, A. L., Colbenson, K. M., et al. (2009). Sexual
abuse and lifetime diagnosis of somatic disorders: A systematic review and meta-analysis. Journal of the American Medical Association, 302(5), 550–561.
Petrie, H. G. (1979). Metaphor and learning. In A. Ortony
(Ed.), Metaphor and thought. New York: Cambridge University Press.
Richards, I. A. (1936). The philosophy of rhetoric. London:
Oxford University Press.
Rosse, R. B., Deutsch, L. H., & Deutsch, S. I. (2000). Medical assessment and laboratory testing in psychiatry. In
B. J. Sadock & V. A. Sadock (Eds.), Kaplan and Sadock’s
comprehensive textbook of psychiatry (7th ed.). Philadelphia: Lippincott/Williams & Wilkins.
Scheflin, A. W., & Spiegel, D. (1998). From courtroom
to couch: Working with repressed memory and avoiding lawsuits. Psychiatric Clinics of North America, 21(4),
847–867.
Schwartz, M. S. (1995). Intake decisions and preparations
of patients for therapy. In M. S. Schwartz & Associates,
Biofeedback: A practitioner’s guide (2nd ed.). New York:
Guilford Press.
Shaw, E. R., & Blanchard, E. B. (1983). The effects of
instructional set on the outcome of a stress management
program. Biofeedback and Self-Regulation, 8(4), 555–565.
Siegelman, E. Y. (1990). Metaphor and meaning in psychotherapy. New York: Guilford Press.
Sommers-Flanagan, J., & Sommers-Flanagan, R. (2008).
Clinical interviewing (4th ed.). New York: Wiley.
Stott, D., Mansell, W., Salkoskis, P., Lavender, A., & Cartwright-Hatton, S. (2010). Oxford guide to metaphors in
CBT. Oxford, UK: Oxford University Press.
Taylor, R. (2007). Psychological masquerade: Distinguishing
psychological from organic disorder (3rd ed.). New York:
Springer.
Turk, D. C., & Melzack, R. (Eds.). (2010). Handbook of pain
assessment (3rd ed.). New York: Guilford Press.
Yucha, C., & Montgomery, D. (2008). Evidence-based practice in biofeedback and neurofeedback. Wheat Ridge, CO:
Association for Applied Psychophysiology and Biofeedback.
Chapter 15
Adherence
Jeanetta C. Rains and Mark S. Schwartz
“Adherence” generally refers to the concordance
between a patient’s behavior and a recommended
treatment plan, and is increasingly recognized
as moderator of outcome with all forms of treatment. Certainly, patients must cooperate with
recommendations if the recommendations are to
be effective. In most cases, treatments are validated with specific protocols that maximize effectiveness and minimize complications. Failure to
adhere to the recommended protocol is likely to
yield an inferior outcome. Evidence from a wealth
of medical and psychological literature confirms
that nonadherence is almost invariably associated
with poorer outcomes across a wide range of disorders and therapies. The World Health Organization (2003) proclaimed that improving patient
adherence to existing treatments would have a
far greater impact on the health of the population than development of new treatments. Thus,
nonadherence acts as a ceiling on effectiveness of
all treatments and warrants specific attention in
administering and evaluating any form of treatment. In biofeedback therapies, nonadherence
may take the form of failure to follow through
with initial referral, missed appointments, failure
to complete homework or enact lifestyle changes
related to treatment, and premature termination
or dropout.
Nonadherence wastes health care resource
(e.g., time, money), portends a poorer therapeutic
prognosis for patients, and may be demoralizing
for practitioners. It seems counterintuitive that
patients with significant and often distressing
symptoms would not avail themselves of available
treatment. Recent attention has been devoted to
development of health behavior models that help
explain the paradox of nonadherence. Some of the
more widely accepted and compelling models take
into account psychological factors such as patient’s
self-efficacy and readiness for change, as well as
external barriers. Rather than merely carrying
out “doctor’s orders,” patients are believed to follow recommendations based on their own implicit
cost–benefit analyses in which the necessity of
treatment is weighed against concerns about the
perceived negative effects or costs. Health behavior theory helps account for the myriad biopsychosocial determinants of adherence and provides a
guiding framework for strategies to facilitate adherence. Appreciation of such concepts may assist in
optimally tailoring interventions to patient needs
through educational, motivational, and behavioral
adherence-enhancing strategies.
We discuss briefly in this chapter the scope and
impact of nonadherence and factors that affect
adherence, and we summarize ideas and conclusions from the extensive literature on adherence
with medical and psychological treatments. We
also presents considerations about conducting a
professional practice. This chapter’s guidelines
and considerations apply to most settings in which
practitioners provide biofeedback.
233
234
Scope of Nonadherence
Across chronic conditions such as asthma, arthritis, headache, hypertension, diabetes mellitus,
and so forth adherence to pharmacological and
psychological therapies is known to be poor. A
review of the medical literature indicated that
nonadherence impacts virtually every aspect of
health care (Dunbar-Jacob et al., 2000; LaGreca,
Bearman, & Roberts, 2003). Conservatively, 30%
of prescriptions are never filled, and only half of
prescribed agents are taken sufficiently to achieve
a therapeutic effect. Not surprisingly, compliance
is poorer for chronic than for acute conditions,
and 50 to 60% of persons with chronic conditions
are nonadherent with prescribed medication regimens.
Failure rates for appointment keeping range
from 8 to 63%. Up to 50% of patients on chronic
medical regimens drop out of care entirely within
1 year of beginning treatment. Compliance with
lifestyle recommendations such as dietary modifications, weight loss, exercise, and smoking cessation occurs less often than adherence to prescribed
medication regimens.
Nonadherence with Psychological Interventions
Though fewer studies have examined adherence
to psychological therapies, evidence confirms that
adherence to psychological treatment is at least
as poor as adherence to pharmacological treatment. A review of three decades of psychotherapy
research found that 30–60% of patients drop out
of treatment (Reis & Brown, 1999). The modal
number of psychotherapy sessions is one single session, and 20–57% of patients do not return after
the initial session. The early phase of treatment
appears crucial for continuation insofar as dropout rates tend to decline as the number of sessions
increases.
There is a dose–response relationship between
number of sessions and clinical improvement
with psychological interventions (Lambert, Hansen, & Finch, 2001) and preliminary evidence for
such a relationship in biofeedback. Reiner (2008)
observed a dose–response effect between compliance with daily practice with respiratory sinus
arrhythmia (RSA) biofeedback and improvement
in anxiety, anger, and sleep measures among 20
patients completing 3–4 weeks of treatment for
a psychophysiological disorder (e.g., anxiety, irritable bowel syndrome, insomnia). Homework
V. PRACTICE ISSUES
included daily practice with a portable, handheld
RSA biofeedback device. Patients were instructed
to practice periodically throughout the day using
the device, which awarded points for the desired
response. The goal involved achieving the desired
response for 20 minutes total per day, or 100 points.
A significant positive correlation was observed
for adherence (number of points per week) with
measures of sleep quality, trait anxiety, and trait
anger (p < .01). Likewise, Byrne, Solomon, Young,
Rex, and Merlino (2007) found that completion of
the full course of treatment (sixsessions) predicted
improvement with biofeedback for fecal incontinence in a large sample of 513 patients, although
authors did not analyze dose–response effects;
compared to those who did not complete treatment, completers exhibited significant improvements in incontinence scores, objective measures
of sphincter function, and quality of life.
Patients’ reasons for discontinuing treatment
have been examined. In a study of 233 patients
referred for cognitive-behavioral therapy (Bados,
Balaguer, & Saldana, 2007), half of the patients
(50.7%) completed the recommended number of
sessions. The majority of dropouts occurred in
the earlier sessions—a cumulative total 40.4%
of patients dropped out after the first or second
session (28.1% after the first session). Patients’
reported reasons for dropping out were consistent
with earlier research. The most common reasons
for leaving therapy were dissatisfaction with therapist or treatment and/or low motivation (46.7%);
external factors such as transportation problems,
timetables, and competing responsibilities (40%);
and the belief that symptoms had improved
(13.3%). Thus, establishing a positive therapeutic
relationship with patients and enhancing motivation within the initial sessions of treatment (especially the first two sessions) appears essential for
continuation and success of treatment.
Predictors of Nonadherence
Barriers to adherence may be environmental, psychological, interpersonal, and financial. Predictors of dropout from a wide range of psychological interventions have been reviewed elsewhere
(DiMatteo, Lepper, & Croghan, 2000; DunbarJacob & Mortimer-Stephens, 2001; Dunbar-Jacob
et al., 2000; Lambert et al., 2001; O’Donohue &
Levensky, 2006; Reis & Brown, 1999) and presumably generalize to biofeedback. Delays in
health care delivery, such as time spent waiting
235
15. Adherence
for consultation, tend to decrease engagement in
treatment and increase dropout, whereas reminding and re-calling patients have been shown to
increase retention. Patient demographic characteristics do not consistently predict dropout, but
lower education and income, poorer interpersonal
relationships, and lack of social support appear to
increase dropout. Experienced therapists appear to
have lower rates of dropout. Patient dissatisfaction
with therapist/therapy and poor patient–therapist communication appears to increase dropout.
Comorbid psychopathology, especially depression,
is one of the most important predictors of dropout
and nonadherence in the medical literature (e.g.,
chronic pain, asthma, diabetes). Substance abuse
is associated with poorer adherence to both psychological and medical treatment. “Self-efficacy,”
or the degree to which a patient believes his or her
actions can achieve the desired outcome, has been
shown to predict treatment adherence in a wide
variety of chronic conditions and is believed to be
one of the most potent psychological determinants
of adherence identified to date.
Model for Health Behavior Change
Grounded in the social learning theory of Albert
Bandura (1977, 1986), adherence with recommendations for health behavior change is believed to
involve three essential constructs: patient’s readiness for change, perceived importance of change,
and confidence in one’s ability to change (self-efficacy) (Elder, Ayala, & Harris, 1999; Kinzie, 2005;
Miller & Rollick, 2002). Interventions addressing
self-efficacy and readiness for change have proven
effective in improving adherence.
Self‑Efficacy
Access to treatment, and skills to carry out such
treatment, are not sufficient to ensure that such
skills will be performed. Rather, patients must have
confidence in their ability to perform the behavior, as well as the expectation that performance of
that behavior will result in the desirable outcome.
Here, self-efficacy represents the degree to which a
person believes he or she can carry out the desired
behavior (e.g., attending biofeedback session, completing homework) and that acquiring biofeedback
skills will in fact relieve presenting symptoms (e.g.,
pain, incontinence, anxiety). Patients who doubt
the efficacy of treatment or lack confidence in
their ability to carry out treatments requirements
are less likely to comply.
Self-efficacy can be augmented through strategies that draw on past successful experiences and
build new skills through modeling and reinforcement of successively more complex adherence
behaviors. Four methods are encouraged to augment self-efficacy (Prochaska & Lorig, 2001), generally beginning with verbal persuasion in describing the expected benefits of the treatment, as well
as encouraging patients to perform the desired
behavior. Biofeedback skills simulated by therapist
or model (vicarious experience) or delivered to the
patient in small, manageable steps of guaranteed
success (performance accomplishment) build selfefficacy. By succeeding with smaller immediate
tasks, patients develop a sense of competence to
continue and persist through challenges. Finally,
teaching patients to expect, identify, and interpret
physiological states as part of the change process
helps them to persist through symptom fluctuations and exacerbations. Providing a realistic
expectation of time and course of change using
biofeedback may help patients anticipate plateaus
and setbacks as a necessary part of the process
and may facilitate perseverance through adverse
experiences. Examples of strategies to enhance
self-efficacy are embedded in subsequent recommendations.
