BRIEF REPORT
Altered Brain Activity in Women Recovered from
Bulimic-Type Eating Disorders after a Glucose
Challenge: A Pilot Study
Guido K. Frank, MD1,2
Angela Wagner, MD1
Sarah Achenbach, BSc1
Claire McConaha, BSN1
Kellie Skovira, BA1
Howard Aizenstein, MD1
Cameron S. Carter, MD3
Walter H. Kaye, MD1*
ABSTRACT
Objective: It is not known whether individuals with bulimic-type eating disorders
have a dysregulation of brain pathways
that modulate appetite. Taste plays a
role in the regulation of appetite and the
purpose of the current study was to determine whether bulimic women have
alterations in the physiologic response to
the blind administration of glucose.
Method: To avoid the confounding
effects of a current eating disorder, and
to assess possibly trait-related disturbances, we studied 10 subjects recovered
(1 year) from a bulimic-type eating disorder and 6 control women. Subjects
were administered a solution of glucose
or artificial saliva (control solution) in
alternating blocks during a functional
magnet resonance imaging scan.
Results: Individuals who recovered
from a bulimic-type eating disorder had
significantly lower activation in the right
Introduction
Bulimic eating patterns are characterized by episodes of overeating and compensatory purging
Accepted 17 April 2005
Presented at the Annual Meeting of the Society of Biological
Psychiatry, New York, April 2004.
Supported by a grant entitled ‘‘The Evaluation of Brain
Neuronal Activity in Response to Taste Stimuli: A Pilot Study and
Method for Taste Activation Assessed by fMRI’’ (to GKF) from the
Obesity/Nutrition Research Center, Pittsburgh, PA, by the
National Institute of Mental Health (NIMH) Training Grant T32MH18399 (to GKF), and by NIMH (WK PI) R01 MH42984 and
MH46001 and K05 MH018940.
*Correspondence to: Walter H. Kaye, MD, University of
Pittsburgh Western Psychiatric Institute and Clinic, Iroquois
Building, Suite 600, 3811 O’Hara Street, Pittsburgh, PA 15213.
E-mail: kayewh@upmc.edu
1
Department of Psychiatry, School of Medicine, University of
Pittsburgh, Western Psychiatric Institute and Clinic, Pittsburgh,
Pennsylvania
2
Department of Psychiatry, University of California, San Diego,
La Jolla, California
3
Department of Psychiatry and Behavioral Sciences, Imaging
Center/UDCMC, University of California, Davis, Davis, California
Published online 27 October 2005 in Wiley InterScience
(www.interscience.wiley.com). DOI: 10.1002/eat.20210
ª 2005 Wiley Periodicals, Inc.
76
anterior cingulate cortex (ACC; Montreal
Neurological Institute [MNI] coordinates x
¼ 8, y ¼ 22, z ¼ 28; cluster size ¼ 18
voxels, T ¼ 5.11, Z-score ¼ 3.78) and in
the left cuneus (occipital cortex; MNI
coordinates x ¼ 12, y ¼ 78, z ¼ 10;
cluster size ¼ 21 voxels, T ¼ 4.27, Z-score
¼ 3.36), when glucose was compared
with artificial saliva.
Conclusion: The ACC plays a role in the
anticipation of reward. Individuals with
bulimic-type eating disorders may have
a reduced reward response to nutrients,
and thus may be vulnerable to overeating. However, this is a small sample and
the current study will need replication in
a larger sample size with investigation of
additional regions of interest. ª 2005 by
Wiley Periodicals, Inc.
Keywords: brain pathways; appetite;
bulimic-type eating disorder; overeating;
(Int J Eat Disord 2006; 39:76–79)
behavior, which may consist of self-induced vomiting and the use of laxatives or diuretics. Such eating
behaviors occur in subjects with bulimia nervosa
(BN) as well as in subjects with the binge/purge
subtype of anorexia nervosa (AN-B/P). These eating disorders (EDs) are cross-transmitted in
families1 and share similar behavioral symptoms,2
suggesting that they are related.
