Regional cerebral blood flow during exposure to food in obese binge eating women

Ž .Psychiatry Research: Neuroimaging Section 99 2000 29]42

Regional cerebral blood flow during exposure to foodin obese binge eating women

Leila J. Karhunena,U, Esko J. Vanninenb, Jyrki T. Kuikkab, Raimo I.Lappalainena,c, Jari Tiihonend, Matti I.J. Uusitupaa

aDepartment of Clinical Nutrition, Uni ersity of Kuopio, P.O. Box 1627, 70211 Kuopio, FinlandbDepartment of Clinical Physiology and Nuclear Medicine, Kuopio Uni ersity Hospital, P.O. Box 1777, 70211 Kuopio,

FinlandcDepartment of Psychology, 33014 Uni ersity of Tampere, Tampere, Finland

dDepartment of Forensic Psychiatry, Niu¨anniemi Hospital, 70240 Kuopio, Finland

Received 16 September 1999; received in revised form 10 April 2000; accepted 13 April 2000

Abstract

Cerebral responses elicited by the sight of food were evaluated in eight obese binge eating, 11 obese and 12Ž .normal-weight non-binge eating women. Regional cerebral blood flow rCBF was mapped while the subjects were

Ž . Ž .looking at a picture of a landscape control or at a portion of food food exposure , and was measured byw99m xTc ethyl-cysteine-dimer and single photon emission computed tomography. Exposure to food was associated with

Ž .different changes in the cerebral blood flow normalized to mean cerebellar counts of the right and left hemispheresin the obese binge eating than in the obese or normal-weight non-binge eating women. As compared with thenon-binge eating groups, the obese binge eating women had, due to food exposure, a greater increase in the cerebralblood flow in the left than right hemisphere, especially in the frontal and pre-frontal regions. In addition, stronglinear correlations were observed in this group between the rCBF of the left frontal and pre-frontal regions and theincrease in the feeling of hunger during the exposure to food. The left hemisphere and its frontal and pre-frontalregions could thus play a role in binge eating behavior in humans. Q 2000 Elsevier Science Ireland Ltd. All rightsreserved.

Keywords: Eating disorders; Cerebrovascular circulation; Tomography, emission-computed, single-photon; Hunger

U Corresponding author. Tel. q358-17-162776; fax: q358-17-162792.Ž .E-mail address: [email protected] L.J. Karhunen .

0925-4927r00r$ - see front matter Q 2000 Elsevier Science Ireland Ltd. All rights reserved.Ž .PII: S 0 9 2 5 - 4 9 2 7 0 0 0 0 0 5 3 - 6

( )L.J. Karhunen et al. r Psychiatry Research: Neuroimaging Section 99 2000 29]4230

1. Introduction

To understand eating and related behavior, themechanisms by which eating is regulated have tobe understood. So far, only limited information isavailable concerning the cerebral functions re-lated to the regulation of eating behavior in hu-mans. Especially, the mechanisms involved in the

Žinitiation of eating are poorly understood Geisel-.man, 1996 . Nevertheless, based on studies in

experimental animals, there is evidence for antici-patory neural activity associated with stimuli pre-dicting food. Populations of neurons have beenidentified in the hypothalamus, amygdala and or-bito-frontal cortex which respond to the sight of

Ž .food in non-human primates Rolls, 1994 . Inaddition, the projections from the hypothalamusto the brainstem autonomic motor nuclei providea route by which food-related stimuli could gener-ate the conditioned autonomic responses, such as

Ž .salivation and insulin release Rolls, 1994 . Thestriatum provides a route for learned stimuli toinfluence behavioral responses by bringingtogether information from many parts of the lim-

Ž .bic system and the cerebral cortex Rolls, 1994 .The cortical areas are expected to contribute to

Žthe control of eating behavior in humans Booth,.1994 .

Functional brain imaging methods have made itpossible to examine the cerebral functions in liv-ing human brain. Using single photon emission

Ž .computed tomography SPECT , we recentlyfound increased regional cerebral blood flowŽ .rCBF in the right parietal and right temporalcortices during exposure to food in obese but not

Ž .in normal-weight women Karhunen et al., 1997 .The cerebral functions of eating disordered hu-mans, i.e. normal- or under-weight bulimic oranorexia nervosa patients, have also been ex-amined, although mainly in situations with no

Žfood-related activity e.g. Wu et al., 1990; Andrea-.son et al., 1992; Delvenne et al., 1995 . However,

to our knowledge, the cerebral functions of obesebinge eating humans have not been examinedbefore. Binge eating is the most common form ofthe eating patterns disorders among obese sub-

jects. It has been estimated that approximately30% of obese individuals who seek treatment fortheir obesity suffer from problems with binge

Ž .eating Spitzer et al., 1992 . Accordingly, the cri-teria for binge eating disorder have been addedas a new research diagnosis in the DSM IVŽ .American Psychiatric Association, 1994 . Basedon these criteria, binge eating disorder is charac-terized by recurrent episodes of eating unusuallylarge amounts of food with a sense of loss ofcontrol and marked distress of eating, but withoutany compensatory behavior. The underlyingpathophysiological mechanisms of binge eating

Žare, however, still poorly known Walsh and.Devlin, 1998 . Therefore, the aim of the present

study is to investigate the cerebral responses ofobese binge eating subjects elicited in response toexposure to food.