Readiness for Change
Patients’ motivations for treatment are dynamic
and vary over time. A useful framework for understanding shifting motivations in the behavior
change process is the transtheoretical model (Prochaska, Redding, & Evers, 1997), which purports
that patients move along a continuum of motivational readiness to change until the behavior ultimately becomes a habit. Change does not necessarily progress in a linear fashion; rather, patients
cycle through five stages of change: precontemplation (not thinking about changing behavior),
contemplation (actively thinking about changing behavior but not trying to change), preparation (beginning to make changes slowly), action
(actively engaged in regular behavior change), and
maintenance (maintaining changes). In the maintenance stage, behaviors are performed habitually
and require little thought. Relapse is movement
from one stage back to a prior stage. Although no
published studies to date have been reported with
biofeedback, low contemplative ratings on the
236
Stages of Change Scale were associated with premature termination from psychotherapy (Derisley
& Reynolds, 2000), and interventions directed at
each of these processes have proven effective in
improving adherence.
Providers may recognize patients’ stage of
readiness for change and tailor their educational,
motivational, and behavioral strategies accordingly. For example, patients referred for biofeedback may be in the stage of precontemplation
and have no consideration of altering their own
behaviors. Here, education linking patients own
behaviors with symptoms is needed, and any
efforts at introducing biofeedback skills acquisition are likely to be lost in this stage. Patients in
the contemplation stage of change warrant education about options and also may benefit from
social support and encouragement. Here biofeedback can be introduced with education describing
the manner in which learned control over physiological responses can be gained through training.
Those in preparation for change stage warrant
clear, step-by-step behavioral instruction, modeling, discussion of barriers, and close follow-up.
The action stage while carrying out biofeedback
warrants the closest assessment of appointment
and homework adherence. Reinforcement is useful in maintaining the behavior through barriers
until the benefits of treatment (symptom relief or
functional improvement) can emerge to maintain
the behavior. Maintenance of treatment gains
includes continued enhancement of self-efficacy
and relapse prevention.
Adherence Facilitation
The ability to engage and motivate patients
enhances the effectiveness of health care professionals, regardless of the treatment’s efficacy and
the professional’s knowledge or good intentions.
Patients do not comply for a variety of reasons,
and it is useful to attempt to understand their
unique cost–benefit perception for requirements of
treatment and tailor efforts to address misunderstandings, competing motivations, and barriers.
The following recommendations are based on evidence primarily gleaned from parallel literatures
and clinical experience. Strategies for adherence
facilitation generally involve preparatory techniques and motivational enhancement. Recommendations are divided into three major categories addressing (1) the professional, (2) the patient,
and (3) evaluation and intervention.
V. PRACTICE ISSUES
The Professional
Professional Setting, and Nontherapy
and Therapy Personnel
Personnel who are friendly, efficient, and professional in their appearance and behaviors affect
patients’ impressions, comfort, confidence, satisfaction, and compliance. Comfortable, neat, and
uncluttered office rooms, with comfortable temperatures, also probably help. Short waiting times
and consistency of therapists promote adherence.
Likewise, reminding and re-calling patients should
be a integrated in clinic administration, since failure to keep appointments is significant in psychological interventions and acts as a ceiling on all
future treatment and adherence efforts. With up
to one-half of patients failing to return for treatment after consultation (Reis & Brown, 1999),
this would appear to be the single greatest opportunity to improve adherence. Calling patients to
remind them of appointments and re-calling those
who miss a scheduled appointment are fundamentally the most cost-effective, adherence-enhancing
strategies. Although not studied in biofeedback,
additional simple strategies have been shown
across other chronic disorders to improve appointment keeping and include reminders (mail, telephone), clinic orientations, and contracting with
patients (Macharia, Leon, Rowe, Stephenson, &
Haynes, 1992).
The Referral Source’s Attitudes and Behaviors
A well-informed referral source can be an ally for
treatment. Practitioners should consider discussing with referral sources their viewpoints about
biofeedback and related therapies. This helps a
practitioner know how much attention to direct
toward building a patient’s confidence in the
therapy rationale, recommendations, and procedures. Formal feedback to referral sources is likely
to increase continuity of care and reinforce the
referral process. Therapists should consider sending educational reading materials and including
useful, informative, and sensible content in letters
to referral sources. Well-written letters to referral
sources, and well-written notes in patients’ records,
influence referral sources’ attitudes and behaviors.
One can avoid both writing a long report or letter and repeating what the reader already knows
about the patient. A sample letter to a referral
source (Figure 15.1) may be supplemented with
data and graphs. Another option is to send a copy
of the intake evaluation and therapy session notes.
237
15. Adherence
Dear Dr. ,
Thank you for referring Ms. for treatment of her tension headaches.
I will not repeat her history, of which you are aware. My evaluation of her head and neck
muscle activity used four sites: bifrontal, bilateral posterior neck, and right and left frontal
posterior neck. Baseline measures were taken with her eyes open and closed while she was
sitting with her back and head supported, and while she was standing. The assessment
involved rest and mild office stressors. The procedures then included feedback to measure her
response and begin to develop her physiological self-regulation.
Muscle activity from the neck while she was sitting with her eyes open showed excess
muscle activity, mostly 4 to 6 microvolts (100- to 200-hertz bandpass). While standing
and trying to relax, she showed higher excessive muscle activity, mostly 9 to 12 microvolts.
Muscle activity from the other sites was only slightly tense for resting muscles while she was
sitting, and only slightly higher while she was standing. Visual feedback helped her reduce
muscle activity, especially while standing. Without the feedback, muscle activity increased and
remained elevated.
I discussed the rationale for therapy and the procedures. Evaluation of psychosocial
factors did not suggest enough to warrant other forms of stress management, and she was not
receptive to other forms of stress management. I provided audiotapes and patient education
booklets to her.
She was seen for five more sessions. She is working on weaving physiological selfregulation into her daily activities. She continues to increase her ability to lower muscle activity
during resting conditions, and to maintain lower muscle activity after feedback stops. In the
last two sessions, her muscle activity during some phases was in a therapeutic range below 2
microvolts. Her symptom log for the last 4 weeks shows a 75% reduction in severe headaches,
a 50% reduction in total hours of headaches, and an 80% decrease in medication use
compared to her initial reports.
Thank you for referring this pleasant lady. I am happy to be of help to her. Contact me with
any questions.
Sincerely yours,
FIGURE 15.1. Sample letter for referral source.
The Practitioner’s Characteristics and Behaviors
Patient satisfaction with the therapist/treatment
and the therapist experience have been associated
with improved adherence. Credibility of practitioners is enhanced by professional presentation,
as well as the amount and quality of time spent
with patients. Our appearance, behaviors, and
personality all affect credibility and trustworthiness. Practitioners increase and maintain patients’
trust in them by being on time and maintaining
confidentiality and consistency. They realistically
discuss therapy goals and expectations. They also
discuss expected and possible changes in therapy
well in advance.
Effective communication and the appearance of
a relaxed therapist, or one who effectively shows
relaxation when needed, conveys an important
teaching model. The therapist might consider
self-disclosure about how he or she has used physiological self-regulation to prevent and manage
symptoms. Communication should be logical and
coherent in order to be understood, accepted, and
remembered. Effective communication must also
be within the patient’s “latitude of acceptance.” A
provider with a rigid conceptual framework and
a rigid approach to therapy may not evoke compliance from patients whose attitudes are outside
the limits of the information and therapy plan.
Patients’ attitudes must be taken into account
when therapists are framing explanations about
the causes of their symptoms, as well as descriptions of the therapy rationale, procedures, and recommendations. Both the wording and the content
of these explanations and descriptions must be
considered. For example, therapists must consider
how such phrases as “letting go” and a “nonstriving
238
attitude” sounds to many patients. Similarly, some
recommendations (e.g., changing work schedules)
are outside the acceptance range for some patients.
A collaborative relationship and enlisting
patients active participation in setting goals is
likely to result in a more positive interaction and
realistic treatment plan. The therapeutic relationship is the cornerstone for effective therapy. Table
15.1 outlines recommendations for interaction
between practitioner and patient. The professional
should clearly explain the role of each factor and
develop a flexible stepped-care approach. Flexibility, laced with clear communication and empathy,
is often more likely to lead to acceptance than
rigid insistence. Consider the following sample
explanation to a patient:
“Mrs. , we talked
about the potential advantages of stopping caffeine, managing your time use more effectively,
and using relaxation therapies. Caffeine interTABLE 15.1. Recommendations for Interaction
between the Practitioner and Patient
1. Attend to your personal characteristics and
behaviors.
2. Spend enough time with the patient.
3. Provide an active interaction.
4. Acknowledge the legitimacy of the patient’s
complaints.
5. Present an organized, systematic, and flexible
approach.
6. Include appropriate, but limited, social
conversation.
7. Provide reassurance, support, and encouragement
8. Provide and reinforce realistic positive
expectations.
9. Provide choices for the patient.
10. Allow the patient to question recommendations.
11. Tailor therapy whenever indicated and practical.
12. Demonstrate and model selected procedures.
13. Provide appropriate self-disclosure.
14. Show attention and interest in the patient through
tone of voice, facial expression, and physical
posture.
15. Convey appropriate affect.
16. Maintain frequent eye contact.
17. Touch the patient appropriately.
18. Observe for signs of anxiety, resistance, and
confusion.
V. PRACTICE ISSUES
feres with effective relaxation. More effective
use of time will help you make time for enough
relaxation to help you reduce anxiety and tension in your life. I know you might not feel ready
for some of these changes. I am not saying that
all of them are completely necessary for you to
reduce or stop your symptoms. However, some
are necessary. Seriously consider these recommendations. The decision about what to do,
both now and later, is up to you. You can start
with for a trial, and then see
how far you get.”
Tailoring Therapy: Providing Choices for the Patient
Although there may be a place for predesigned
therapy programs, tailoring therapy to patients’
individual situations and preferences is almost
invariably needed. Tailoring also includes consideration of the patient’s attitudes, schedule, and
finances. In addition, the therapist considers the
symptoms during the first days and weeks of therapy and the first few biofeedback sessions. Having
choices, and knowing the potential advantages
and disadvantages of each, gives a patient the
sense of actively participating in the design of his
or her program. Such an approach also conveys
that the practitioner is considering the patient’s
situation, preferences, and needs. Consider these
factors:
1. Symptoms sometimes improve after a physician
reassures a patient about the nonserious nature
of his or her symptoms.
2. Starting a new medication or changing dosage
can improve symptoms before physiological
effects of the change may be expected.
3. Making changes in lifestyle, work, exercises,
and/or diet can result in a significant decrease
in symptoms. (Note. In these first three examples, compliance with a time-consuming treatment will often be less than ideal, especially if
the patient believes that the symptom changes
result from these factors. In such circumstances, tailoring the therapy plan can involve
deferring more biofeedback.)
4. Another factor that alerts practitioners to the
need for tailoring is a lack of physiological tension and arousal, and/or a rapid return to a
“therapeutic range”1 after intentional arousal
during office sessions. This suggests that the
patient can relax but needs to apply his or her
ability to do so.
239
15. Adherence
5. Some patients show significant reduction
of symptoms in the first weeks of therapy,
regardless of the physiological self-regulation
shown in the office. It is sometimes difficult to
explain rapid improvement of symptoms. The
scientist in each of us wants to know why and
how this is occurring; the skeptical and cautious part of us is suspicious. However, the clinician in us, especially the cost-conscious and
pragmatic part, accepts the progress and may
defer more office therapy. If the symptoms
increase later, patients may better accept the
role of the therapy and the need for compliance.
6. Therapists should consider deferring or stopping office-based biofeedback sessions with
patients who are nonadherent with necessary
parts of therapy so as not to associate the failure experience with therapy and inoculate the
patient against future treatments.
7. Assessment and feedback sites may need tailoring. For example, a patient may perceive
repeated sessions semireclined, with eyes
closed, and with feedback from only one area
as meaningless. Tailoring the sites, body positions, and conditions can make the sessions
more sensible to the patient and increase the
patient’s confidence and compliance.
8. Physiological arousal in response to cognitive
stressors and activities offers data with which
to tailor therapy. For example, let us suppose
that a patient shows good relaxation in multiple muscle areas, warm hands, and low skin
conductance during resting baselines. Furthermore, suppose that the patient can do
this during a standard cognitive stressor and
in different positions. Now let us suppose that
this patient has significantly decreased finger
temperatures while imagining or talking about
work or family stress. The therapist may consider repeating the arousal scenes several times,
with and without feedback, and exploring the
content. Or the therapist may encourage relaxation before, during, and immediately after
work situations.