It is not known whether individuals with BN have
a primary disturbance of appetite regulation. The
modulation of appetite is complex and involves
smell, taste, conditioned learning behavior, various
neurotransmitter systems, and peripheral as well as
central neurotransmission.3,4 Altered taste perception has been found in women with BN, who preferred sweeter stimuli compared with women
without an ED.5,6 Other studies found that altered
hedonic response to sucrose may be related to
conditioned thoughts about weight gain,7 and that
taste response in BN may extend to other nutrient
solutions such as fat preparations.8 Recent imaging
studies suggest alterations of frontal, temporal, and
cingulate brain regions in individuals with BN and
AN-B/P.9–11 The orbital frontal cortex, anterior
Int J Eat Disord 39:1 76–79 2006
ALTERED BRAIN ACTIVITY
cingulate cortex (ACC), anteromedial temporal cortex, and insula contribute to the modulation of taste
and other aspects of feeding, such as the hedonic
or rewarding aspects of food consumption.12–14
Recently, we developed a method for the assessment of brain activity in response to taste stimuli
using functional magnet resonance imaging (fMRI)
scanning.15 In that methods study, we examined
nonfasted control women (CW). The current study
was conducted to test the possibility that BN subjects have a disturbance of appetitive modulation as
reflected in altered processing of glucose (GLU) in
regions that might play a role in the hedonic or
rewarding aspects of feeding. Because all our other
imaging studies in EDs have a standardized design
with subjects fasted overnight, we changed the original fMRI design and studied fasted subjects.
To avoid the confounding effects of the extremes of
feeding behavior and malnutrition, our group has
investigated women who had recovered for 1 year
from an ED. Some, but not all studies, showed that a
disturbance of serotonin activity persists after recovery from an ED.16 Moreover, anxiety, perfectionism,
and obsessionality, as well as altered attitudes toward
food and body image, have been found to occur premorbidly, and persist after recovery from an ED.16
Together, these data raise the possibility that vulnerabilities, such as altered appetite regulation, may be
traits that contribute to the risk for developing an ED.
Methods
Subjects recovered from a bulimic-type ED (RBN; 7 BN
without a history of AN and 3 BN with a history of AN
[AN-B/P]) were compared with 6 CW. Criteria for the
recovered state and CW have been described previously.9
Subjects were studied during the first 10 days of the
follicular phase of the menstrual cycle. The study was
approved by the local institutional review board, and
written informed consent was obtained from all subjects.
Subjects were fasted overnight before the study. The
methods for taste activation using a customized design
and fMRI have been described previously.15 In brief, a customized pump delivered taste stimuli during fMRI scanning.
GLU was contrasted with artificial saliva (AS; control solution). 1 molar GLU solution was chosen to be able to compare results with studies from the Rolls group.17 fMRI
scanning was performed on a 3T MR scanner. Each taste
stimulus was followed by 6 T2* weighted functional scans
(TE 7, TR 2000, flip angle 70 ; reverse spiral acquisition).
Statistical parametric mapping (SPM2, London, UK; http://
www.fil.ion.ucl.ac.uk/spm/spm2.html) was used for image
analysis. Images were realigned, coregistered (12-parameter
Int J Eat Disord 39:1 76–79 2006
affine transformation) in Montreal Neurological Institute
(MNI) space, and smoothed (8 mm isotropic Gaussian kernel). Data were analyzed in an event-related design. For
each subject, contrast images were generated for GLU >
AS, and then used for a second-level, random-effects regression analysis (independent sample t test). The statistical
maps were thresholded at a significance level of p .001
(uncorrected for multiple comparisons), 7-voxel minimum
cluster size to correct for multiple comparisons.18 The SPSS
statistical software package19 was used for demographic
statistical analyses. Due to the relatively small sample size,
between-group comparisons were made with nonparametric Mann–Whitney U two-independent-samples tests
calculating two-sided exact significance levels.
This research was reviewed and approved by an institutional review board.
