Pain Perception in Recovered Bulimia
Nervosa Patients
Daniel Stein,1,2 Walter H. Kaye,2* Hisato Matsunaga2,3 and Daniel Myers,4
Israel Orbach,5 Dov Har-Even,5 Guido Frank,2 and Rhadika Rao2
1
Sheba Medical Center, 21Tel Hashomer, Israel
Western Psychiatric Institute and Clinic, University of Pittsburgh, Pennsylvania
3
Department of Neuropsychiatry, Osaka City University Medical School, Osaka, Japan
4
Department of Oral Medicine and Pathology, University of Pittsburgh Medical School,
Pittsburgh, Pennsylvania
5
Department of Psychology, Bar Ilan University, Ramat Gan, Israel
2
Accepted 18 December 2002
Abstract: Objective: Decreased pain sensitivity is found in individuals who are ill with
bulimia nervosa (BN). The purpose of this study is to determine whether altered pain
perception persists after recovery from bulimia nervosa (RBN). Methods: Eleven women
who were recovered from BN for more than 1 year were compared with 15 healthy
volunteer women. The participants received two pain evaluations—thermal pain stimulation
(TPS), which evaluates threshold and tolerance to heat, and the submaximal effort tourniquet
test (SETT), which assesses threshold and tolerance to ischemic pain induced by inflation of a
blood pressure cuff. Results: Compared with the controls, the RBN women showed elevated
pain threshold as measured with the SETT and a tendency to elevated pain threshold on the
TPS. Discussion: Decreased pain sensitivity persists after recovery from BN and may reflect
altered modulatory function in this illness. # 2003 by Wiley Periodicals, Inc. Int J Eat Disord
34: 331–336, 2003.
Key words: bulimia nervosa; pain threshold; pain tolerance
INTRODUCTION
Acutely ill bulimia nervosa (BN) patients generally show elevated levels of thermally
induced (Lautenbacher, Pauls, Strian, Pirke, & Krieg, 1991; de Zwaan, Biener, Bach,
Wiesnagrotzki, & Stacher, 1996a; de Zwaan, Biener, Schneider, & Stacher, 1996b)
and/or mechanically induced (pressure) (Faris, Raymond, de Zwann, Howard, Eckert,
& Mitchell, 1992; de Zwaan et al., 1996a,b; Raymond, Eckert, Hamalainen, Evanson,
Thuras, & Faris, 1999; Faris, Raymond, de Zwann, Howard, Crosby, Mitchell, Meller, Kim,
*Correspondence to: Walter H. Kaye, MD, University of Pittsburgh, Western Psychiatric Institute and Clinic,
3811 O’Hara Street, Pittsburgh, PA 15213. E-mail: kayewh@msx.upmc.edu
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/eat.10164
#
2003 by Wiley Periodicals, Inc.
332
Stein et al.
Hartmen, & Eckert, in press) pain threshold and/or tolerance compared with controls.
The mechanisms underlying these phenomena have yet to be determined. Recent data
suggest that decreased pain sensitivity may be related to severity of bulimic symptoms
(Faris et al., in press). Furthermore, pain detection threshold may increase during the
interval between bingeing/purging episodes (Faris et al., 1998), and BN patients do not
report greater pain unpleasantness compared with controls, despite showing significantly longer time before reaching pain tolerance (Girdler, Koo-Loeb, Pedersen, Brown,
& Maixner, 1998). These findings raise the speculation that BN-related diminished pain
sensitivity and pain unpleasantness may play a role in the maintenance of BN in
minimizing the physical discomfort and increasing the self-reported calm and euphoria
associated with the binge/purge cycle (Faris, Kim, Meller, Goodale, Hofbauer, Oakmon,
Howard, Stevens, Eckert, & Hartmen, 1998; Girdler et al., 1998). Girdler and associates
(1998) have additionally speculated that complex interactions between elevated opioid
activity and vagal activity may be implicated in the hypoalgesia of BN.