9. Long geographic distance between the patient
and the treatment office require tailoring of
the therapy. Office sessions twice a day for 2
or more consecutive days, with a few weeks
between such phases, may be preferable for
some patients who live too far away for weekly
sessions. Such a “massed-practice” schedule
conveys that the therapist is willing to extend
him- or herself, and may help increase patient
compliance.
Providing Appropriate Self‑Disclosure
Limited, appropriate self-disclosure (Table 15.1,
item 13) includes brief and proper descriptions of
how practitioners effectively use applied psychophysiology in their own lives. In this manner, therapists can communicate that they know firsthand
that this works and potentially enhance patients’
self-efficacy through vicarious experience.
Professional Verbal and Nonverbal Behavior
The practitioner’s voice, facial expressions, body
posture, and nonverbal behaviors all convey interest, trust, sincerity, experience, and confidence
(Table 15.1, items 14 and 17). Likewise, touch is
important. Psychologists and others providing biofeedback do not typically use physical contact in
their other contacts with patients. Even those with
experience may not know how to use touch to convey sincerity, support, and encouragement. There
are several obvious chances to use touch properly,
aside from the initial handshake. These include
times when the therapist is attaching electrodes
and other transducers to a patient, and when he
or she is directly helping in relaxation or muscle
reeducation. For example, the therapist might
consider how he or she moves a patient’s hair and
grasps the patient’s arm. It can be enlightening
for the practitioner to ask him- or herself, “How
would this contact feel if I were the patient?” The
practitioner can gently but firmly place a hand
on the patient’s forearm, and may give a mild
squeeze or brief pat on the arm, but for no longer than about 2 seconds. A therapist must always
be sensitive about contact; some patients may not
like or may distrust such contact. However, if not
overdone in intensity or frequency, the technique
can have positive effects. Other chances for reassuring with touch include occasions when patients
express frustration, fear, life stress, and difficulty
with physiological self-regulation. Notably, touch
should not be overdone, insofar as touch can have
a negative effect. For example, consider the possible impressions of female patients touched by male
practitioners. I am not suggesting that one should
avoid such contacts altogether. However, the duration and frequency of the contacts can convey the
wrong message. Consider the difference between a
possible undesirable message from a touch of about
240
5 seconds, and a desirable one of about 2 seconds.
If the practitioner’s hands are cold, moist, or both,
he or she should consider touching clothed parts of
the arm and not bare skin.
Cognitive Preparation of the Patient
Cognitive preparation includes (1) the rationale
for physiological self-regulation, (2) therapy process, (3) therapy goals, (4) use of medications, (5)
generalization, (6) therapy options, (7) stepped
care, and (8) a symptom log. It may be helpful to
tailor this information to patient’s apparent stage
of readiness for change. A therapist should try to
avoid overloading a patient during one session.
Some information may need presentation during the first few sessions, then be repeated later
for emphasis and to help patients recall. Carefully
developed patient education scripts, booklets,
checklists, and audiotapes help. Because patients
often forget much of what they hear, therapists should consider repeating key points. Most
patients will have at least some compliance problems. Shaping acceptance of self-responsibility is
one of the challenging aspects of clinical practice.
Cognitive preparation also includes expecting
slow progress, plateaus, and setbacks. Practitioners
want to communicate realistic positive expectations. It can be useful to show patients graphs of
developing physiological self-regulation and symptom changes from prior patients. In addition, it
can be helpful to tell a patient something like the
following:
“You have a good chance of making progress in
reducing or stopping your symptoms. I do not
know how long this will take. Some patients
show much improvement within days or weeks.
Others progress gradually over several weeks
or a few months. Plateaus and even temporary
reversals happen. If they occur, remember that
they are normal and a natural part of learning. You need not feel discouraged. You know
athletes and musicians expect unevenness in
developing and keeping their skills. Also, keep
in mind that even accomplished athletes have
off days. Not even a great baseball player, golfer,
or tennis player always hits the ball well.”
Patients often display resistance, skepticism, and
pessimism. Often they have gone to several doctors
and have tried various treatments. They have had
positive expectations but have been disappointed
with the results. For some patients, our treatment
V. PRACTICE ISSUES
approach appears too simplistic, despite careful
explanations and professional reputations. If their
healthy skepticism leads to our being defensive
or rejecting of them, then we unnecessarily risk
dropout. Checking and discussing patient’s perceptions may be needed to explain the rationale
for a therapeutic trial, discuss options, and gently
and empathetically overcome patient resistance.
It is too easy to give up and label such patients
as unsuitable for therapy. For such patients, the
greater challenge to their therapists is to establish
a therapeutic alliance and mobilize these patients
realistically. In addition, the challenge is to shape
their attitudes and perceptions of the therapy.
Thus, therapists shape patients’ self-confidence,
their optimism, and their willingness to engage in
a realistic therapy trial. A therapist should consider
an empathetic and nondefensive response:
“If I put myself in your place, I would be skeptical too. I know you went to several doctors
and tried several treatments without success. I
understand that you were hopeful and then disappointed. Part of you is asking yourself, ‘Why is
this going to be any different?’ You do not want
to get your hopes up, because you do not want
that disappointment again. I understand that,
and I think it is perfectly normal. You know, I
cannot promise that you will improve or tell you
how much improvement you will have. However, I can tell you that many patients treated
did well, despite being unsuccessful with past
therapies. Thousands of professionals all over
the country report the same experiences with
their patients. There are several approaches we
can take. If the first is less than ideal, then there
are variations and other approaches. Biofeedback and related therapies involve many therapies. Ask me any questions you wish, or express
any concerns and doubts you may have. Some
parts of the therapy may appear to you as too
simple to work. I do not want them to appear
complicated. They are not as simple as they
appear, but neither are they very complicated.
Even long-term symptoms often do not require
complex solutions. For example, I treated many
patients with [this patient’s symptoms] for years.
[Practitioners can also insert their personal
experiences.] These treatments helped most of
them. Many of those patients thought these
treatments were probably not enough for them,
yet they got better. Relaxation and biofeedback
treatments are often enough alone. However,
sometimes we also need other therapies.”
241
15. Adherence
The Patient
A patient’s perceptions, expectations, and mood
impact therapy. The list below provides a catalogue of many specific patient perceptions, expectations, affective and symptom-related factors, and
other factors. All of these can affect compliance
and therapy results. This list should be considered during evaluation, cognitive preparation, and
treatment.
Perceptions of Biofeedback/Relaxation
Patients May Perceive Biofeedback/
Relaxation as Psychological Treatment
Because mental health professionals commonly
provide biofeedback-assisted relaxation, it is understandable that many patients perceive biofeedback
as a psychological approach. Many health care
professionals also consider biofeedback a psychological technique. However, these therapies are
multidisciplinary and not uniquely or exclusively
within the province of psychology. When patients
reject or fear stigma of “psychological” therapies,
the therapist should explain that many professionals do not view biofeedback and relaxation
as psychological treatments. Such a discussion
is pertinent when treating patients with medical
disorders such as migraine. One may not need to
discuss the multidisciplinary nature of biofeedback
with patients who voluntarily seek mental health
help. Furthermore, nurses, physical therapists,
occupational therapists, and other non-mentalhealth professionals providing or supervising these
therapies need not have this type of discussion
with patients. A mental health professional might
consider portraying physiological self-regulation
and biofeedback as having many unique features,
in addition to any psychological aspects.
Patients May Perceive Biofeedback/
Relaxation as Insufficient, Useless,
or a Waste of Time
Prior unsuccessful therapies or exposure to diluted
forms of therapy (e.g., self-administered biofeedback tools, relaxation tapes) can diminish the
credibility of self-regulation therapies. Patients
may harbor these perceptions yet not openly
express them. A corollary of the insufficiency/
uselessness perception is that relaxation is simply a
waste of time. Such a perception is especially common among medical patients who have chronic
physical symptoms. A therapist should consider
showing awareness of these perceptions, for example, by saying:
“You may be thinking, ‘How is this therapy going
to help me? I have had these symptoms so long
and tried so many treatments. I got my hopes up
before and the treatments did not help. How is
this going to be different? I have a busy schedule. I am having trouble believing this is not a
waste of time.’ If you are having thoughts like
this, let’s discuss them.”
The therapist can then explain the rationale for
therapy and discuss how it can help the patient,
despite the chronicity of his or her symptoms and
any prior unsuccessful treatments. Time spent
in the present treatment can be described as an
investment.
Compliance with homework assignments is
often crucial within many therapies, including
applied psychophysiology. Reviews are available
elsewhere (Kazantzis, 2000; Scheel, Hanson, &
Razzhavaikina, 2004). Most studies support the
role of homework adherence and improvement in
therapy, including the dose–response relationship
observed between RSA biofeedback homework
and outcome describer earlier (Reiner, 2008),
whereas other studies have not found such support. Generally, homework in biofeedback facilitates generalization of skills, and many therapists
consider completion of homework to be a barometer of the patient’s motivation and commitment
to therapy. Homework assignments should have a
dedicated purpose and not be burdensome to the
patient. When homework assignments are made,
the documents should be reviewed in session and
patients reinforced for adherence. Homework is
discussed in a later section of this chapter.
Patients May Perceive Biofeedback/
Relaxation as the Last Hope
Patients who have tried several therapies without
success may perceive relaxation and biofeedback
as therapies of “last resort.” This perception is
especially common among patients seen in tertiary medical centers and considered refractory.
As with other misperceptions, patients are usually
hesitant to report this perception spontaneously.
Such a perception is an added source of anxiety.
It can interfere with patients’ attending to what
professionals present and complying with recommendations. The increased tension and symptoms
that sometimes result can lead to more frustra-
242
tion and discouragement. Some patients then
consider giving up entirely and dropping out of
therapy. Providers should emphasize that different
therapies and approaches are available. There are
lifestyle changes, dietary changes, cognitive stress
management therapies, other stress management
approaches, and various combinations of these.
This information can significantly alter the perception of biofeedback as the therapy of last resort.
Patients May Perceive Therapy
as Preprogrammed and May Resist
Such a Program
Practitioners sometimes provide biofeedback in
preprogrammed packages, with a standard number
of sessions, placements of transducers, body positions, and treatment conditions. There are often
specific physiological criteria for proceeding to the
next stage or changing to a different strategy. Some
patients (and practitioners) resist preprogrammed
packages and prefer tailored therapy. Therapists
need to be aware of these potential perceptions
and adjust their treatments accordingly.
Patients May Perceive That Other
Therapies Are Needed, Based on Prior
Medical Consultations
Physicians with whom a patient has consulted
have probably discussed and recommended therapy options. These include new or different medications, dosage changes, surgery, or psychotherapy.
Before pursuing these other therapies, the patient
has come to a provider who uses biofeedback and
related therapies. However, the patient often
perceives the other options as viable and possibly effective. The biofeedback therapist should
uncover these perceptions and deal with them as
soon as it is practical to do so.
Patients May Perceive Aspects
of Biofeedback/Relaxation
as Silly or Embarrassing
Patients may fail to understand certain aspects
of relaxation, such as tensing and releasing facial
muscles, diaphragmatic breathing, listening to
audiotaped relaxation, and relaxing in public
places. Patients rarely volunteer such perceptions
and feelings. The next question is how to revise
them. Some practitioners and therapists model
some aspects of biofeedback and relaxation. A
therapist’s self-disclosure that he or she uses these
V. PRACTICE ISSUES
procedures in daily life can be reassuring. Some
patients also may feel more comfortable once they
know that many professionals, executives, athletes, entertainers, and others also often use these
procedures. Providing a credible rationale for, and
explaining the application of, each procedure can
help to put it in better perspective as well.
Patients May Perceive the Therapy as Too
Costly, Impractical, or Time Consuming
Some patients’ perceptions that the costs and
duration of treatment are beyond their capabilities
may undermine adherence. Solving this problem
is not easy, but a stepped-care model would adjust
the number of office sessions. The therapeutic
relationship can help the patient accept the costs
and duration of therapy. A practitioner can support reduced fees when standard fees are a hardship for a patient, when there is a clear need for
therapy, and when fee reduction is allowable. Such
humane and generous adjustments result in appreciation and can increase compliance. However, fee
adjustments have become more difficult in recent
years because of reimbursement problems throughout the health care system.