2. Methods

2.1. Subjects

Ž .Eight obese binge eating women OB-BEDw Ž .age 36.1"9.3 years mean"S.D. , body mass

2 xindex 35.2"5.0 kgrm participated in the study.The OB-BED subjects were recruited into thestudy from the participants of weight-reductionprograms carried out in the University of Kuopio,Kuopio University Hospital and local health cen-ters, and with the aid of an advertisement in alocal newspaper. Binge eating episodes were as-sessed using the self-report questionnaires BITEŽ .Henderson and Freeman, 1987 , Binge Eating

Ž . Ž .Scale BES Gormally et al., 1982 and the Ques-Ž .tionnaire of Eating and Weight Patterns QEWP

Ž .Spitzer et al., 1993 . To be classified as a bingeeater, a subject was required to have at least a

Žtotal combined score of 25 in the BITE Hender-. Žson and Freeman, 1987 , 20 in the BES Gormally

. Žet al., 1982 and 6 in the QEWP i.e. binge eating.disorder; Spitzer et al., 1993 . In addition, the

presence of binge eating was confirmed by a

( )L.J. Karhunen et al. r Psychiatry Research: Neuroimaging Section 99 2000 29]42 31

Table 1Individual scores obtained from the self-report questionnaires assessing the presence of binge eating behavior in obese binge eatingsubjects

aStudy BITE Binge Eating Questionnaire of Eatingb c,dŽ .subjects symptomrseverity Scale and Weight Patterns

1 27r8 41 62 25r9 37 63 22r4 30 64 23r7 37 65 22r4 37 66 26r3 36 67 28r10 38 68 24r6 29 6

a Ž .Henderson and Freeman 1987 .b Ž .Gormally et al. 1982 .c Ž .Spitzer et al. 1993 .dScores confirmed by clinical interview.

clinical interview that was based on the questionsof the QEWP. Thus, as compared with the DSM-

ŽIV criteria of binge eating disorder American.Psychiatric Association, 1994 , all subjects also

fulfilled these criteria. The individual scores ob-tained from the self-report questionnaires arepresented in Table 1.

Ž . ŽIn addition, 11 obese OB-non-BED age 45.02 ."10.0 years, body mass index 32.7"4.0 kgrm

Ž . Žand 12 normal-weight NW age 39.8"9.7 years,2 .body mass index 22.2"1.6 kgrm women with

no specific eating disorder who had participatedin our previous study with an identical study

Ž .design Karhunen et al., 1997 were included intothe present study as comparison groups. Theobese non-binge eating subjects had been re-cruited among the participants of the weight-reduction program carried out at the University

Ž .of Kuopio Uusitupa et al., 1996 . At the time ofthe study they had all completed the active weightloss period of the program. The normal-weightsubjects were recruited from the university staff.

All subjects in these three groups were right-handed. Based on the clinical and psychiatrichistory and routine laboratory tests, they did nothave any other disorders or medications known toaffect the variables examined in the present study.Subjects gave their informed consent for partici-pation in the study. The study was approved by

the Ethics Committee of the University of Kuopioand Kuopio University Hospital, and was in ac-cordance with the Helsinki declaration.

2.2. Procedure

The study design was the same for all subjects.The study consisted of two experiments: the con-trol and the food exposure. Based on the currentthinking of Pavlovian conditioning, i.e. that condi-tioned associations are formed not just betweenthe primary events presented, but with the whole

Žcontext in which they are presented Rescorla,.1988 , all subjects experienced the control experi-

ment first, in order to avoid conditioned anticipa-tory responses during the control experiment. Forthe same reason, the subjects were not told inadvance that food would be a part of the experi-ment.

The experiments were performed after anovernight fast on two separate days between 09.00and 12.00 h. The mean interval between the con-trol and the food-exposure experiments was ap-

Ž .proximately 9"7 days range 3]25 days in theŽ .OB-BED, 21"21 days range 6]62 days in the

Ž .OB-non-BED, and 20"23 days range 3]72 daysŽ .in the NW group Karhunen et al., 1997 . Both

the control and food-exposure experiments con-sisted of two parts: a baseline and exposure to a


Ž . Ž .Fig. 1. Study designs of the a control and b food-exposure experiments. S, blood sampling; T, injection of tracer; SPECT, singlephoton emission computed tomography scan; H and D, assessments of feelings of hunger and desire to eat, respectively. The

Ž .measurements are presented in the figure on a time scale, expressed in minutes min .

Ž .stimulus Fig. 1 . The rCBF was mapped at thebeginning of the exposure period while the sub-ject was sitting in front of a table and was lookingat a control or a food stimulus at a distance of 50cm. During the injection of the tracer, the subjectwas advised to remain silent and relaxed, and toconcentrate on looking at the stimulus. Bloodsamples were taken at 2-min intervals during thebaseline and exposure periods. In addition, thefeelings of hunger and desire to eat were assessedthree times during the food-exposure experiment:Ž .1 just before the beginning of the food exposurewhen the subject did not yet know that she wasgoing to be exposed to food in a little while, andŽ . Ž2 at the beginning i.e. 2 min after the injection

. Ž . Žof the tracer and 3 at the end i.e. approx. 13.min after the injection of the tracer of the food-

exposure period. Feelings of hunger and desire toeat were not assessed during the control experi-ment in order to keep it totally free from foodstimuli. At the end of both experiments, the sub-ject rated the pleasantness of the experiment and

Ž .the food she had eaten cf. food stimulus . Inaddition, she filled in the Beck Depression Inven-

Ž . Ž .tory BDI Beck et al., 1961 to assess the pres-ence and severity of symptoms of depression.

2.3. Control stimulus

ŽA colorful picture of a landscape 41 cm wide.and 31.5 cm high was used as the stimulus in the

control experiment. It was placed on the table infront of the subject just before the injection ofthe tracer. The control exposure and the baselineperiods of both experiments were performed in aquiet room free from food-related cues.

2.4. Food stimulus

A warm, freshly cooked lunch, consisting of thesubject’s self-selected warm main course, salad,bread, beverage, and dessert, was used as a stimu-lus in the food-exposure experiment. The food-exposure period was performed in a kitchen, adja-cent to the room where the baseline period hadtaken place. The food was on the table in front ofthe subject during the injection of the tracer andfor the whole duration of the food-exposure pe-riod. To confirm that the subject really liked thefood, and had a prior experience of eating it inthat particular situation, the subject was allowedto choose and eat a similar lunch after the controlexperiment. She was, however, not told then that

Ž .the lunch would be a part i.e. the food stimulus


of the second experiment. The subject also atethe food during the food-exposure experiment,but only after the food-exposure period.