The aim of this study was to assess pain perception in women who were recovered
from BN to determine whether decreased pain sensitivity is only a phenomenon related
to pathological eating behavior or whether it persists after recovery.
METHODS
Patients
Participants consisted of 11 recovered BN female patients (RBN) and 15 healthy
matched volunteer women. All RBN patients had a previous lifetime DSM-IV (APA,
1994) diagnosis of BN and were required to have no history of AN. These participants
also reported that in the past year they maintained a stable and normal weight (85% to
115% of average body weight [ABW]) (Metropolitan Life Tables, 1959); had not binged,
purged (both vomiting and use of laxatives or diuretics), or engaged in restrictive eating
patterns; and had regular menstrual cycles.
RBN participants fulfilling all the inclusion but none of the following exclusion
criteria were admitted to the study: (1) any axis I (DSM-IV, 1994) disorder in the past year.
(2) Ever having a bipolar disorder (I and II), a schizophrenic spectrum disorder, organic
brain syndrome, and dementia. The RBN patients had, in addition, no evidence of
present or past medical illness, neurological illness, or chronic pain condition; had not
used psychoactive medications in the past year; and had no current or regular use of
analgesics or of any other medications. These patients were previously treated in the
eating disorders treatment program at the Western Psychiatric Institute and Clinic,
Pittsburgh, Pennsylvania.
Healthy volunteers consisted of 15 age-matched women whose weight has been
between 90% and 115% of ABW since menarche. These volunteers had no lifetime or
present psychiatric, medical, or neurological illness; no chronic pain condition; no regular
use of analgesics or any other medications; and no stigmata suggestive of an eating
disorder (ED). These volunteers came from the staff of the Western Psychiatric Institute
and Clinic, Pittsburgh, Pennsylvania. All RBN patients and healthy controls were
Caucasian.
To control for menstrual variations in pain sensitivity (Lautenbacher, Pauls, Strian,
Pirke, & Krieg, 1990), all RBN and control participants were studied during the first
14 days (follicular phase) of their menstrual cycle. ABW for each participant was assessed
Pain Perception
333
with the Metropolitan Life Tables (1959), and body mass index was calculated as kilograms divided by height in meters squared. All participants gave informed consent to
participate in the study, which was approved by the center’s institutional review board
and were paid for participating in the study.
Instruments
Measurement of Pain Perception
Thermal Pain Stimulator (TPS). This instrument is modeled after Procacci (1979).
In the first stage, a 1.5-cm radius area was colored in black ink on the participant’s wrist.
A 75-watt bulb was then held 7 cm from the wrist. Pain threshold was defined as the total
time measured (with a stopwatch) from onset of illumination until the participant
reported feeling pain, and pain tolerance was the total time measured from onset of
illumination until the participant withdrew her wrist. To ensure thermal levels, the bulb
was illuminated for 5 minutes before use with each participant. The procedure was
stopped after 8 minutes according to the requirements of previous studies. It was
repeatedly shown to be a valid and reliable pain measure (Procacci, 1979; Orbach,
Palgi, Stein, Har-Even, Loten-Pelag, Asterov, & Elizar, 1996).
Submaximal Effort Tourniquet Test (SETT). This technique has been extensively used in
the study of experimental pain in humans (Moore, Duncan, Scott, Gregg, & Ghia, 1978),
as well as in a previous study of ill BN patients (Girdler et al., 1998). The first step was
raising the arm for 10 seconds and temporary application of a tight bandage to drain
residual blood. Ischemia of the forearm was achieved by inflation of a blood pressure cuff
to 250 mmHg. Maximum handgrip was determined immediately before ischemia by use
of a handgrip dynamometer. The participant was instructed to squeeze the dynamometer
with as much force as possible. Once ischemia was produced, participants squeezed the
dynamometer 20 times at 30% maximal grip force. Each squeeze was for 2 seconds, with
an intergrip interval of 2 seconds. The onset, duration, and magnitude of each handgrip
were signaled by computer-controlled signals. After the 20 grips were completed, participants were asked to report the time they started to feel any pain (pain threshold,
measured with a stopwatch from the onset of cuff inflation) and the time the pain
becomes intolerable (pain tolerance) (Maixner Fillingim, Booker, & Sigurdson, 1995).