Patients May Perceive
the Therapy Procedures, Logs,
and Practice as Burdensome
Patients must relate treatment demands and procedures to symptom reduction. Otherwise, they
may perceive treatment as being too complex or
irrelevant for them. Self-monitoring and practice should be tailored to patients’ symptoms
and lifestyle. Impractical and complex logs can
lead to a lack of records, contrived data, or withdrawal from therapy. Biofeedback/relaxation time
commitments can take time away from patients’
other activities. Taking an hour or more daily for
applying recommendations—which therapists
often suggest—is more than some patients can
do. Of course, when patients’ schedules are very
full, there is often a greater need for balance and
therapy. However, patients may not perceive their
situations in this way. Examples of people whose
schedules do not permit time for ideally applying treatments are farmers, tax accountants, and
other seasonal workers whose schedules change
significantly. In such cases, therapists need to be
flexible and wait rather than “beating their [professional] heads against a stone wall.” Shorter daily
relaxation sessions, and/or instructions for blend-
243
15. Adherence
ing therapy activities into daily activities, should
be considered. Another strategy would be to defer
some treatment and focus therapy on altering the
patient’s schedule and priorities—in other words,
to practice “time use therapy.” For example, some
people need to delegate, limit, or learn to do some
tasks with less perfection.
Patients May Mistrust Health
Care Professionals
Some patients have learned to mistrust health
care providers. Such distrust often stems from
negative personal or vicarious experiences. Many
such patients have experienced mistreatment,
have been misled, or have received treatment
from insensitive providers. Professionals who have
added to the negative perceptions of patients may
have provided inadequate relaxation, biofeedback, and associated therapies. For example, many
patients react negatively to being alone for most
or all sessions. These patients report feeling abandoned, anxious, confused by what to do, and frustrated. These feelings have added to their negative perception of biofeedback and of professionals
offering such services. Professionals can modulate
such experiences by the quality of interactions
with their patients. They should inquire about
prior experiences if the patients do not volunteer
the information. Some patients candidly describe
their negative experiences. Others either sit quietly with distrustful looks on their faces or provide
no obvious clues of their distrust. Still others simply do not want to be in therapists’ offices or follow
therapists’ advice—no matter who the therapists
are, what they say, and how pleasant they are.
Patients May Not Perceive Their
Practitioners as Allies
Even patients without negative experiences may
not perceive practitioners as allies in the battle
against symptoms. We may be credible and even
highly competent. However, patients may see us as
too formal, too distant, not devoted enough, or too
busy to provide the time they perceive they need.
We need to get out from behind our desks and
our formality, and convey that we do care about
patients as people, not just as patients or cases.
It is difficult for some professionals to adopt Will
Rogers’s dictum, “I never met a man [or woman] I
didn’t like.” However, it helps to keep striving to
show our liking for all patients . . . perhaps with a
few exceptions!
Patients May Perceive Their Symptoms
as Organic and Requiring Solely
Biological Therapy
A patient’s perception may be that “if only my doctors believed me and did more tests, they would
find the cause of my problems.” Such patients
continue to believe that an organic cause is the
major factor explaining their symptoms. They may
hold this perception despite the fact that highly
expert medical examinations and laboratory tests
have ruled out an organic explanation for the
symptoms. Patients often limit compliance in such
cases. Adherence necessitates that patients accept
the rationale and believe that stress, tension, or
arousal may cause or worsen their symptoms. They
need to believe that organic factors are minor or
nonexistent, or that they can be overcome. Superficial compliance with the mechanics of therapy
may still occur; however, these patients only “go
through the motions.” They often wait for another
chance to get more medical tests. They may even
view such compliance as a chance to show that
the symptoms indeed are organic. They do this
by failing to improve. Such a situation is typically
very difficult to manage. Such patients often go to
highly credible tertiary medical centers. Resulting
examinations and tests sometimes do find organic
disease explanations for some patients’ symptoms.
Ruling out organic causes by the best credible
medical examinations can help patients accept
physiological self-regulation and related therapies.
It is often wiser to defer applied psychophysiological therapies if a patient does not yet sufficiently
accept these therapies. Otherwise, one risks souring the patient’s experience with failure from a
treatment that might later be successful when the
patient is more accepting. However, the presence
of some patient skepticism does not prevent starting applied psychophysiological therapies.
Patients May Perceive Symptoms as Out
of Their Control
Patients may believe that their symptoms are
beyond their control, even if they accept a functional or psychophysiological explanation and
believe that psychophysiological therapies can
ameliorate their symptoms. Such patients have
been described as “precontemplative,” because
they do not perceive a relationship between their
behavior and symptoms and therefore have not
even contemplated changing their behavior. Other
patients may lack sufficient self-efficacy; while
244
they do see the relationship between behavior and
symptoms, they lack the confidence to change.
They also may perceive the intensity or chronicity of their symptoms to be so severe that these
therapies cannot possibly work for them. A skilled
clinician can convince patients of the potential for
help, despite the chronicity and severity of their
symptoms, and can mobilize these patients to
comply enough to achieve successful results.
Patients May Fear
Using Passive Therapies or Loss of Control
with Relaxation Therapy
Some patients perceive relaxation therapies as
tantamount to becoming passive or losing control; this is a threatening perception for certain
patients. Therapists should look for fear as the
cause when patients are avoiding relaxation practice. Such patients need to be reassured that relaxation is in fact an active response that increases
rather than decreases self-control. Therapists can
try briefer periods of relaxation, or suggest that
patients periodically raise themselves out of the
relaxed state. The latter can reassure patients that
they can do this anytime the need arises.
Patients May Perceive a Lack of Cooperation
from Significant Others
Social support is known to facilitate adherence,
and participation of significant others may facilitate treatment. Some patients believe that these
people will not be understanding, accepting, and
cooperative. Often, such perceptions are accurate.
Therapists should consider four options. First,
give patient education booklets and/or tapes to
patients and encourage them to share these with
significant others to gain their understanding,
acceptance, and cooperation. Second, consider
directly contacting the other people, with the
patient’s permission. This option can be pursued
when cooperation from the other people is necessary but the patient is not getting it. Third, suggest
that the patient use relaxation procedures only
when other people are not around. This is the
least desirable option. Fourth, consider enlisting
others to give direct cooperation and help to the
patient. For example, others can take care of some
household or work responsibilities, reduce noise,
and answer phone calls for a few minutes. Others
can remind the patient about body postures and
relaxation. However, cooperation and help may
threaten some patients’ perceptions of self-suffi-
V. PRACTICE ISSUES
ciency. Some patients prefer to conduct their therapy without help from other people. Others have a
strong preference and capability for independence
and a good history of self-discipline.
Patients May Have Unrealistic Positive
or Negative Expectations
Unrealistic negative expectations often interfere
with compliance, as noted in several of the contexts discussed earlier. However, unrealistic positive expectations are also common and involve
expecting greater or faster benefit than usual.
Unrealistic positive expectations can result in disappointment when the expectations do not match
reality. Cognitive preparation includes promoting
realistic positive expectations.
Patients May Have Had a Prior
Inadequate Biofeedback/Relaxation Trial
and Expect Failure
Patients may perceive that biofeedback/relaxation
therapy is inadequate because they did not have
successful results with it in the past. However,
the prior therapy may have been less than ideal.
Reversing negative expectations requires extra
care, especially if one is to avoid disparaging the
prior practitioner. A therapist should consider saying something like the following:
“The treatment you received in the past sounds
like what many professionals provide. Some
professionals devote more time to developing
specialized knowledge, skills, and procedures.
Others are less specialized. Some professionals
are not in places that allow them to get preferred instruments. Most practitioners mean
well, but often they do not know what they do
not know. There is much we can add to the
therapy you had in the past. This can be important and helpful for you.”
Patients May Be Reluctant to Speak Candidly
about Psychological, Interpersonal,
and Other Stressful Matters
Some patients view a referral for relaxation and
biofeedback therapy as more “face-saving” than
a referral for psychological evaluation. These
patients sometimes expect that psychological topics will not be a focus of the evaluation and treatment. They are often reluctant to speak candidly
about psychological matters that are of potential
245
15. Adherence
importance in the formulation of an effective
treatment program. For some of them, practitioners can skirt such topics, at least in the early sessions. It is often sufficient to focus on physiological self-regulation therapies, reducing or stopping
the use of chemical stressors, and making related
changes.
Symptoms May Be Reinforcing and Symptom
Relief May Be Perceived as Threatening
Symptoms can provide reinforcement or secondary gain, despite the discomfort and impairment
they cause. Some patients may find significant
symptom relief to be threatening. Practitioners
often struggle with ways to discuss this delicate
topic with patients. Discreet practitioners consider using examples when discussing this topic
and checking for the possible role of this factor in
maintaining patients’ symptoms and in compliance problems. For instance, persons with chronic
obesity (often from their adolescence onward)
may encounter resistance from significant others or find themselves in unfamiliar heterosexual
situations when they lose considerable weight and
their figures approach normal size. They may not
believe that they have the interpersonal skills and
stress management strategies to adjust to such
conditions. They may or may not be aware of the
approach–avoidance conflict; in either case, they
will probably be reluctant to discuss their feelings
spontaneously. A common result is nonadherence
with a weight loss program.
Patients May Not Want to Stop Using
Vasoactive Dietary Chemicals, Chewing Gum,
and/or Taking Unnecessary Medications
Many patients accept recommendations to eliminate caffeine, chewing gum, and other vasoactive chemicals. However, they may resist stopping
nicotine, alcohol, and unnecessary medications.
Practitioners can advise patients of these substances’ potential negative effects on treatment,
and explain the physiological effects of caffeine,
nicotine, and other chemicals. They try to persuade patients to avoid such chemicals during at
least the hour or so before relaxation and biofeedback sessions. The effects of gum chewing on
temporalis muscles are not clear to most patients;
hence, there is a need for patient education and
demonstration. A general statement to patients
about these substances may be something like the
following:
“I know your symptoms are very distressing
to you, and you want to reduce or stop them.
I’ve explained how interferes
with therapies and self-regulation of your body.
I know you want to make progress as fast as is
practical. You probably want to limit the number of office sessions you need. Continuing to
consume these chemicals can detract from your
progress and prolong therapy. The decision
about whether you stop them or not is yours.
However, I do not want you to waste your time
and money. I will do everything I can to help
you withdraw from and stop these chemicals.
Please give this some thought, and we can discuss this further.”
Patients with Impaired
Neurocognitive Functioning
Impaired neurocognitive functioning often results
in patients’ inability to remember instructions and
therapy recommendations, and to attend to therapy tasks. Practitioners can sometimes successfully
treat such patients with help from cooperative persons living with the patients.
Evaluation and Intervention
Patient engaging in biofeedback should be
encouraged to self-monitor progress. Recommended self-report outcome measures are listed
in Table 15.2. Self-monitoring homework conveys
the importance of the information requested,
as well as the intent of the therapist to review
this information. However, requesting complicated record keeping can be counterproductive.
Patients’ records need not be 100% accurate for
practitioners to obtain useful information. A log
is often enough to allow practitioners to check
adherence and stimulate discussion. Some portion of each office session should be devoted to
reviewing the log. Clinicians may utilize simple
questions to elicit useful information,. For example, consider the value of the response “yes” to
the following questions:
• “Are you practicing your relaxation?”
• “Are you practicing your relaxation as
instructed?”
• “Are you feeling relaxed when you use the relaxation?”
• “Are you practicing your relaxation daily?”
• “Are you practicing your relaxation at various
times each day?”
246
TABLE 15.2. Recommended Methods
for Assessing Adherence
with Biofeedback Homework
1. Patient self-monitoring of frequency and duration
of relaxation practice, and times when relaxation
is used.
2. Self-report log of symptoms.
3. Self-monitoring of subjective physiological
sensations associated with relaxation.
4. Self-monitoring of cognitions associated with
relaxation.