2.5. Cerebral responses

SPECT was used to assess the cerebral respon-w99m xses. A dose of 550 MBq of Tc ethyl-cysteine-

Ždimer Neurolite, DuPont PharmarDurham APS,.Kastrup, Denmark was injected into the subject’s

antecubital vein to determine rCBF. Thew99m xTc ethyl-cysteine-dimer was used as a tracerbecause of its greater specificity for measuringrCBF and better radio-chemical stability than theother commonly used tracer, hexamethyl-propan-

Ž . Žediamine HMPAO Slosman and Magistretti,.1997 . The SPECT scan was carried out 45 min

after the injection with a three-headed SiemensMultiSPECT 3 gamma camera equipped with fan

Žbeam collimators Siemens Medical Systems,. Ž .Hoffman Estates, IL, USA Kuikka et al. 1993 .

The subject’s head was positioned similarly inboth scans using positioning lasers in a head-holder specifically built for Siemens MultiSPECT3. A total of 5]7 million counts were acquired forthe entire head using an angular step of 38 over

Ž .3608 matrix size 128=128 and 120 projections .The imaging resolution was 7]8 mm. The radia-tion load on the subject from the two experiments

Žwas moderate, being 6]8 mSv effective dose.equivalent .

ŽThree-millimeter-thick trans-axial oriented in.the orbito-meatal line , sagittal and coronal slices

were reconstructed after the Chang attenuationcorrection with the uniform attenuation coeffi-cient of 0.1rcm. In addition, three-dimensionalsurface shaded plots were used for illustrative

Ž .purposes Fig. 2 . Two consecutive trans-axialslices were summated in order to obtain a slicethickness of 6 mm and these were saved onto ahard disk for further analysis.

A semi-automatic brain quantification programfrom Siemens was used to analyze the cerebralregions of interest: pre-frontal, frontal, temporal,parietal and occipital cortex, thalamus, hypothala-

Žmus, basal ganglia, cerebellum and amygdala mayalso include some other parts of the temporal

. Ž .lobe, e.g. hippocampus Karhunen et al., 1997 .

w99m xFig. 2. An example of three-dimensional Tc ECD SPECTŽ .surface shaded images. Images from the top control and

Ž .bottom food-exposure experiments in a 26-year-old bingeeating woman. Note increased uptake in the left frontal and

Žpre-frontal cortex during the food-exposure experiment bot-.tom .

First, the slices were rotated and realigned so thatŽ . Ž .trans-axial x-direction , sagittal y-direction and

Ž .coronal z-direction ones were at 908 angles toeach other. Secondly, the regions of interest were


drawn onto aligned trans-axial slices on the rightand then mirrored on the left. The average region-al counts were normalized to mean cerebellar

Ž .counts i.e. region rcerebellum . It was reasonedithat the normalization of the regional counts withthe mean cerebellar counts would take into ac-count uncontrollable non-specific stimulation, ifany, between the experiments, unrelated to themanipulated stimuli. The cerebellum was used asa reference region because the reactivity of thecerebellum to the sight of food or a picture wasexpected to be minimal and not to differ fromeach other.

2.6. Emotional responses

Feelings of hunger and desire to eat were de-termined with 10-cm visual analog scales rangingfrom absent to extreme, and the pleasantness ofthe experiment and of the lunch eaten with ver-bally anchored nine-point category scales. Thefeeling of hunger was defined as a general feelingwith no specific food item in mind, and the desireto eat as a specific desire to eat the exposed foodwhen looking at or thinking of it. The emotionalresponses were expressed as within-subjectchanges in emotions, since they were consideredas more objective determinants of emotional re-sponses than their absolute levels.

2.7. Peripheral physiological responses

Blood samples for the determinations of seruminsulin and plasma glucose concentrations weretaken five times during the baseline period and

Ž .five times during the exposure period Fig. 1 . Theserum leptin and cortisol determinations weremade from two samples taken during the baselineperiod and three samples taken during the expo-

Ž .sure period samples at 4-min intervals . Plasmanoradrenaline and adrenaline concentrations weredetermined from samples that were taken at theend of each period. All the samples were takenthrough an intravenous cannula inserted into thesubject’s antecubital vein 15 min before the firstsamples. The samples were placed in pre-chilledtubes, and centrifuged and stored without delayat y708C until analyzed.

Plasma glucose was measured by a glucoseŽoxidase method Glucose Auto & Stat, Model

.GA-110, Daiici, Kyoto, Japan . Radioimmunoas-say methods were used for the measurement of

Žserum insulin Phadeseph Insulin RIA 100, Phar-.macia Diagnostics, Uppsala, Sweden , leptin

Ž .Linco Research, St. Louis, MO and cortisolŽCortisol Radioimmunoassay, Orion Diagnostica,

.Espoo, Finland . Plasma noradrenaline andadrenaline concentrations were determined with

Ž .high pressure liquid chromatography HPLC .

2.8. Statistical analyses

Statistical analyses were performed with SPSSŽfor Windows statistical program SPSS Inc.,

.Chicago, IL, USA . The differences in the rCBF,and in the emotional and peripheral physiologicalvariables between andror within the control andfood-exposure experiments were assessed withanalysis of variance for repeated measurementsŽ .GLM . The hemisphere and experimental condi-tion were included in the analysis as the within-subjects factors and the group as the between-subjects factor when analyzing the differences inthe rCBF between the control and food-exposureexperiments. The experimental period and experi-mental condition were included into the analysisas the within-subjects factors and the group as thebetween-subjects factor when analyzing the dif-ferences in the peripheral physiological variablesbetween and within the experiments, and theexperimental period as the within-subjects factorand the group as the between-subjects factor whenanalyzing the differences in the emotional vari-ables within the experiment. Bonferroni correc-tions were made when multiple comparisons wereperformed between the groups. The Spearman

Ž .rank correlation coefficient two-tailed was usedto analyze the relationship of the emotional andperipheral physiological responses or the BDIscores to the rCBF in the cerebral regions withsignificant differences between the control andfood-exposure experiments. A value of PF0.05was used as the criterion for statistical signifi-cance.