The cuff was deflated at that time or after 25 minutes, whichever came first, to guarantee
the participants’ safety (Maixner et al., 1995).
Psychiatric diagnosis was assessed with the Structured Clinical Interview for DSM-IV
Axis I Disorders-Patient Edition (SCID-I/P, Version 2.0) (First, Spitzer, Gibbon, &
Williams, 1995). Recovery from past BN, or not ever having an ED in the controls, was
assessed with the Eating Disorders Family History Interview (EDFHI) (Strober, 1987).
Procedure
A screening interview was first conducted by a psychiatrist (DS). If the participant was
screened positively, she received the SCID-I/P, Version 2.0, the EDFHI, and a thorough
evaluation of medical and neurological condition, including measurement of weight and
height. If the participant fulfilled all inclusion and exclusion criteria, she was accepted
into the protocol.
All participants were then admitted for 1 day to the Eating Disorders Research
Laboratory of the Center for Overcoming Problem Eating, Pittsburgh, Pennsylvania, and
participated in an identical study design. The participants were seated in a comfortable
chair, and all evaluations took place in the supine position. TPS was administered first
334
Stein et al.
and the SETT 60 minutes after completing the TPS. The two brain hemispheres handle
pain differently, and the maximal handgrip used in the SETT is expected to be different
in the dominant compared with the nondominant hand. Therefore, we assigned the
participants randomly to be tested with either the TPS or the SETT in their dominant
hand. To control for circadian variations in pain response (Moore et al., 1978), we
performed all evaluations between 12:00 to 16:00. Evaluators were trained in the use of
the TPS and SETT and were blind to the diagnosis of the participants.
Statistical Analysis
Two-tailed t tests and chi-square tests were used for the between-group comparison of
the demographic and clinical parameters and the TPS and SETT pain threshold. Pain
tolerance cannot be reached until pain is detected, namely, within an individual it is
dependent on the pain threshold score. We therefore used an analysis of covariance with
pain threshold as a covariate (ANCOVA) to assess the between-group differences in pain
tolerance. All results are described as mean SD.
RESULTS
The mean age of the RBN participants was 28.8 5.5 years and of the controls 25.1 4.8 years (t[df ¼ 24] ¼ 1.80; NS). The mean ABWs for the RBN participants and controls
were 107.8% 12.1% and 104.6% 10.6%, respectively (t[24] ¼ 0.70; NS), and the mean
body mass indexes were 22.4 2.5 and 21.7 2.3 kg/m2, respectively (t[24] ¼ 0.76; NS).
The RBN women were recovered for a mean period of 5.5 years (range, 1–12 years).
In terms of the TPS test, the pain threshold for 7 of the 11 RBN patients and 3 of the 15
controls was more than 8 minutes, defined as the time limit for stopping the experiment.
This precluded any meaningful analyses concerning TPS pain tolerance. Nevertheless, a
significantly greater percentage of the RBN patients had TPS pain threshold of more than
8 minutes compared with the controls (2[df ¼ 1] ¼ 4.03; p < 0.04).
For the SETT, we found that pain threshold was 3.10 2.7 minutes for the RBN patients
and 1.15 1.6 minutes for the controls (t[df ¼ 24] ¼ 2.08; p < 0.05). Although a significant
difference was also shown for pain tolerance between the RBN (8.88 5.3 minutes)
and control participants (3.63 3.5 minutes) (t[df ¼ 24] ¼ 2.20; p < 0.04), this difference
failed to retain its significance when using ANCOVA with pain threshold as a covariate
(F[1,21] ¼ 0.04; NS). The evaluation of the TPS and SETT in either the dominant or
nondominant hand had no significant influence on any of the pain measures.