5. Self-monitoring of caffeine, nicotine, and other
chemical use.
6. Self-monitoring of physiological parameters (e.g.,
skin temperature, pulse, blood pressure) before,
during, and after relaxation sessions.
7. Practitioner interviews.
8. Periodic psychophysiological assessments.
9. Practitioner observation of a patient’s breathing
and body postures.
10. Reports from other people in a patient’s daily life.
11. Patients’ motor functioning.
12. Patients’ daily activities.
These and similar questions are inadequate
alone to elicit useful information. Patients can
answer such questions with a “yes” or “no,” but
they may be interpreting the questions in ways
they prefer, which are not always the ways the
practitioner intends. For instance, when patients
say “yes” to “practicing daily,” does this really
mean “daily,” or does it mean “most days” or just
“some days”? Limited time to question patients in
detail, or limited interviewing skills, can result
in not enough information for assessing compliance and progress. If there is no self-report log or
if it is incomplete, the therapist might consider
the following examples of interview questions.
Open-ended questions may yield greater information. One might start by saying, “Let’s review
your relaxation practice. As accurately as you can
recall, please tell me how often and when you are
using relaxation procedures [or other procedures].”
Depending on the completeness of the patient’s
response to such a question, the therapist might
consider more specific questions:
• “How many times each day, during the past
week, did you relax for 15–20 minutes each?
How many brief relaxations of 2–10 minutes
each?”
V. PRACTICE ISSUES
• “Are there days on which you were not been
able to relax? How many days?”
• “Let’s talk about those days. What are the problems?”
• “How do you feel during your relaxation sessions? What sensations are you experiencing
during relaxation? What do you experience
after your relaxation sessions? How long do
those sensations and feelings last?”
• “In what situations did you use relaxation during the past week?”
• “When could you benefit from relaxation but
are not using it?”
Therapists should avoid misusing and misinterpreting physiological measures during office sessions to assess adherence. It is reasonable to assume
that warmer baseline skin temperatures, faster
increases in skin temperature, lower baseline muscle activity, and/or faster drops of muscle activity
reflect good relaxation experiences between office
sessions. However, a therapist cannot assume
that these are reliable indicators of compliance.
Improved baseline psychophysiological functioning in office sessions can reflect increasing comfort or habituation to the office, instrumentation,
and the therapist. Psychophysiological data are
often consistent with verbal reports of adherence
with relaxation, reduction of undesired chemicals, and/or other lifestyle changes. In such cases,
the therapist can use the data as a valuable basis
for positive verbal reinforcement of the patient.
Conversely, nonadherence should not be assumed
from the lack of psychophysiological progress during baselines. Observations of significant others
may help therapists assess and reinforce adherence
with observation of patients’ restlessness, posture,
breathing, facial muscles, and other visible cues
of tension and relaxation. Increasing use of technologies, such as portable biofeedback devices for
home use, electronic diaries, and Internet-based
interventions enhance objectivity of adherence
monitoring (Clough & Casey, 2011).
Summary and Recommendations
Nonadherence with regard to referrals, keeping
appointments, homework, and treatment recommendations undermines the effectiveness of psychological interventions and adherence should
be recognized a moderator of effectiveness in biofeedback treatment. Practitioners are encouraged
to take into account patients’ risk factors for non-
247
15. Adherence
adherence, such as dissatisfaction with therapist
or treatment, low self-efficacy, and other barriers.
Therapists are encouraged to adopt adherence
facilitation strategies as a mode of practice. There
are many tools that promise to increase adherence,
including clinic structure, practitioner communication, the therapeutic relationship, and behavioral homework. Patients present with diverse
experiences and beliefs, and providers may wish
to consider the patient’s individual cost–benefit
analysis for treatment. Likewise, it should be noted
that there is no cure for low adherence, and adherence facilitation should be maintained as long as
treatment is needed.
This chapter helps to answer a common question
asked by practitioners: “What can I do to ensure,
or increase the likelihood, that my patients will do
what I recommend?” The most pertinent recommendations are summarized in Table 15.3. It is useful for us as practitioners to remind ourselves that
we are here to serve our patients—not the reverse.
We fulfill many of our responsibilities when we provide patients a rationale for our recommendations
and present practical and achievable methods to
reach therapy goals. We fulfill other responsibilities by maintaining our credibility in the views of
our patients and preserving positive rapport with
them. It is abundantly clear that adherence is a
complex and many-sided concept that requires
great care, preparation, ingenuity, persistence, and
patience by practitioners. It requires professionals
to review their own professional behaviors, setting,
and procedures. It requires tailoring interventions
and giving patients adequate time to apply recommendations. Practitioners also need to tolerate and
function with ambiguity, and within the less than
ideal world of clinical practice. Providers should
strive toward cultivation and growth of skills to
help patients cultivate compliance, healthy attitudes, and health-promoting behaviors.
TABLE 15.3. Summary of Adherence
Recommendations
References
Persons and factors that impact adherence:
•• Professional setting and office personnel.
•• Professional’s characteristics and behaviors.
•• Interaction between the professional and patient.
•• Cognitive preparation of the patient.
•• Patient’s perceptions, expectations, and affect.
•• Patient’s family members and other significant
individuals.
Specific adherence facilitation interventions:
•• Use readily accessible, easy-to-use self-report
record systems.
•• Ask patients to record readily observable and
meaningful behaviors.
•• Instruct patients why and how to self-monitor.
•• Reinforce patients’ accuracy and completeness.
•• Convey that the patient’s records will be
reviewed.
•• Encourage patients to record behaviors,
experiences, and symptoms when they occur.
•• Establish subgoals, and review and revise them as
needed.
Important general considerations:
•• Be willing to accept less than ideal compliance
and therapeutic progress.
•• Successively approximate and shape compliance.
•• Allow patients to set their own goals and
subgoals, and discuss cost–benefit considerations.
Note
1. We prefer the term “therapeutic range,” which may
or may not be the same as a “relaxed range.” The criteria for
a relaxed range differ among practitioners; they also differ
at different stages of therapy and for different therapeutic
goals.
Bados, A., Balaguer, G., & Saldana, C. (2007). The efficacy
of cognitive-behavioral therapy and the problem of drop
out. Journal of Clinical Psychology, 63(6), 585–592.
Bandura, A. (1977). Self-efficacy: Toward a unifying theory
of behavioral change. Psychological Review, 84, 191–215.
Bandura, A. (1986). Social foundations of thought and action:
A social cognitive theory. Englewood Cliffs, NJ: Prentice
Hall.
Byrne, C. M., Solomon, M. J., Young, J. M., Rex, J., & Merlino, C. L. (2007). Biofeedback for fecal incontinence:
Short-Term outcomes of 513 consecutive patients and
predictors of successful treatment. Diseases of the Colon
and Rectum, 50(4), 417–427.
Clough, B. A., & Casey, L. M. (2011). Technological
adjuncts to enhance current psychotherapy practices: A
review. Clinical Psychology Review, 31, 279 –292.
Derisley, J., & Reynolds, S. (2000). The transtheoretical
stages of change as a predictor of premature termination, attendance and alliance in psychotherapy. British
Journal of Clinical Psychology, 39(4), 371–382.
DiMatteo, M. R., Lepper, H. S., & Croghan, T. W. (2000).
Depression is a risk factor for noncompliance with medical treatment: Meta-analysis of the effects of anxiety
and depression on patient adherence. Archives of Internal Medicine, 160, 2101–2107.
Dunbar-Jacob, J., Erlen, J. A., Schlenk, E. A., Ryan, C. M.,
Sereika, S. M., & Doswell, W. M. (2000). Adherence
in chronic disease. Annual Review Nursing Research, 18,
48–90.
248
Dunbar-Jacob, J., & Mortimer-Stephens, M. K. (2001).
Treatment adherence in chronic disease. Journal of Clinical Epidemiology, 54(Suppl. 1), S57–S60.
Elder, J. P., Ayala, G. X., & Harris, S. (1999). Theories and
intervention approaches to health-behavior change in
primary care. American Journal Preventative Medicine,
17, 275–284.
Kazantzis, N. (2000). Power to detect homework effects in
psychotherapy outcome research. Journal of Consulting
and Clinical Psychology, 68(1), 166–170.
Kinzie, M. B. (2005). Instructional design strategies for
health behavior change. Patient Education and Counseling, 56, 3–15.
LaGreca, A. M., Bearman, K. J., & Roberts, M. C. (2003).
Adherence to pediatric treatment regimens. In M. C.
Roberts (Ed.), Handbook of pediatric psychology (3rd ed.,
pp. 119–140). New York: Guilford Press.
Lambert, M. J., Hansen, N. B., & Finch, A. E. (2001).
Patient-focused research: Using patient outcome data
to enhance treatment effects. Journal of Consulting and
Clinical Psychology, 69(2), 159–172.
Macharia, W. M., Leon, G., Rowe, B. H., Stephenson, B. J.,
& Haynes, R. B.(1992). An overview of interventions
to improve compliance with appointment keeping for
medical services. Journal of the American Medical Association, 267(13), 1813–1817.
Miller, W. R., & Rollnick, S. (2002). Motivational interviewing: Preparing people for change (2nd ed.). New York:
Guilford Press.
V. PRACTICE ISSUES
O’Donohue, W. T., & Levensky, E. R. (2006). Promoting
treatment adherence: A practical handbook for health care
providers. Thousand Oaks, CA: Sage.
Prochaska, J. O., & Lorig, K. (2001). What do we know
about what works: The role of theory in patient education. In K. Lorig (Ed.), Patient education: A practical
approach (3rd ed., pp. 21–55). Thousand Oaks, CA: Sage.
Prochaska, J. O., Redding, A., & Evers, K. E. (1997). The
transtheoretical model and stages of change. In K.
Glanz, F. M. Lewis, & B. K. Rimer (Eds.), Health behavior and health education (pp. 60–84). San Francisco:
Jossey-Bass.
Reiner, R. (2008). Integrating a portable biofeedback
device into clinical practice for patients with anxiety
disorders: Results of a pilot study. Journal of Applied Psychophysiology and Biofeedback, 33, 55–61.
Reis, B. F., & Brown, L. G. (1999). Reducing psychotherapy
dropouts: Maximizing perspective convergence in the
psychotherapy dyad. Psychotherapy: Theory, Research,
Practice, Training, 36(2), 123–136.
Scheel, M. J., Hanson, W. E., & Razzhavaikina, T. I. (2004).
The process of recommending homework in psychotherapy: A review of therapist delivery methods, client
acceptability, and factors that affect compliance. Psychotherapy: Theory, Research, Practice, Training, 41(1),
38–55.
World Health Organization. (2003). Adherence to longterm therapies: Evidence for action. Geneva, Switzerland:
Author.
Ch a p ter 16
Problems Associated with
Relaxation Procedures and Biofeedback,
and Guidelines for Management
Mark S. Schwartz, Nancy M. Schwartz, and Vincent J. Monastra
• Meditation (Carrington, 2007)
• Behavioral relaxation training (Poppen, 1998)
• Electroencephalographic (EEG) biofeedback
(Thompson & Thompson, 2007; Hammond,
2010)
Most people use relaxation therapies and biofeedback-assisted relaxation without problems. Nevertheless, a few people do experience negative reactions and other problems. These difficulties can
seem alarming to the subject1 (e.g., patients, clients)
and professionals, and can result in stopping potentially useful interventions. Even if subjects experiencing such difficulties continue interventions,
they might reduce adherence with recommended
relaxation procedures and therefore reduce their
chances for improvement (Borkovec et al., 1987).
Fortunately, significant negative reactions are
uncommon and can usually be avoided or easily
managed. Practitioners who are aware of these
and other potential difficulties and their possible
causes can often prevent or lessen these effects.
Readers are referred to selected published discussions and some suggested solutions:
The reader is referred to these excellent discussions and their suggested solutions.
Negative Reactions
A partial list of the potential negative reactions
reported includes the following:
1. Musculoskeletal activity. Examples of such activ-
ities are tics, cramps, myoclonic jerks, spasms,
and restlessness.