Table 2Ž .Experimental conditions in the control and food-exposure experiments mean"S.D.

Obese binge eating women Obese non-binge eating women Normal-weight womenŽ . Ž . Ž .ns8 ns11 ns12

Control Food exposure Control Food exposure Control Food exposure

Ž .Time of the experiment h 10.6"0.8 10.5"0.6 10.2"0.6 10.3"0.7 10.3"0.8 10.2"0.8Ž .Dose of the tracer MBq 545"17 554"16 554"12 558"15 546"39 559"17

Duration of the fasting 14.6"1.1 14.9"1.1 14.4"1.1 14.6"1.2 13.8"1.0 13.7"0.9Ž .period h

Pleasantness of the 6.0"1.4 6.0"1.7 7.0"0.8 6.4"0.9 6.8"1.5 6.1"1.2aexperiment

Pleasantness of the 8.3"0.9 7.9"0.8 8.5"0.5 8.5"0.7 8.3"0.8 7.8"1.1alunch

aA nine-point category scale.

3. Results

The control and food-exposure experimentswere carried out in comparable situations in all

Ž .three groups Table 2 . The average cerebellarŽ .blood flow results total counts were also compa-

Žrable OB-BED: control 164"51 countsrvoxel,food exposure 159"42 countsrvoxel; OB-non-BED: control 215"50 countsrvoxel, food expo-sure 208"42 countsrvoxel; NW: control 235"40countsrvoxel, food exposure 221"31 countsr

.voxel; GLM, Fs0.210, d.f.s2,28, Ps0.81 .The rCBF of the right and left hemispheres

changed differently among the OB-BED, OB-non-BED and NW groups between the control

Žand food-exposure experiments GLM, Fs3.643,.d.f.s2,28, Ps0.04 . In the OB-BED group, the

exposure to food was associated with the mostmarked increase in the blood flow in the lefthemisphere, whereas there were no markedchanges or the increase was localized in the righthemisphere in the NW and OB-non-BED groups,

Ž .respectively Fig. 3 . The differences among theŽgroups were seen especially in the frontal GLM,

.Fs5.340, d.f.s2,28, Ps0.01 and pre-frontalŽ .GLM, Fs4.307, d.f.s2,28, Ps0.02 cortices,where the hemispheric rCBF of the OB-BEDsubjects differed significantly from that of the

ŽOB-non-BED frontal: t-test, tsy3.204, two-tailed, d.f.s102, Ps0.006; pre-frontal: t-test, t

.sy2.354, two-tailed, d.f.s102, Ps 0.05 andŽNW subjects frontal: t-test, tsy2.964, two-

tailed, d.f.s102, Ps0.01; pre-frontal: t-test, ts. wy2.542, two-tailed, d.f.s102, Ps0.03 signifi-

cances of contrasts were multiplied by numberŽ .x Ž .contrasts Bonferroni statistics Table 3, Fig. 2 .

Other cerebral regions did not show this differ-Ž .ence Table 3 .

The feeling of hunger increased significantly,but not differently, during the food-exposure ex-

Žperiment in all three groups OB-BED: GLM,Fs23.56, d.f.s2,14, P-0.001; OB-non-BED:GLM, F s 10.69, d.f.s 2,18, P s 0.001; NW:

.GLM, Fs11.96, d.f.s2,22, P-0.001; Table 4 .So did the desire to eat in the OB-non-BEDŽ .GLM, Fs17.21, d.f.s2,18, P-0.001 and NWŽ .GLM, Fs6.24, d.f.s2,22, Ps0.007 groupsŽ .Table 4 . Furthermore, in the OB-BED groupthe increase in the feeling of hunger during the

Žfood exposure i.e. the change from the beginning.of the food-exposure period to the end of it was

associated with higher rCBF in the left frontalŽSpearman correlation coefficient, two-tailed, rs

. Ž0.88, ns8, Ps0.004 and pre-frontal Spearmancorrelation coefficient, two-tailed, rs0.88, ns8,

. Ž .Ps0.004 cortices Fig. 4 . No such associationwas observed in the OB-non-BED or NW groups,either in the feeling of hunger or desire to eat.

ŽBDI scores OB-BED 17.1"5.2, OB-non-BED.9.3"8.0, NW 2.4"3.4 were not associated with

the rCBFs in these two cerebral regions in eithergroup.

There were no significant differences in thechanges of the serum insulin, leptin, cortisol,


Fig. 3. Changes in cerebral blood flow from the control to the food-exposure experiment in the right and left hemispheres in theŽ . Ž .obese binge eating OB-BED , obese non-binge eating OB-non-BED and normal-weight subjects. Difference among the groups:

analysis of variance for repeated measurements, Fs3.643, d.f.s2,28, Ps0.04, Mean"S.E.

plasma glucose, noradrenaline or adrenaline con-centrations during the control or food-exposure

Ž .experiments among the three groups Table 4 .

4. Discussion

To our knowledge, this is the first time thatcerebral functions in obese binge eating humanshave been examined. In order to examine thepathophysiological mechanisms contributing toaberrant eating behavior more specifically, thestudy was performed in a food-exposure situation.As a result, significant differences due to theexposure to food were observed between the obesebinge eating and the obese and normal-weightnon-binge eating women in hemispheric bloodflow in frontal and pre-frontal regions. Interest-ingly, in contrast to that we had observed previ-ously in the obese non-binge eating women, whoshowed an increase in the rCBF in the right

Žhemisphere during food exposure Karhunen et.al., 1997 , in the obese binge eating women the

food exposure was associated with the mostmarked increase in the rCBF in the opposite, lefthemisphere.