Results
The RBN subjects (27.5 ± 7 years) and CW (22 ± 2
years) were of similar ages (Mann–Whitney U ¼
18.5; p ¼ .2). Current body mass index (BMI) was
22 ± 3 for RBN and 21 ± 2 for CW (Mann–Whitney
U ¼ 19; p ¼ .6). The length of recovery was 12–180
months (45 ± 60 months). Within groups (data not
shown), the CW had greater activation (GLU > AS)
in the right ACC and posterior cingulate cortex,
whereas RBN subjects had greater activation in
the right lateral prefrontal cortex, subcallosal
gyrus, and lingual gyrus. Between-group analysis
(GLU > AS contrast images; Figure 1) showed that
the CW had significantly higher activation compared with RBN in the right ACC (MNI coordinates
x ¼ 8, y ¼ 22, z ¼ 28; cluster size ¼ 18 voxels, T ¼
5.11, Z-score ¼ 3.78), as well as in the left cuneus
(occipital cortex; MNI coordinates x ¼ 12, y ¼
78, z ¼ 10; cluster size ¼ 21 voxels, T ¼ 4.27, Zscore ¼ 3.36). There was no area with higher activation in RBNs compared with CW. When contrasting subjects with BN without a history of AN (n ¼ 7)
with the CW, the ACC with the similar MNI coordinates (x ¼ 8, y ¼ 22, z ¼ 28) was significant at p .01 (uncorrected for multiple comparisons).
Conclusion
To our knowledge, this is the first imaging study to
investigate whether individuals with BN have an
altered brain response to GLU. We found that
women recovered from BN had reduced activity,
compared with CW, in response to GLU, in the
ACC and cuneus. Other imaging studies have
77
FRANK ET AL.
FIGURE 1. Areas of significant greater activation in control women (CW) compared with recovered bulimic women
(RBN) in the SPM format. (A) Glass brain with activation in the anterior cingulate cortex (ACC; Montreal Neurological
Institute [MNI] coordinates x = 8, y = 22, z = 28; cluster size = 18 voxels, T = 5.11, Z-score = 3.78) and in the cuneus
(MNI coordinates x = 12, y = 78, z = 10; cluster size = 21 voxels, T = 4.27, Z-score = 3.36), at p .001,
uncorrected. (B) ACC activation overlaid on a reference image.
found altered serotonergic activity in similar cingulate regions in BN subjects.11,20 Moreover, Uher et
al.21 reported that pictures of food increased activity in the ACC and other regions in BN subjects
compared with CW.
The ACC plays an important role in initiation,
motivation, and goal-directed behaviors,22 as well
as in the anticipation of reward.23 It is possible that
altered ACC activity could result in a disturbance of
taste reward expectancy in individuals with BN.
Thus, they may overeat or seek to eat sweeter
foods. Consistently, a previous study found that
BN subjects prefer sweeter stimuli compared with
CW.5,6 All subjects reported the GLU solution as
sweet and pleasant. This pilot study did not include
a formal questionnaire for taste preference. Thus, it
is not known whether the altered brain response in
recovered BN subjects correlated with subjective
pleasantness ratings. This question is currently
being addressed in our ongoing studies.
Little information exists on taste perception after
recovery from an ED.24 It is not known whether
alterations after recovery were present before the
onset of BN, were a ‘‘scar’’ caused by the illness, or
were an adaptive process related to recovery. Taste
experience is strongly influenced not only by biologic processes, but also by conditioned processes.25 This is also the case for BN subjects.7 In
addition, altered taste response has been related to
altered perceptional physiology of the tongue in ill
BN subjects.26 It is therefore possible that any of
these factors could be implicated in the altered
brain response in this group of RBN subjects.
Those questions are targets for future research in
our studies. Moreover, taste is only a part of appetite modulation3,4 and more complex studies that
78
discern biologic from cognitive factors will help to
understand pathologic eating behavior in BN.