2. Disturbing sensory experiences. These experiences include sensations of heaviness, warmth,
or cooling; feelings of depersonalization,
misperceived body size, or floating; and a variety of visual, auditory, gustatory, and olfactory
experiences.
3. Sympathetic nervous system activity. These reactions include increased heart rate and electrodermal activity.
• Progressive muscle relaxation (Bernstein, Carlson, Schmidt, 2007; Bernstein & Borkovec,
1973; Bernstein & Carlson, 1993; McGuigan &
Lehrer, 2007)
• Autogenic therapy (Linden, 2007; Schultz &
Luthe, 1969).
• Dysfunctional2 breathing (van Dixhoorn, 1997,
2007)
249
250
4. Disturbing cognitive and/or emotional reactions.
Examples include feelings of sadness, anger, and
depression; disturbing and intrusive thoughts
or mind wandering; tearfulness; and increased
anxiety and fears (e.g., losing control).
5. Other possible negative side effects. These
include hypotensive reactions, headache, sexual arousal, and psychotic symptoms.
Incidence of Relaxation‑Associated3
Negative Reactions
The very few studies about these problems involve
very few subjects. Available reports are mostly
anecdotal surveys of mental health professionals, observations of subjects or patients in studies, and small samples of patients with anxiety
disorders. Survey studies are helpful, but they are
fraught with methodological problems. See www.
marksschwartzphd.com for a summary of papers by
Jacobson and Edinger (1962), Edinger and Jacobson (1982), Blanchard, Cornish, Wittrock, and
Fahrion (1990), and Braith, McCullough, and
Bush (1988). In Blanchard et al. (1990), among
73 hypertension patients receiving temperature
biofeedback and relaxation, there were 4–9% who
reported minor negative sensations or experiences
and none reported relaxation-induced anxiety
(RIA)-type reactions. In another study, among 30
undergraduates with chronic anxiety (Braith et al.,
1988) 5 or 17% reported increased anxiety during
a taped session of progressive relaxation.
Relaxation‑Associated Negative Reactions
The term “relaxation-induced anxiety” (RIA)4 is
sometimes still used to denote a variety of negative
reactions associated with relaxation procedures.
However, there has been cogent criticism of the
term and its connotations (van Dixhoorn, 2007;
Lehrer, Woolfolk, & Sime, 2007; McGuigan &
Lehrer, 2007; Bernstein et al., 2007), as summarized in this chapter.
Heide and Borkovec (1983) defined RIA as
“paradoxical increases in cognitive, physiological,
or behavioral components of anxiety as a consequence of engaging in systematic relaxation training” (p. 171). Carrington (1977) described intense
restlessness, profuse perspiration, shivering, trembling, pounding heart, and rapid breathing associated with a type of meditation. Others refer to
subjects being frightened by some of the sensations
associated with relaxation and having fears about
V. PRACTICE ISSUES
losing control and experiencing anxiety and worrisome cognitions.
van Dixhoorn (2007) focuses on fear of losing control, especially associated with relaxation
methods such as autogenic procedures that have
suggestive instructions in comparison with procedures focused on muscle tension and movement
(e.g., progressive muscle relaxation). When anxiety/panic does occur during relaxation procedures,
van Dixhoorn (2007) and Ley (1988) attributes it
to unintended hyperventilation. This is consistent
with the view that Jacobson’s progressive relaxation, in contrast to the “post-Jacobsonian . . .
techniques” or “briefer methods” (Bernstein et al.,
2007, pp. 70–71), does not cause anxiety.
Proposed Causes of Relaxation‑Induced Negative
Reactions and Risk Factors
There are several hypothesized causes of relaxation-induced negative reactions (RANRs). Any
one or a combination of them might apply to a
specific individual. The first five were suggested by
Heide and Borkovec (1984) discussing RIA.
1. Cognitive fear of unfamiliar sensations. Some
people have cognitive fear of the sensations associated with relaxation, such as tingling, heaviness,
warmth, and muscle jerks. Patients may view these
physiological–behavioral reactions and the related
cognitive–affective reactions as uncomfortable or
unfamiliar rather than as positive signs that relaxation is occurring. This may be more common in
people who rarely or never attend to body sensations, or among those who interpret these sensations as negative.
A related hypothesis (Denny, 1976) suggests
that stimuli produced by relaxation may become
conditioned to fear when paired with a history of
punishment during relaxation or times of safety.
2. Fear of losing control. Some people are preoccupied with maintaining control over their
physical and psychological processes (Braith et al.,
1988; Lehrer, 1982) and fear losing control (van
Dixhoorn, 2007). Western culture assumes that
exercising control requires active effort. Seligman
(1975) defined “control” as being able to change
outcomes by voluntary actions. Such patients
may display a pattern of trying too hard. They
maintain a high degree of activity out of fear that
without it, they will waste time and accomplish
nothing. This fear of inactivity may lead to more
anxiety.
16. Problems with Relaxation Procedures and Biofeedback, and Management Guidelines
The fear of losing control may be more common among people who avoid rest and reflection,
and among those intent on maintaining control
with active and effortful activity. For people with
generalized anxiety, “daily maintenance of higherthen-normal tension may be a learned avoidance
response to relaxation” (Borkovec et al., 1987,
p. 887).
Associated with the loss of control is the perception that relaxation “signifies vulnerability, lack
of control over anger and sexual desire, overpassivity, etc.” (Lehrer, 1982, p. 424). This may have
been the cause of the angry feelings evoked in five
patients who were asked to relax, as reported by
Abramowitz and Wieselberg (1978).
3. Fear of experiencing anxiety. RIA is more
common in persons who are chronically anxious.
Relaxation methods often direct people to focus
away from external stimuli and on body sensations
or thoughts. This may increase their awareness of
current internal cues, which are often associated
with higher levels of anxiety. In the past, these
cues were distressing: The person viewed them as
meaning “out of my control,” or associated the cues
with heightened anxiety or even panic. For example, specific thoughts about anxiety may result in
cognitive anxiety (Norton, Rhodes, Hauch, &
Kaprowy, 1985).
4. Fear of encountering oneself. This is the
hypothesized fear of attending to the heightened
awareness of internal experience in general. Some
professionals view this phenomenon as one resulting from dissatisfaction with oneself or from fearing the increased awareness of inner conflicts.
5. Situation-produced worry, or intrusive
thoughts and worries. Patients may find, as they
reduce their focus on external stimuli, that their
own thoughts and worries arise and become more
dominant. This phenomenon is similar to cognitive intrusions’ interference with sleep onset. Note
that these thoughts do not relate to relaxation but
become associated with the relaxation experience
(see also Lichstein, 1988, p. 138).
6. Breathing-related physical changes. Another
hypothesis with support is that of van Dixhoorn
(2007), with similarities to that of Ley (1988).
Physical changes from breathing occur during
relaxation and cause or increase the chance of
having RANR. Chronic hyperventilation alters
the amount of carbon dioxide and other body
chemicals. See van Dixhoorn (2007, p. 310) for
his summary. Briefly, he refers to persons exerting
251
excessive effort, breathing too deeply, with pauses
that are too short after inhalations and/or exhalations, and/or using regulated breathing for excessively long durations.
7. Parasympathetic or trophotropic responses.
An old hypothesis is the parasympathetic or trophotropic response hypothesis. According to this
view, some people tend to have more parasympathetic responses. RIA is a compensatory ergotropic sympathetic nervous system response. This is
similar to theories such as the one proposed by
Stampler (1982), suggesting that relaxation may
directly stimulate a “complex interplay of psychological and physiological factors” (Cohen, Barlow,
& Blanchard, 1985, p. 99) that can lead to RIA in
susceptible people.
For example, DeGood and Williams (1982)
reported the case of a 40-year-old female patient
treated with autogenic training and electromyographic biofeedback for low back pain and leg
pain. They monitored her finger temperature and
skin conductance. She developed acute headaches
with nausea soon after each of the first two sessions. Revising the training procedure to having
the patient sit upright with her eyes open helped to
stop the postsession symptoms. DeGood and Williams speculated that the negative symptoms were
caused by “vagal rebound” or “parasympathetic
overcompensation” (p. 464) after physiological
deactivation during the relaxation. This explanation focuses on the possible role of the anterior
hypothalamus (Gellhorn, 1965, 1967; Mefford,
1979). According to this view, “lowered somatic
activity . . . tends to be accompanied by increased
activation of the trophotropically dominant anterior hypothalamus and related structures (DeGood
& Williams, 1982, p. 464).
8. Switching from passive to active coping.
Another possible factor involves switching from a
passive, immobilized, nonpreparatory and relaxed
state to anticipation of or preparation for action
(Elliott, 1974; Obrist, 1976, 1981; Cohen et al.,
1985). Heart rate is slower during passive coping.
The person switching from a passive to an active
coping method could experience large accelerations of heart rate, according to this explanation
(Cohen et al., 1985). Such accelerated heart rate
is presumably not because of increased anxiety,
but because of cardiac–somatic coupling. This
is the close relationship between heart rate and
striated muscle activity. Thus, preparation for
action may increase somatic arousal and heart
rate.
252
9. Other explanations. Some patients who experience RANR are competitive with themselves
and fear failure. In other individuals, relaxation
may arouse thoughts and feelings of sexual arousal.
Still others take certain medications and may confuse the side effects or interactions with the feelings induced by relaxation; this confusion may
result in RANR.
Another useful categorization of negative effects
was provided by van Dixhoorn (2007). Although
his work focused on breathing, the categorization
is broader, with implications for other relaxation
methods. He addressed the negative effects topic
under six categories: hyperventilation, increased
unpleasant awareness, relaxation overdose, RIA,
cathartic responses, and cardiac arrhythmia. See
van Dixhoorn (2007) for details. A summary follows:
Increased Unpleasant Awareness. Unpleasant
bodily experiences occur in patients with hyperventilation symptoms, including during intervention, in addition to desired and pleasant experience, thus lead Dixhoorn (2007) to recommend
including individual evaluation of unpleasant
experiences.
Relaxation Overdose. This refers to practicing
too long in an upright position, standing or sitting,
or getting up too quickly from a reclined position.
The result can be dizziness, lightheadedness, or
even faintness associated with an excessive drop in
blood pressure. The faintness typically lasts a few
minutes but can last longer. Risk factors include,
but are not limited to, being tired or after an illness. The original discussion by van Dixhoorn
(2007) is worth reviewing.
Relaxation-Induced Anxiety. For van Dixhoorn
(2007) the focus is on fear of losing control and
is associated with relaxation techniques involving suggestion rather than movement of muscles.
One implication is for breathing interventions to
include movements that affect breathing, rather
than sitting or lying quietly for too long. He also
refers to the potential for unplanned hyperventilation during any type of relaxation procedure,
which is also discussed by Ley (1988).
Cathartic Responses. The negative responses
include spontaneous movements (e.g., shivering, yawning, movements of legs or arms, whole
body jerks) or emotionally charged reactions (e.g.,
experiencing historical traumas with associated
emotions). The movements might dissipate naturally, resulting in the patient feeling refreshed
and relieved. However, if perceived as distressing,
V. PRACTICE ISSUES
solutions recommended are voluntary movements
or deep inhalation and breath holding for a few
seconds, either of which typically stops the movements.
Cardiac Arrhythmia. This refers to patients with
cardiac arrhythmias for whom certain breathing
patterns may provoke the arrhythmia. We direct
interested readers to discussion of this by van Dixhoorn (2007, pp. 312–313). In summary, based on
his very extensive experience, research, and publications regarding use of breathing therapy for cardiac patients, he stated that “sometimes breathing
is too slow and exhalation pauses too long in comparison with the state of agitation of the whole
system. Also, exhalation constricts the intrathoracic space and may stimulate the heart mechanically. The heart responds with extra beats, mostly
supraventricular; but also premature ventricular
contractions (PVSs) or even ventricular tachycardia may occur” (p. 313). For more discussion and
solutions, see van Dixhoorn (2007) and his other
publications at www.methodevandixhoorn.com/centrum/index.htm for more discussion and solutions.