There is some previous evidence supporting theview that normal cerebral asymmetry could be

altered in patients with eating disorders. Normal-weight bulimic subjects have shown left-greater-than-right hemispheric asymmetry in several cor-tical regions and not the normal right-greater-than-left asymmetry in, e.g., the frontal regionswhen assessed in a situation with no food-related

Ž .activity Wu et al., 1990; Andreason et al., 1992 .Similarly, although before eating, the bulimicpatients had greater blood flow in the right sideof the frontal, parietal and occipital cortices, aftereating the blood flow was greater in the left side

Ž .of all cortical regions Nozoe et al., 1995 .Anorexia, in turn, has been suggested to be aright-hemispheric disorder; the profile most fre-quently associated with anorexia nervosa is theright posterior hypo- and right anterior hyper-metabolism associated with right-graded abnor-

Žmal electroencephalogram spiking Braun and.Chouinard, 1992 . The balance of cerebral hemi-

sphere activity could thus play a role in bingeeating behavior in humans, although it could alsodepend on the binge eating patient’s eating stateŽ . Ži.e. anorexic or binge eating phase Hirano et

.al., 1999 .The mechanism contributing to the differential

cortical sensitivity of the two hemispheres is notknown, but it has been suggested that the frontal


Tab

le3

Ž.

Ž.

Reg

iona

lcer

ebra

lbl

ood

flow

rCB

Fno

rmal

ized

toth

erC

BF

ofth

ece

rebe

llum

inth

eco

ntro

lan

dfo

od-e

xpos

ure

expe

rim

ents

mea

n"

S.D

.

aŽ

.Ž

.C

ereb

ralr

egio

nO

bese

bing

eea

ting

wom

enn

s8

Obe

seno

n-bi

nge

eatin

gw

omen

Nor

mal

-wei

ghtw

omen

ns

12P

Ž.

ns

11

Con

trol

Foo

dex

posu

reC

ontr

olF

ood

expo

sure

Con

trol

Foo

dex

posu

re

Pre-

fron

tal,

righ

t1.

02"

0.10

1.01

"0.

120.

93"

0.06

0.94

"0.

080.

91"

0.05

0.93

"0.

100.

01Pr

e-fr

onta

l,le

ft0.

96"

0.08

0.99

"0.

100.

90"

0.06

0.88

"0.

070.

89"

0.06

0.90

"0.

10F

ront

al,r

ight

1.09

"0.

111.

07"

0.09

0.99

"0.

050.

99"

0.06

1.00

"0.

051.

01"

0.07

0.02

Fro

ntal

,lef

t1.

04"

0.09

1.06

"0.

090.

96"

0.04

0.96

"0.

050.

96"

0.04

0.97

"0.

08T

empo

ral,

righ

t1.

10"

0.07

1.10

"0.

090.

99"

0.04

1.02

"0.

061.

04"

0.06

1.05

"0.

080.

81T

empo

ral,

left

1.06

"0.

061.

07"

0.07

0.97

"0.

041.

00"

0.09

1.03

"0.

061.

03"

0.09

Pari

etal

,rig

ht1.

11"

0.07

1.09

"0.

090.

93"

0.05

0.97

"0.

061.

05"

0.07

1.05

"0.

110.

11Pa

riet

al,l

eft

1.08

"0.

061.

09"

0.06

0.93

"0.

050.

94"

0.08

1.06

"0.

081.

07"

0.13

Occ

ipita

l,ri

ght

1.27

"0.

061.

33"

0.10

1.21

"0.

081.

24"

0.10

1.23

"0.

091.

22"

0.12

0.89

Occ

ipita

l,le

ft1.

22"

0.13

1.29

"0.

071.

16"

0.10

1.19

"0.

111.

18"

0.09

1.19

"0.

10T

hala

mus

,rig

ht1.

26"

0.08

1.28

"0.

071.

14"

0.08

1.17

"0.

081.

15"

0.07

1.16

"0.

070.

63T

hala

mus

,lef

t1.

27"

0.11

1.28

"0.

061.

11"

0.09

1.11

"0.

081.

16"

0.06

1.17

"0.

11B

asal

gang

lia,r

ight

1.30

"0.

111.

29"

0.09

1.16

"0.

041.

18"

0.06

1.17

"0.

051.

18"

0.10

0.40

Bas

alga

nglia

,lef

t1.

29"

0.08

1.30

"0.

111.

16"

0.06

1.18

"0.

051.

19"

0.06

1.20

"0.

09A

myg

dala

,rig

ht0.

92"

0.12

0.89

"0.

080.

88"

0.07

0.89

"0.

090.

84"

0.06

0.84

"0.

080.

53A

myg

dala

,lef

t0.

87"

0.11

0.90

"0.

120.

88"

0.07

0.88

"0.

080.

82"

0.08

0.83

"0.

10b

Hyp

otha

lam

us0.

92"

0.10

0.92

"0.

100.

88"

0.09

0.92

"0.

070.

90"

0.09

0.91

"0.

100.

63

aŽ

.A

naly

sis

ofva

rian

cefo

rre

peat

edm

easu

rem

ents

GL

Mw

ithth

ehe

mis

pher

ean

dex

peri

men

tal

cond

ition

asth

ew

ithin

-sub

ject

sfa

ctor

san

dth

egr

oup

asth

ebe

twee

n-su

bjec

tsfa

ctor

,int

erac

tion:

hem

isph

ere

byex

peri

men

tal

cond

ition

bygr

oup.

bG

LM

with

expe

rim

enta

lco

nditi

onas

the

with

in-s

ubje

cts

fact

oran

dth

egr

oup

asth

ebe

twee

n-su

bjec

tsfa

ctor

,int

erac

tion:

expe

rim

enta

lco

nditi

onby

grou

p.