DISCUSSION
To our knowledge, this is the first study showing that elevated pain threshold persists
after recovery from BN. The persistence of decreased pain sensitivity after recovery raises
the possibility that it is a trait associated with the pathophysiology of BN, although it is
possible that it is a consequence of this disorder. Recent findings in acutely ill BN patients
suggest a putative role of pain in the maintenance of BN. That is, bingeing and purging
may be a means of temporary normalization of elevated pain threshold (Faris et al., 1998).
Perhaps pain threshold reflects exaggerated modulation of other phenomena such as
Pain Perception
335
mood states, which binge and purge behaviors may also temporarily blunt (Steinberg,
Tobin, & Johnson, 1990).
The etiology underlying persistent decreased pain sensitivity after recovery from BN
has yet to be determined. Recent studies raise the possibility that altered central serotonergic activity may persist after recovery from BN (Kaye, Greeno, Moss, Fernstrom,
Lilenfeld, Waltzin, & Mann, 1998). Decreased pain sensitivity after recovery may also
be related to altered serotonin activity. In this respect, drugs that act on serotonin systems
are known to have effects on pain sensitivity (e.g., Schreiber, Stein, & Floman, 1996).
Our study has several limitations. The sample size is relatively small. This precludes
the assessment of a possible confounding effect of the order of administration of the pain
measurements. The TPS may be considered less accurate than currently used computercontrolled pain measure devices (Lautenbacher et al., 1990). The requirement to stop this
procedure after 8 minutes precludes meaningful analyses of TPS pain threshold and
tolerance. Nevertheless, the finding that the TPS pain threshold of most RBN patients is
more than 8 minutes compared with only a few of the controls may be of relevance,
because it is consistent with the results of the SETT. Our results need to be interpreted
with caution, because techniques used for the determination of pain thresholds at which
stimuli of different nature are perceived as painful are prone to yield results influenced
by factors not necessarily related to pain perception (de Zwaan et al., 1996b).
The advantages of our study include the careful selection of the RBN patients and
controls and the exclusion of participants having other DSM-IV axis I diagnoses that may
affect pain perception (e.g., affective disorders or substance use) (Lautenbacher & Krieg,
1994). In addition, ischemic pain generated with the SETT is considered an experimental
pain-induction method that particularly mimics pain sensations similar to those experienced in the natural environment (Moore et al., 1978).
In conclusion, this study is the first to suggest that pain threshold may be elevated not
only in acutely ill but also in recovered BN patients. Further studies in larger groups of
RBN patients assessing the association of decreased pain sensitivity with elevated vagal,
opioid, and/or serotonergic activity, as well as with psychological correlates that persist
after recovery, are required to establish the importance of pain perception in the pathophysiology of BN.
REFERENCES
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (3rd ed.),
Washington, DC: Author.
de Zwaan, M., Biener, D., Bach, M., Wiesnagrotzki, S., & Stacher, G. (1996a). Pain sensitivity, alexithymia, and
depression in patients with eating disorders: are they related? Journal of Psychosomatic Research, 41, 65–70.
de Zwaan, M., Biener, D., Schneider, C., & Stacher, G. (1996b). Relationship between thresholds to thermally and
to mechanically induced pain in patients with eating disorders and healthy controls. Pain, 67, 511–512.
Faris, P.L., Raymond, N.C., de Zwaan, M., Howard, L.A., Eckert, E.D., & Mitchell, J.E. (1992). Nociceptive, but
not tactile, thresholds are elevated in bulimia nervosa. Biological Psychiatry, 32, 462–466.
Faris, P.L., Kim, S.W., Meller, W.H., Goodale, R.L., Hofbauer, R.D., Oakman, S.A., Howard, L., Stevens, E.R.,
Eckert, E.D., & Hartman, B.K. (1998). Effects of ondansteron, a 5-HT3 receptor antagonist, on the dynamic
association between bulimic behaviors and pain thresholds. Pain, 77, 297–303.