Guidelines for Avoiding, Minimizing,
and Managing RANR
People who experience RANR are among those
who often most need applied psychophysiological
interventions The implication for practitioners
is not to avoid relaxation but to be aware of and
anticipate reactions. Practitioners should provide
understandable and realistic patient education
regarding the specifics of the relaxation procedures, and should select types of relaxation that
are less likely to result in these reactions with particular patients. A positive therapeutic alliance
can help practitioners manage these reactions.
A practitioner’s preparation of subjects for
relaxation includes an explanation that subjects
might experience certain sensations and thoughts
during relaxation, including the normal signs
that relaxation is taking place. This explanation
is especially important for persons with chronic
anxiety. Moreover, he or she should caution some
persons to expect intrusive thoughts in early sessions of relaxation and expect that these will
diminish as their skills and confidence increase.
In addition, the practitioner should explain that
relaxation increases rather than diminishes control. He or she explains that people often achieve
relaxation proficiency and increased autonomic
nervous system control through less rather than
more effort.
16. Problems with Relaxation Procedures and Biofeedback, and Management Guidelines
The practitioner can also consider a switch to
a different type of relaxation method. For example, if a person is having difficulty with a bodily
focus type, consider a switch to a more cognitive
approach. If a person is having trouble with a
cognitive method, consider a switch to an active,
external attentional focus or to a bodily focus type.
An example of this would be focusing on external sounds in or outside the office rather than a
mental focus on body awareness (Wells, 1990; see
also the discussion of distraction and intrusive
thoughts below).
Other Reactions and Problems,
and Guidelines for Avoiding, Reducing,
and Managing Them
1. Embarrassment. Some people feel embarrassed or self-conscious about selected relaxation
procedures (e.g., tensing facial muscles, closing
one’s eyes).
Solutions: Model the procedures and offer reassurance and supportive statements. In earlier sessions, the practitioner can look away from the subject during part of the procedure.
2. Gender-related or sexual problems. Some
subjects feel sexually aroused, self-conscious, or
threatened during relaxation. Reclining in a darkened room and using suggestive or other relaxation
terms with double meanings can add to the potential for subjective discomfort. Self-consciousness
and similar discomfort can also occur with people
(especially males) who are unaccustomed to the
passive role in any situation. Now they find themselves asked to recline passively. This can be psychologically uncomfortable for them. We assume
that many, or most, of these people are either not
aware of these reasons or they will not explain
because of their discomfort.
Solutions: Be sensitive to this potential and
gender-related factors, and adjust language and
procedures accordingly (e.g., consider starting
selected subjects in a comfortable upright posture to which they are more accustomed). Maintain appropriate boundaries and be careful about
touching subjects.
3. Script content problems. The contents of
relaxation scripts are comfortable for some patients
but uncomfortable for others. This depends on
patients’ perceptions, attitudes, and fantasies.
For example, the practitioner might consider the
potential effect of using the term “feelings of
253
heaviness” with patients who have actual or perceived weight problems.
Solutions: Tailor scripts to the subject as indicated.
4. Distraction and intrusive thoughts. There are
many sources of distraction from the concentration needed during relaxation and body awareness
procedures. These include associations to the contents of the relaxation script. Some people think
about their life and responsibilities at these times.
Solutions: Discuss this with subjects and reassure them that these distracting thoughts and
images are normal. Then provide assistance to
lessen them (e.g., consider shorter sessions, more
breaks, or having subjects start with eyes open or
partially open). Practitioners also might consider
using active sensory awareness exercises, such as
the example by Wells (1990) of external attentional focus on sounds in the environment.
One can expand this to training to focus on
sensory awareness of the environment rather than
to strive for internal awareness and a self-attentive
focus. For example, provide guidance or direction
to attend to the texture of the armchair, sounds
in the office, and the color of a wall. Then provide guidance by switching from one to another to
increase the subject’s ability to choose and to control his or her mental focus. Do these slowly, but
focus only a few seconds on each, as in this example: “Right now, think about the color of the wall.
Right now, think about the texture of the chair.
Right now, think about the sound of _________.
Now, switch your attention from one to another.”
Another suggestion is for patients to think of
the distracting thoughts as words or pictures on a
television or movie screen. Patients can then imagine the screen becoming smaller and smaller, until
it becomes tiny and distant or disappears entirely.
Patients can imagine that they are moving farther
away or that the screen is moving farther away.
Sitting close to a big color screen is more distressing than seeing the same words and images on a
3-inch black-and-white screen several feet away.
5. Restlessness and related problems. Being silent
and being motionless are paradoxically uncomfortable for some people. These people become
restless with longer sessions. Some may have features of the syndrome sometimes known as adult
attention-deficit/hyperactivity disorder (ADHD).
Problems with laughing, talking, coughing, sneezing, and other body movements are related.
Solutions: Assess and prepare for restlessness
early. Ask whether a subject has any concern about
254
sitting quietly for the planned amount of time. If
a practitioner anticipates subject discomfort, discuss this early, reassure the patient, and suggest
adjusted body positions and durations as indicated.
Patients should be given choices about physical positions, lighting, and time to make adjustments. Sessions can be shortened or interspersed
with breaks. Subjects may keep their eyes open or
partially open. Avoid long silences without verbal
instructions, discussion, changes in feedback displays and tasks, or physiological changes that are
obvious to patients.
6. Low self-efficacy and fear of failure. People
often do not have the needed self-confidence in
their abilities to develop effective relaxation skills.
The theme in their self-statements is “I cannot do
it.” They also may not have realistic goals; they
may expect the goal of therapy to be mastery. Similarly, fear of failure is a common problem. Such
people say, at least to themselves, “Am I doing
this right?”; “Am I doing this better than the last
time?”; or “I will never get the feelings and benefits
I need!”
Solutions: Provide explanation and reminders
that learning any new physical or mental skill is
a process with peaks, valleys, and plateaus. Moreover, developing or cultivating low or lower tension and arousal is often a gradual process. Using
examples from the acquisition of athletic, musical,
or other skills is often helpful. Provide encouragement and reminders to avoid hurrying and to
apply the “three P’s”—patience, practice, and persistence.
Discussing fear of failure early and periodically
can help to replace negative thoughts with positive ones. Provide reminders that most people can
make progress, and guide subjects away from viewing physiological self-regulation as something that
they “pass or fail.” In addition, encourage people
to allow relaxation to happen, or to let go, rather
than trying to make it happen. Focus on increasing
awareness of feelings associated with physiological
self-regulation. Finally, shift the goals away from
specific numbers and using performance competition as a model.
7. Increased awareness of tension. People, typically patients, sometimes report increased symptoms or other signs of dysfunction during the early
stages of relaxation therapies. Because general
relaxation permits more awareness of tension of
selected body areas, people may perceive themselves as more tense than they were before. This
does not mean that patients are actually more
V. PRACTICE ISSUES
tense, simply that they are more body-aware. An
increased awareness of tension can also result
from increased focus on symptoms through the
use of self-report symptom logs. Some relaxation
procedures, such as tensing muscles, can increase
some symptoms as well. Some body areas remain
tense during some relaxation procedures, including tense–release procedures.
Solutions: Discuss this phenomenon and reassure patients that such perceptions are common
and normal. By noticing tension earlier, one
can reduce the tension and prevent symptoms.
Patients can reframe the belief in increased tension as increased awareness.
8. Problems with significant others. Family members and other significant people in a subject’s life
during relaxation therapy may not be understanding and cooperative.
Solutions: If such people are not present during office sessions, provide education materials
explaining the rationale, procedures, and need
for cooperation from others. Practitioners may
need to provide counseling on how to discuss recommended relaxation with others. and how to
increase cooperation from family members.
9. Factors not related to starting relaxation. Factors other than starting relaxation therapy can
increase symptoms. If a person is not cognitively
prepared for the intervention, he or she can experience increased concern, emotional arousal, and
tension, therefore increasing symptoms. Another
factor is the presence of continuing or even
increasing stress that is added to the symptoms.
Solutions: Discussion of current life events and
counseling, adjustments, and reassurance are
appropriate. Rule out other causes, such as an
incorrect diagnosis and/or intervention strategy.
10. Viewing treatment as stressful. There are
patients who view some aspects of treatment as
stressful, and this can add to symptoms. Consider the time allotments and other arrangements
required for patients to attend office sessions.
These include time away from other duties and
responsibilities, explanations to employers and
supervisors, and often extra work when patients
get back from appointments. People invest time
and effort in carrying out homework assignments,
maintaining self-report records, and completing
questionnaires. All of these pressures are stressful,
and added to them are the expenses for intervention. Thoughts about any or all of these added
sources of stress can intrude.
16. Problems with Relaxation Procedures and Biofeedback, and Management Guidelines
Solutions: Sensitivity and flexibility on the practitioner’s part about scheduling and assignments
can help to decrease the effects of this stress.
11. Disregarding instructions. Subjects (e.g.,
patients) sometimes disregard instructions during biofeedback and other relaxation procedures
in the practitioner’s office and elsewhere. For
example, they may not imagine the stress stimuli
the practitioner presents, or they may imagine the
stress stimuli only part of the time. In addition,
some subjects intentionally think of topics other
than the biofeedback signal or verbal relaxation
instructions. People are unlikely to admit to such
diversions without careful questioning. Be careful in discussing this, in order to avoid giving the
impression of being critical.
Solutions: Sensitivity and flexibility on the practitioner’s part about assignments and instructions
can help. Once again, provide adequate information. A practitioner can change the content of
the relaxation script and procedures, or consider
saying something like the following: “Sometimes
you might be thinking about other topics during
relaxation and biofeedback. It is normal for that to
occur at times, and I understand. I need to know
if this is happening. Please share it with me when
it is happening.”
12. Not focusing on physiology. Some people
listen to prerecorded relaxation but do not focus
on their physiology. Similarly, some just listen to
or watch feedback signals, but with minimal or no
focus on their physiology. It is as if they expect or
hope that the relaxation instructions and biofeedback signals themselves will induce the desired
outcome.
Solutions: Anticipate this potential problem and
discuss it. Provide guidance away from a scenario
of passively expecting the prerecorded relaxation
procedures or feedback signal to be therapeutic by
themselves.
13. Falling asleep. Some people doze or start
sleeping during relaxation procedures.
Solutions: Be aware of this potential. Schedule
sessions earlier in the day, or avoid relaxation after
meals, unless it is needed for postprandial symptoms. In persistent cases, consider a sleep disorder
evaluation to check for sleep apnea, psychophysiological insomnia, sleep–wake schedule disorders,
and narcolepsy.
14. Misuse of relaxation audiotapes. Some
patients are dependent on prerecorded relaxation
methods and use them too often and too long.
255
They rely on them and do not learn to relax without them.
Solutions: Clarify the proper role of prerecorded
scripts and encourage people to avoid dependence
on them. If dependency does occur, taper via fading and related behavioral techniques. Consider
using progressively briefer scripts. People can turn
the recording off progressively earlier in the script,
or lower the volume gradually and continue the
relaxation.
15. Not having or taking enough time to practice
and apply relaxation. This is a very common problem. Substantial time use problems are common.
Some people do not know how to or have not
applied effective time use management in their
lives.
Solutions: Practitioners can conduct, or refer
such people for an evaluation of and education
about, time use management. For example, help
such people learn to set goals and priorities, to
delegate responsibility appropriately, to avoid or
reduce activities that waste time, to reduce perfectionism, and to manage procrastination. Encourage scheduling relaxation and making practice
and application a high priority.
16. Specific problems with biofeedback procedures. Problems specific to biofeedback can occur
during biofeedback-assisted relaxation sessions
(Gaarder & Montgomery, 1981, p. 94). We include
some of these problems here and suggest a few
solutions. Experienced practitioners can develop
their own repertoires.
a. Very small changes in a physiological
parameter, or no patient perception the
feedback signal is changing.
Solutions: Use the threshold or change
it to ease the task. Increase the gain, so
that the visual display feedback or audio
feedback changes are more obvious, with
smaller physiological changes. Encourage
shifting attention to other sites or cognitions, as well as switching to a different
task or feedback site. Use varied verbal
relaxation instructions or change the
visual display. Ask the subjects to close
their eyes for a few moments, then freeze
the visual display if it changes in the
desired direction. Then ask subjects to
open their eyes gently to see the change.