( )L.J. Karhunen et al. r Psychiatry Research: Neuroimaging Section 99 2000 29]4238T

able

4Ž

.E

mot

iona

lan

dpe

riph

eral

phys

iolo

gica

lres

pons

esdu

ring

the

base

line

and

expo

sure

peri

ods

ofth

eco

ntro

lan

dfo

od-e

xpos

ure

expe

rim

ents

mea

n"

S.D

.

bc

Ž.

Obe

sebi

nge

eatin

gw

omen

Obe

seno

n-bi

nge

eatin

gw

omen

Nor

mal

-wei

ghtw

omen

ns

12P

aŽ

.Ž

.n

s8

ns

11

Exp

erim

enta

lC

ontr

olF

ood

expo

sure

Con

trol

Foo

dex

posu

reC

ontr

olF

ood

expo

sure

cond

ition

Exp

erim

enta

lB

asel

iner

Bas

elin

erB

asel

iner

Bas

elin

erB

asel

iner

Bas

elin

erpe

riod

expo

sure

expo

sure

expo

sure

expo

sure

expo

sure

expo

sure

gF

eelin

gof

hung

er,

Not

done

1.7"

1.6r

Not

done

3.9"

2.6r

Not

done

4.0"

2.4r

0.51

de

ee

Ž.

VA

Scm

4.1"

2.8

5.8"

2.2

6.1"

2.8

ff

f6.

4"3.

27.

3"2.

47.

1"2.

4g

Des

ire

toea

t,N

otdo

ne4.

9"3.

7rN

otdo

ne4.

2"2.

5rN

otdo

ne5.

4"2.

4r0.

45d

ee

eŽ

.V

AS

cm6.

0"3.

36.

7"2.

27.

1"2.

2f

ff

7.0"

3.1

8.1"

1.9

7.8"

2.2

Seru

min

sulin

93.8

"91

.3r

117.

4"10

5.5r

55.5

"24

.3r

55.6

"38

.8r

38.4

"14

.5r

35.2

"8.

9r0.

08h

Ž.

pmol

rl10

2.5"

91.7

114.

6"96

.951

.9"

22.4

56.4

"45

.935

.4"

12.6

33.3

"9.

8Se

rum

lept

in34

.6"

17.8

r38

.2"

18.6

r27

.7"

14.5

r26

.0"

17.0

r8.

8"4.

5r7.

6"3.

1r0.

08i

Ž.

ngrm

l34

.8"

18.4

35.5

"16

.928

.2"

16.3

25.3

"15

.78.

5"4.

57.

5"3.

1Se

rum

cort

isol

336.

9"14

6.5r

279.

9"81

.9r

270.

0"52

.3r

251.

7"65

.1r

325.

9"12

7.1r

344.

5"13

7.2r

0.78

iŽ

.nm

olrl

318.

5"20

8.9

275.

6"91

.525

8.6"

106.

322

4.6"

63.5

291.

1"10

0.2

308.

1"11

8.9

Plas

ma

gluc

ose

5.3"

0.4r

5.3"

0.3r

5.6"

0.4r

5.4"

0.6r

5.0"

0.3r

5.0"

0.3r

0.55

hŽ

.m

mol

rl5.

3"0.

65.

3"0.

45.

6"0.

45.

4"0.

64.

9"0.

45.

0"0.

2Pl

asm

ano

r-3.

07"

1.55

r3.

03"

1.05

r2.

35"

0.89

r2.

05"

0.63

r2.

81"

1.22

r2.

58"

1.04

r1.

00Ž

.ad

rena

line

nmol

rl3.

29"

1.01

3.35

"1.

152.

67"

0.55

2.48

"0.

843.

09"

1.24

2.94

"1.

34Pl

asm

aad

rena

line

0.58

"0.

80r

0.25

"0.

06r

0.20

"0.

08r

0.19

"0.

06r

0.37

"0.

09r

0.32

"0.

08r

0.22

Ž.

nmol

rl0.

29"

0.12

0.26

"0.

080.

22"

0.12

0.19

"0.

070.

25"

0.07

0.30

"0.

12

aF

eelin

gof

hung

er,d

esir

eto

eat,

seru

min

sulin

,lep

tin,c

ortis

olan

dpl

asm

agl

ucos

e,n

s10

.bPl

asm

agl

ucos

e,n

s11

;pl

asm

aad

rena

line,

ns

7.c

Ž.

Ana

lysi

sof

vari

ance

for

repe

ated

mea

sure

men

tsG

LM

with

the

expe

rim

enta

lper

iod

and

expe

rim

enta

lcon

ditio

nas

the

with

in-s

ubje

cts

fact

ors,

and

the

grou

pas

the

betw

een-

subj

ects

fact

or,i

nter

actio

n:ex

peri

men

tal

peri

odby

expe

rim

enta

lco

nditi

onby

grou

p.d

Ž.

A10

-cm

visu

alan

alog

uesc

ale

VA

S.

eT

hebe

ginn

ing

ofth

efo

od-e

xpos

ure

peri

od.

f The

end

ofth

efo

od-e

xpos

ure

peri

od.

gG

LM

,the

expe

rim

enta

lpe

riod

asth

ew

ithin

-sub

ject

sfa

ctor

,the

grou

pas

the

betw

een-

subj

ects

fact

or,i

nter

actio

n:ex

peri

men

tal

peri

odby

grou

p.h

Val

ues

are

mea

nsof

five

conc

entr

atio

nsdu

ring

each

peri

od.

i Val

ues

are

mea

nsof

two

conc

entr

atio

nsdu

ring

the

base

line

and

thre

edu

ring

the

expo

sure

peri

ods.