To our knowledge, altered cuneus activation has
not been reported in BN. This area may have functions that extend beyond visual activity including
borderline personality,27 which shares behavioral
traits with BN. We did not inquire about mood
states in this pilot study. An exploratory analysis
showed a relation between bulimia symptom
scores, as assessed by the Eating Disorders Inventory,28 and the ventrolateral striatum (MNI coordinates 22, 16, 6, cluster size ¼ 15 voxels, p ¼
.005). Others have found binge eating to be associated with striatal activation.29
A previous study found regional cerebral blood
flow normal in RBN compared with CW,30 suggesting this was not a contributing factor. In addition,
data from our laboratory (data not shown), using
positron emission tomography, suggested that
there was similar resting blood flow between recovered pure BN and anorexic-bulimic subjects. There
was an age difference between groups that was not
significant. An exploratory analysis of an age relation with the outcome variable (contrast image) did
not show any age relation for the studied groups.
The cluster size in the current study was small
compared with studies that stimulated emotional
reactions.21 However, only a subset of neurons may
respond to reward expectancy.23 In addition, expecting or receiving a food reward, or their combination,
may have different activation patterns.31
We did not find differences in orbitofrontal activation within or between subject groups. Recent
studies,32 as well as our preliminary data, suggest
that orbitofrontal activation may be particularly
related to sensory-specific satiety or to subjects’
Int J Eat Disord 39:1 76–79 2006
ALTERED BRAIN ACTIVITY
fasting status. That is, we found orbitofrontal activation in nonfasted subjects in our previous study
that described the methods for the current investigation,15 as well as in a second cohort of five nonfasted CW (unpublished data), but pronounced
cingulate activation in fasted CW in the current
study. Thus, the status of food ingestion before
fMRI studies may have important influence on
results.32
A limitation of this pilot study is its small sample
size. A larger sample will be needed to replicate this
finding as well as to investigate more regions of
interest. This may help us to understand the pathways of central taste transmission in BN.
In summary, the current study suggests lower
activity in the ACC and cuneus in response to a
GLU challenge in RBN compared with CW. This
raises the possibility that women with BN may
have alterations of mechanisms that model the
taste reward aspect of food. The fact that this is
found after recovery supports the speculation that
this may be a trait-related disturbance.
The authors thank Eva Gerardi for her editorial help with
preparation of the article, and the staff at the University
of Pittsburgh functional magnetic resonance imaging
facility for their help with conducting the study.
References
1. Kaye WH, Greeno CG, Moss H, et al. Alterations in serotonin
activity and psychiatric symptoms after recovery from bulimia nervosa. Arch Gen Psychiatry 1998;55:927.
2. Garner DM, Garfinkel PE, O’Shaughnessy M. The validity of
the distinction between bulimia with and without anorexia
nervosa. Am J Psychiatry 1985;142:581.
3. Blundell J, Lawton C, Hill A. Mechanisms of appetite control
and their abnormalities in obese patients. Horm Res 1993;
39:Suppl 3:72.
4. Rolls ET. Smell, taste, texture, and temperature multimodal
representations in the brain, and their relevance to the control
of appetite. Nutr Rev 2004;62:11 Pt 2:S193; discussion S224.
5. Drewnowski A, Halmi KA, Pierce B, et al. Taste and eating
disorders. Am J Clin Nutr 1987;46:442.
6. Franko DL, Wolfe BE, Jimerson DC. Elevated sweet taste
pleasantness ratings in bulimia nervosa. Physiol Behav
1994;56: 969.
7. Eiber R, Berlin I, de Brettes B, et al. Hedonic response to
sucrose solutions and the fear of weight gain in patients with
eating disorders. Psychiatry Res 2002;113:173.
8. Sunday SR, Halmi KA. Taste perceptions and hedonics in
eating disorders. Physiol Behav 1990;48:587.
9. Frank G, Bailer U, Henry S, et al. Neuroimaging studies in
eating disorders. CNS Spectr 2004;9:539.