Repeat as indicated.
b. Movement of the feedback signal in the
undesired direction.
256
V. PRACTICE ISSUES
Solutions: Stop instrumentation-based
feedback. Discuss “trying too hard” and
provide quiet, brief, and clear verbal
feedback when the signal moves in the
desired direction. Observe the person’s
posture, breathing, and movements, and
suggest adjustments as needed. Shift the
focus to other sites, tasks, or techniques.
Use varied verbal relaxation instructions, ask about cognitions, and discuss
and suggest changes as needed. Increased
control is the ability to move the signal in
either direction. The experienced practitioner will notice what makes it go in the
wrong direction, as this can give clues for
moving it in the desired direction.
c. Patient fatigue with the feedback signal or
task.
Solutions: Change the feedback displays.
With computer-based systems, there are
a wide variety of feedback options. Conduct shorter sessions or adjust the goals
of a session to increase the chance of
obvious successes and reinforce changes.
17. Inability to exhibit a relaxation posture due
to physical limitations. Poppen (1998) discusses
the specialized procedures he developed and calls
behavioral relaxation training (BRT).5 Physical
limitations include scoliosis, arthritis, or unequal
leg length that can interfere with the patient’s
ability to implement BRT.
Solutions: Be alert to these limitations and tailor
adjustments in the relaxation process and criteria.
18. Breathing with open mouth while relaxing
results in dry oral cavity.
Solutions: Breathe nasally and close mouth.
19. Frustration from feeling that performance of
specific procedures is impossible. Examples are being
unable to keep eyes closed without twitching of
the eyelids, breathing at a slower rate, or swallowing less often.
Solutions: Provide reassurance that 100% perfection is not expected. Also provide positive feedback anytime the relaxed behavior occurs.
Side Effects of EEG Biofeedback
As Thompson and Thompson (2007) discussed,
there are a few reports of side effects and some
theoretically possible side effects associated with
EEG biofeedback.
Alpha–Theta
“Abreaction,” or spontaneously occurring images,
as observed and reported by Peniston and Kulkosky
(1989), occurred in some of their patients diagnosed with alcoholism and being treated with
alpha and theta enhancement. This recollection
of traumatic events is possible in patients with a
history of abuse or psychological trauma and is
presumably associated with increased theta as it is
a “state in which the unconscious becomes more
accessible,” as noted by Thompson and Thompson
(2007, p. 265). They also note that there are psychoanalysts “who use EEG feedback to help their
patients get into a hypnagogic state . . . in which
the person in analysis can access free associations
and dream-like states . . . [allowing] memories and
fantasies to emerge, and therein lies its potential
for use in psychotherapy, and work with positive
replacement imaging (Peniston & Kulkowsky,
1990) as well as its danger” (p. 265). The risk is
that “any suggestions occurring in this state can
solidify the mixture of fantasy and memory, with
the unfortunate production of ‘false memories’ ”
(Thompson & Thompson, 2007, p. 265).
Precipitating a seizure is theoretically possible
in vulnerable persons, with increased drowsiness
associated with increased theta. Neurotherapy
with added techniques, for example, audiovisual
stimulation or HEG (hemoencephalography or
brain blood flow biofeedback) theoretically might
be associated with more side effects, although none
were noted by Thompson and Thompson (2007).
Beta training has been reported by some professionals to increase overactivity after a session
(Thompson & Thompson, 2007), although the
authors observed this in only one of about 2000
clients in which inadvertent training occurred.
Greater hypnotizability among clients with
ADHD is associated with their excess theta,
as noted by Thompson and Thompson (2007).
This might theoretically result in unwanted and
unplanned expectations if clients are asked questions about agitation or excitability from the training. These authors do not report any such cases in
their own experience or that of others.
Training the wrong (intentionally or unintentionally) EEG frequencies at specific sites might,
theoretically, result in undesired side effects. For
example, Lubar (1991) intentionally reversed EEG
feedback reinforcement conditions. After subjects
successfully learned to suppress theta activity (4–8
Hz) and increase beta (16–20 Hz), the experimenters intentionally reversed this to increase
16. Problems with Relaxation Procedures and Biofeedback, and Management Guidelines
theta and inhibit beta. In contrast to the previously achieved desired improvement of attention
and behaviors, the reversal resulted in the patients
becoming more disorganized, inattentive, and
hyperactive. Resumption of the initial protocol
resulted in a return to improved functioning. This
reversal illustrates the dramatic effect that operant
conditioning of specified EEG frequency bands
can exert on human functioning.
Monastra, Monastra, and George (2002) noted
certain adverse side effects associated with combining EEG biofeedback with medication. Stimulant therapy (Ritalin) was administered to all
100 patients with ADHD in their study, and a
titrated dose was maintained for a 1-year period.
During that time, 51 of the patients received EEG
biofeedback; the other 49 did not. Only patients
receiving the EEG biofeedback developed
increased cortical arousal over the central midline region, which was associated with sustained
improvement on behavioral and neuropsychological measures. By the conclusion of the first
year, evidence of increased irritability was noted
in approximately 20% of the patients who had
received EEG biofeedback. This irritability was
eliminated once stimulant therapy was reduced or
discontinued. Consequently, clinicians treating
patients diagnosed with ADHD with a combination of stimulant therapy and EEG biofeedback
may find it useful to monitor patients as treatment progresses, and to consider a reduction of
stimulant dose should irritability emerge.
Side Effects Associated with the Use of Heart Rate
Variability Biofeedback
Lehrer (2007) discussed some people’s tendency
to hyperventilate when they start relaxed breathing during the early sessions of heart rate variability (HRV) biofeedback. Thus, his training
manual (Lehrer, Vaschillo, & Vaschillo, 2000;
Lehrer, 2007) guides clients/patients toward shallow breathing initially, and educates and sensitizes them regarding hyperventilation symptoms.
He suggests using a capnometer to monitor endtidal CO2 in early sessions, if available, although
Thompson and Thompson (2007) note that only
a “small number of practitioners measure CO2. We
do not know the percentage, but we, too, suspect
that it is only a small percentage and typically is
not practical. However, if available, and assuming
practitioner training and expertise with a capnometer, it could be useful for selected situations such
as these.
257
Research on Negative Reactions
and Other Problems
More research is needed on the incidence and
mechanisms of negative reactions and other problems associated with relaxation therapies, including biofeedback-assisted relaxation (Poppen, 1984;
Edinger, 1984). Practitioners also need to conduct
more research on patient variables associated with
these difficulties and on preventive and management procedures. Retrospective survey research
is fraught with enough methodological problems
to preclude its value for estimating the incidence
of these difficulties. (For more on this, see www.
marksschwartzphd.com.) Prospective research
needs to control for many variables, including
relaxation procedures, biofeedback modality and
procedures, presence or absence of therapist, subject preparation, duration of sessions, details of
the negative reactions, whether eyes are open or
closed, antecedent events, postural information,
lighting information, therapist characteristics,
types of physical and mental symptoms and disorders, patients’ prior experience with these and
other therapies, and medications and doses.
Cautions and Contraindications
There are other, more serious factors to consider
when practitioners are providing physiological
self-regulatory therapies. These constitute the cautions and contraindications (see M. S. Schwartz,
Chapter 14, this volume, for a discussion of these).
Therapists can expect potentially serious problems
to occur if they provide these therapies to patients
for whom such cautions and contraindications
apply. However, they can use special approaches
for carefully selected patients with some of these
disorders and conditions, if the practitioners are
knowledgeable about and experienced with these
disorders.
Conclusion
Various negative effects and other problems can
occur as part of some relaxation therapies and
some types of biofeedback. The experience of
significant negative reactions can interfere with
interventions and reduce adherence with recommendations. However, very few people are at
risk, and very few experience negative reactions.
Among those who do experience various prob-
258
lems, there are good solutions. Prudent practitioners use available information, wisdom from experience, skills, precautions, patient education, and
good judgment in patient selection and implementation of treatments.
Interventions with relaxation procedures and
biofeedback should be provided by professionals
with appropriate credentials, training, and experience, or by someone under the direct and personal
supervision of a professional so qualified.
Notes
1. For convenience in this chapter, we use the term
“subject” to denote patients, clients, trainees, performers,
athletes, and research subjects.
2. van Dixhoorn (1997)) recommended the concept
and term “dysfunctional breathing” rather than “hyperventilation syndrome” (HVS). The replacement term is more
encompassing and includes the hypocapnia associated with
HVS.
3. In prior editions, the term “relaxation-induced” was
used. The term used here, “relaxation-associated,” indicates
that the reactions are thought to be due to multiple factors
(e.g., procedural, relaxation, breathing, cognitive, past history, diagnoses). This also is more consistent with Poppen
(1998), who suggested the term “training-induced arousal.”
4. Poppen (1998) also questioned the accuracy of the
term “relaxation-induced anxiety” for describing the phenomenon. He suggested a more general and descriptive
term such as “training-induced arousal.” The rationale he
gives is “studies of the phenomenon have not indicated that
any degree of relaxation occurs prior to the trainee’s upset;
instead arousal occurs before training or right at the beginning” (p. 86).
The second reason is that the term “anxiety” implies
a specific set of behaviors, but that term does not always
accurately describe the discomfort experienced by some
persons. The examples given by Poppen (1998) involve
nausea or falling sensations, which he described as “not
properly labeled anxiety.”
5. BRT is useful for various populations, but it is particularly useful for special populations, such as those with
intellectual disabilities (i.e., severe retardation, acquired
brain injury, hyperactivity disorder, and schizophrenia).
Poppen uses the terms “trainee” and “client” rather than
“patient.” Items 17 through 20 are gleaned from Poppen
(1998).
References
Abramowitz, S. I., & Wieselberg, N. (1978). Reaction to
relaxation and desensitization outcome: Five angry
treatment failures. American Journal of Psychiatry, 135,
1418–1419.
V. PRACTICE ISSUES
Bernstein, D. A., & Borkovec, T. D. (1973). Progressive
relaxation training. Champaign, IL: Research Press.
Bernstein, D. A., & Carlson, C. R. (1993). Progressive
relaxation: Abbreviated methods. In P. M. Lehrer & R.
L. Woolfolk (Eds.), Principles and practice of stress management (2nd ed., pp 53–87). New York: Guilford Press.
Bernstein, D. A., Carlson, C. R., & Schmnidt, J. E. (2007).
Progressive relaxation: Abbreviated methods. In P. M.
Lehrer, R. L. Woolfolk, & W. E. Sime (Eds.), Principles
and practice of stress management (3rd ed.). New York:
Guilford Press.
Blanchard, E. B., Cornish, P. J., Wittrock, D. A., & Fahrion,
S. (1990). Subjective experiences associated with thermal biofeedback treatment of hypertension. Biofeedback
and Self-Regulation, 15(2), 145–159.
Borkovec, T. D., Mathews, A. M., Chambers, A., Ebrahimi,
S., Lytle, R., & Nelson, R. (1987). The effects of relaxation training with cognitive or nondirective therapy
and the role of relaxation-induced anxiety in the treatment of generalized anxiety. Journal of Consulting and
Clinical Psychology, 55(6), 883–888.
Braith, J. A., McCullough, J. P., & Bush, J. P. (1988). Relaxation-induced anxiety in a subclinical sample of chronically anxious subjects. Journal of Behavior Therapy and
Experimental Psychiatry, 19(3), 193–198.
Carrington, P. (1977). Freedom in meditation. Garden City,
NY: Doubleday/Anchor.
Carrington, P. (2007). Modern forms of mantra meditation.
In P. M. Lehrer, R. L. Woolfolk, & W. E. Sime (Eds.),
Principles and practice of stress management (3rd ed.).
New York: Guilford Press.
Cohen, A. S., Barlow, D. H., & Blanchard, E. B. (1985).
Psychophysiology of relaxation-associated panic attacks.
Journal of Abnormal Psychology, 94(1), 96–101.
DeGood, D. E., & Williams, E. M. (1982). Parasympathetic
rebound following EMG biofeedback training: A case
study. Biofeedback an