Faris, P.L., Raymond, N.C., de Zwaan, M., Howard, L.A., Crosby, R.D., Mitchell, J.E., Meller, W.H., Kim, S.W.,
Hartman, B.K., & Eckert, E.D. (in press). Elevated pain thresholds in bulimia nervosa: relationship to disease
symptoms. Physiology and Behavior.
First, M.B., Spitzer, R.L., Gibbon, M., & Williams, J.B.W. (1995). Structured Clinical Interview for DSM-IV Axis I
Disorders—Patient Edition (SCID-I/P, Version 2.0). New York: Biometrics Research Institute, New York
State Psychiatric Institute.
336
Stein et al.
Girdler, S.S., Koo-Loeb, J., Pedersen, C.A., Brown, H.J., & Maixner, W. (1998). Blood pressure-related hypoalgesia in bulimia nervosa. Psychosomatic Medicine, 60, 736–743.
Kaye, W.H., Greeno, C.G., Moss, H., Fernstrom, J., Fernstrom, M., Lilenfeld, L.R., Weltzin, T.E., & Mann, J.J.
(1998). Alterations in serotonin activity and psychiatric symptomatology after recovery from bulimia
nervosa. Archives of General Psychiatry, 55, 927–935.
Lautenbacher, S., Pauls, A.M., Strian, F., Pirke, K.M., & Krieg, J.C. (1990). Pain perception in patients with eating
disorders. Psychosomatic Medicine, 52, 673–682.
Lautenbacher, S., Pauls, A.M., Strian, F., Pirke, K.M., & Krieg, J.C. (1991). Pain sensitivity in anorexia nervosa
and bulimia nervosa. Biological Psychiatry, 29, 1073–1078.
Lautenbacher, S., & Krieg, J.C. (1994). Pain perception in psychiatric disorders: a review of the literature. Journal
of Psychiatric Research, 28, 109–122.
Maixner, W., Fillingim, R, Booker, D., & Sigurdson, A. (1995). Sensitivity of patients with painful temporomandibular disorders to experimentally evoked pain. Pain, 63, 341–351.
Metropolitan Life Insurance Company. (1959). New weight standards for men and women. Statistical Bulletin,
40, 1–11.
Moore, P.A., Duncan, G.H., Scott, D.S., Gregg, J.M., & Ghia, J.N. (1978). The submaximal effort tourniquet test:
its use in evaluating experimental and chronic pain. Pain, 6, 375–382.
Orbach, I., Palgi, Y., Stein, D., Har-Even, D., Lotem-Peleg, M., Asherov, J., & Elizur, A. (1996). Tolerance for
physical pain in suicidal patients. Death Studies, 20, 327–341.
Procacci, P. (1979). Methods for the study of pain threshold in man. In J.J. Bonica, J.C. Leibskind, & D.G. AlbeFessard (Eds.), Advances in pain research and therapy (vol. 3) (pp. 781–790). New York: Raven Press.
Raymond, N.C., Eckert, E.D., Hamalainen, M., Evanson, D., Thuras, P.D., & Faris, P.L. (1999). A preliminary
report on pain thresholds in bulimia nervosa during a bulimic episode. Comprehensive Psychiatry, 40,
229–233.
Schreiber, S., Stein, D., & Floman, Y. (1996). Psychologic approaches to the treatment of chronic low back
pain. In S.W. Wiesel, J.N. Weinstein, H. Herkowitz, J. Dvorak, & G. Bell (Eds.), The lumbar spine (2nd ed.)
(pp. 1018–1041). Philadelphia: WB Saunders Co.
Steinberg, S., Tobin, D., & Johnson, C. (1990). The role of bulimic behaviors in affect regulation: different
functions for different patient subgroups? International Journal of Eating Disorders, 9, 51–55.
Strober, M. (1987). The Eating Disorders Family History Interview. Los Angeles: University of California.