10. Kaye WH, Frank GK, Meltzer CC, et al. Altered serotonin 2A
receptor activity in women who have recovered from bulimia
nervosa. Am J Psychiatry 2001;158:1152.
Int J Eat Disord 39:1 76–79 2006
11. Tiihonen J, Keski-Rahkonen A, Lopponen M, et al. Brain
serotonin 1A receptor binding in bulimia nervosa. Biol Psychiatry 2004;55:871.
12. O’Doherty J, Rolls E, Francis S, et al. Representation of pleasant and aversive taste in the human brain. J Neurophysiol
2001;85:1315.
13. Small DM, Zald DH, Jones-Gotman M, et al. Human cortical
gustatory areas: a review of functional neuroimaging data.
Neuroreport 1999;10:7.
14. Small DM, Voss J, Mak YE, et al. Experience-dependent neural
integration of taste and smell in the human brain. J Neurophysiol 2004;92:1892.
15. Frank G, Kaye W, Carter C, et al. The evaluation of brain
activity in response to taste stimuli-a pilot study and method
for central taste activation as assessed by event-related fMRI.
J Neurosci Methods 2003;131:99.
16. Kaye W, Strober M, Jimerson D. The neurobiology of eating
disorders. In: Nestler DSCaEJ, editor. Neurobiology of mental
illness. New York: Oxford Press; 2004, p 1112.
17. Francis S, Rolls E, Bowtell R, et al. The representation of
pleasant touch in the brain and its relationship with taste
and olfactory areas. Neuroreport 1999;10:453.
18. Forman S, Cohen J, Fitzgerald M, et al. Improved assessment
of significant activation in functional magnetic resonance
imaging (fMRI): use of a cluster-size threshold. Magn Reson
Med 1995;33:636.
19. SPSS I. SPSS for windows. Release 11.0.1. Chicago: SPSS; 2001.
20. Bailer U, Price JC, Meltzer CC, et al. Altered 5-HT2A receptor
activity after recovery from bulimia-type anorexia nervosa:
relationships to harm avoidance and drive for thinness.
Neuropsychopharmacology 2004;29:1143.
21. Uher R, Murphy T, Brammer M, et al. Medial prefrontal
cortex activity associated with symptom provocation in eating disorders. Am J Psychiatry 2004;161:1238.
22. Devinsky O, Morrell M, Vogt B. Contributions of anterior
cingulate cortex to behaviour. Brain 1995;118:279.
23. Richmond B, Liu Z, Shidara M. Neuroscience. Predicting
future rewards. Science 2003;301:179.
24. Sunday S, Halmi K. Taste perceptions and hedonics in eating
disorders. Physiol Behav 1990;48:587.
25. McClure S, Li J, Tomlin D, et al. Neural correlates of behavioral preference for culturally familiar drinks. Neuron 2004;
44:379.
26. Rodin J, Bartoshuk L, Peterson C, et al. Bulimia and taste:
possible interactions. J Abnorm Psychol 1990;99:32.
27. Juengling F, Schmahl C, Hesslinger B, et al. Positron emission
tomography in female patients with borderline personality
disorder. J Psychiatr Res 2003;37:109.
28. Garner D. Eating Disorder Inventory-2 professional manual.
Odessa, FL: Psychological Assessment Resources; 1991.
29. Inoue K, Kiriike N, Okuno M, et al. Prefrontal and striatal
dopamine metabolism during enhanced rebound hyperphagia induced by space restriction—a rat model of binge eating. Biol Psychiatry 1998;44:1329.
30. Frank GK, Kaye WH, Greer P, et al. Regional cerebral blood
flow after recovery from bulimia nervosa. Psychiatry Res
2000;100:31.
31. O’Doherty JP, Deichmann R, Critchley HD, et al. Neural
responses during anticipation of a primary taste reward.
Neuron 2002;33:815.
32. Kringelbach ML, O’Doherty J, Rolls ET, et al. Activation of the
human orbitofrontal cortex to a liquid food stimulus is
correlated with its subjective pleasantness. Cereb Cortex
2003;13:1064.
79