Ž . Ž . Ž .Fig. 4. Correlation of the regional cerebral blood flow rCBF of the a left frontal and b left pre-frontal cortex with the change inŽ .the feeling of hunger during the food exposure in the obese binge-eating women OB-BED . VAS, visual analog scale.

regions of the two hemispheres may be differen-tially sensitive to upstream influences from the

Ž .limbic system Jacobs and Snyder, 1996 . Thedifferential cortical sensitivity could also be afunction of an asymmetrical distribution of neu-rotransmitter mechanisms. Accordingly, it hasbeen suggested that the normally asymmetric

Ž .serotonin system Arato et al., 1991 could be´rendered more imbalanced in eating disordersŽ .Regard and Landis, 1997 . The role of the sero-tonergic system in the pathogenesis of humanbinge eating behavior was supported in a recentstudy showing abnormalities in the serotonergicsystem persisted after recovery from bulimia ner-

Ž .vosa Kaye et al., 1998 . Our preliminary observa-tions on obese binge eating women also support

Ž .this possibility Kuikka et al., 2000 .In the present study, the differences in hemi-

spheric asymmetry due to food exposure wereseen especially in the frontal and pre-frontal re-gions. Although the meaning of this finding is notknown, there is some previous evidence suggest-ing a role for these cerebral regions in the regula-tion of eating behavior. Populations of neurons inthe frontal regions, especially in the orbito-frontalcortex, have been shown to respond to the sight

Ž .of food in non-human primates Rolls, 1997 .

Rapid learning of visual-to-taste association isalso a feature of the neural processing occurring

Ž .in the orbito-frontal area Rolls, 1997 . In addi-tion, reward ‘expectancy’ related to, e.g., the pre-sentation of food has been shown to be processed

Ž .in the pre-frontal cortex Watanabe, 1996 .Human studies have also shown increased rela-tive activation of the left hemisphere anteriorregions being associated with heightened ap-proach-related positive affect, decreased with-

Ždrawal-related negative affect or both Jacobs.and Snyder, 1996 .

On the other hand, the frontal lobe is thecerebral region that is influenced by cognitive

Ž .functions related to eating Nozoe et al., 1995 .Accordingly, a feeding-suppression mechanism

Žhas been suggested to be located there Nozoe et.al., 1995 . The frontal regions have also been

postulated to exert suppressing effects on theother cerebral regions, such as hypothalamus, that

Žrespond to food-related stimuli Rolls, 1994;.Tataranni et al., 1999 . In the present study the

rCBF was examined before eating when the sub-jects were looking at the food, knowing that theycould eat the food after the measurements. Theobserved changes in the rCBF of the frontal andpre-frontal regions due to the food exposure could


thus also be associated with the cognitive func-tions related to eating, possibly in relation to thebinge eaters’ problematic relationship with food.In addition, as the frontal lobe is the corticalregion having the most intimate connections with

Ž .the arousal system Nauta, 1971 , the enhancedrCBF in the frontal region during the food expo-sure could reflect different arousal levels elicitedduring the food- and non-food-exposure situa-tions, especially when a preoccupation with foodand eating is one characteristic feature of binge

Ž .eating subjects Bruce and Wilfley, 1996 .As a further support for the present findings,

strong positive correlations were observed in theobese binge eating women between the increasein the feeling of hunger and the rCBFs of the leftfrontal and pre-frontal regions during the expo-sure to food. Studies in non-human primates haveshown the orbito-frontal cortex to be the cerebralregion whose responses to visual representationsof food are modulated by an animal’s hunger

Ž .status Critchley and Rolls, 1996 . Frontal regionshave also been shown to be associated with hungerand satiation in healthy men: hunger with in-creased rCBF in the posterior orbito-frontal cor-tex and satiation with increased rCBF in theventromedial and dorsolateral pre-frontal corticesŽ .Tataranni et al., 1999 . Although we did notseparately analyze the orbito-frontal regionhere, our pre-frontal region includes this areaŽ .Karhunen et al., 1997 . In line with the previousfindings, the present observations thus point to arole for the pre-frontal, and thereby also for theorbito-frontal region, in food-related behaviors inhumans. Nevertheless, whether these regions areinvolved in the mediation of hunger responses bythemselves, or whether they rather reflect a cog-nitive response to increased feelings of hunger,cannot be determined on the basis of the presentstudy. Instead, as similar associations were notobserved with the desire to eat, different cerebralregions seem to be involved in processing theseemotional responses.

Some cerebral metabolic abnormalitiesobserved in the patients with eating disordershave been associated with depression whereassome others seem to be independent of the mood

Ž .state Andreason et al., 1992 . In the present

study, the depressive symptoms, as assessed byŽ .the BDI Beck et al., 1961 , were not associated

with rCBF in the frontal or pre-frontal regions.The observed differences in the asymmetry of thehemispheric blood flow between the binge andnon-binge eating subjects could thus be suggestedto be associated with the core features of eatingbehavior, rather than with depression.

Both the control and food-exposure experi-ments consisted of visual stimulation. The percep-tual integration of a visual stimulus is a multi-stage process with extensive interactions betweenthe visual pathways at almost all cortical levelsŽ .Kandel, 1995 . It could thus be suggested thatbesides the effect purely produced by food-related stimulation, visual stimulation could alsohave activated deeply complicated neural net-works in the cortices. Therefore, although nodifferences were observed among the binge andnon-binge eating subjects in the rCBF of theprimary visual regions, we cannot rule out thepossibility that there could have been some otherdifferences in visual processing among the groupsthat could have contributed to the present obser-vations. To clarify this is, however, a question foranother study.

In conclusion, exposure to food elicits differentchanges in hemispheric blood flow in obese bingeeating than in obese or normal-weight non-bingeeating women. This is especially seen in frontaland pre-frontal regions of the left hemisphereshowing greater increases in blood flow due toexposure to food. The rCBF of left frontal andpre-frontal cortices is also associated with an en-hanced feeling of hunger during food exposure.Left hemisphere and its frontal and pre-frontalregions could thus play a role in binge eatingbehavior in humans. The significance of this find-ing in the regulation of human eating behaviorremains to be examined in further studies.

Acknowledgements

The authors thank Ms Kaija Kettunen and thepersonnel of the Department of Clinical Physi-ology and Nuclear Medicine for their excellenttechnical assistance, and Pirjo Halonen, M.Sc.,


and Veikko Jokela, M.Sc., for their valuable helpin performing statistical analyses. The study wassupported by a grant from F. Hoffmann-La RocheLtd. to the University of Kuopio, and grants fromKuopio University Hospital, the Finnish CulturalFoundation, and the Academy of Finland, Re-search Council for Health.

References

American Psychiatric Association, 1994. DSM-IV: Diagnosticand Statistical Manual of Mental Disorders, 4th ed. Ameri-can Psychiatric Press, Washington, D.C.

Andreason, P.J., Altemus, M., Zametkin, A.J., King, A.C.,Lucinio, J., Cohen, R.M., 1992. Regional cerebral glucosemetabolism in bulimia nervosa. American Journal of Psy-chiatry 149, 1506]1513.

Arato, M., Frecska, E., Duncan, J., MacCrimmon, D.J., Dus-´cott, R., Saxena, B., Tekes, K., Tothfalusi, L., 1991. Seroto-tonergic interhemispheric asymmetry: neurochemical andpharmaco-EEG evidence. Progress in Neuropsychopharma-cology and Biological Psychiatry 15, 759]764.

Beck, A.T., Ward, C.G., Mendelson, M., Mock, T., Erbaugh,T., 1961. An inventory for measuring depression. Archivesof General Psychology 4, 561]571.

Booth, D.A., 1994. Psychology of Nutrition. Taylor & Francis,London.

Braun, C.M.J., Chouinard, M.J., 1992. Is anorexia nervosa aneuropsychologic disease? Neuropsychological Reviews 3,171]212.

Bruce, B., Wilfley, D., 1996. Binge eating among the over-weight population: a serious and prevalent problem. Jour-nal of the American Dietetic Association 96, 58]61.

Critchley, H.D., Rolls, E.T., 1996. Hunger and satiety modifythe responses of olfactory and visual neurons in the pri-mate orbitofrontal cortex. Journal of Neurophysiology 75,1673]1686.

Delvenne, V., Lotstra, F., Goldman, S., Biver, F., De Maerte-laer, V., Appelboom-Fondu, J., Schoutens, A., Bidaut, L.M.,Luxen, A., Mendlewicz, J., 1995. Brain hypometabolism ofglucose in anorexia nervosa: a PET scan study. BiologicalPsychiatry 37, 161]169.

Geiselman, P.J., 1996. Control of food intake: a physiologicallycomplex, motivated behavioral system. Endocrinology andMetabolism Clinics of North America 25, 815]829.

Gormally, J., Black, S., Daston, S., Rardin, D., 1982. Theassessment of binge eating severity among obese persons.Addictive Behaviors 7, 47]55.

Henderson, M., Freeman, C.P.L., 1987. A self-rating scale forbulimia. The ‘BITE’. British Journal of Psychiatry 150,18]24.

Hirano, H., Tomura, N., Okane, K., Watarai, J., Tashiro, T.,1999. Changes in cerebral blood flow in bulimia nervosa.Journal of Computer Assisted Tomography 23, 280]282.

Jacobs, G.D., Snyder, D., 1996. Frontal brain asymmetry pre-dicts affective style in men. Behavioral Neuroscience 110,3]6.

Kandel, E., 1995. Construction of the visual image. In: Kandel,Ž .E., Schwartz, J.H., Jessell, T.M. Eds. , Essentials of Neural

Science and Behavior. Appleton & Lange, New York, pp.387]405.

Karhunen, L.J., Lappalainen, R., Vanninen, E., Kuikka, J.T.,Uusitupa, M.I.J., 1997. Regional cerebral blood flow duringfood-exposure in obese women. Brain 120, 1675]1684.

Kaye, W.H., Greeno, C.G., Moss, H., Fernstrom, J., Fern-strom, M., Lilenfeld, L.R., Weltzin, T.E., Mann, J.J., 1998.Alterations in serotonin activity and psychiatric symptomsafter recovery from bulimia nervosa. Archives of GeneralPsychiatry 55, 927]935.

Kuikka, J.T., Tenhunen-Eskelinen, M., Jurvelin, J., Kiiliainen,¨H., 1993. Physical performance of the Siemens Multi-SPECT 3 gamma camera. Nuclear Medicine Communica-tions 14, 490]497.

Kuikka, J.T., Tammela, L., Rissanen, A., Bergstrom, K.,¨Muhonen, M., Vanninen, E., Karhunen, L., Karhu, J.,Tiihonen, J., Uusitupa, M., 2000. Altered serotonintransporter density in obese binge eating women: initialfindings. European Journal of Nuclear Medicine 27, A31.

Nauta, W.J.H., 1971. The problem of the frontal lobe: areinterpretation. Journal of Psychiatric Research 8,167]187.

Nozoe, S.I., Naruo, T., Yonekura, R., Nakabeppu, Y., Soejima,Y., Nagai, N., Nakajo, M., Tanaka, H., 1995. Comparison ofregional cerebral blood flow in patients with eating dis-orders. Brain Research Bulletin 36, 251]255.

Regard, M., Landis, T., 1997. ‘Gourmand syndrome’: eatingpassion associated with right anterior lesions. Neurology48, 1185]1190.

Rescorla, R.A., 1988. Pavlovian conditioning. It’s not what youthink it is. American Psychologist 43, 151]160.

Rolls, E.T., 1994. Neural processing related to feeding inŽ .primates. In: Legg, C.R., Booth, D. Eds. , Appetite: Neural

and Behavioural Bases. Oxford University Press, Oxford,pp. 11]53.

Rolls, E.T., 1997. Taste and olfactory processing in the brainand its relation to the control of eating. Critical Reviews inNeurobiology 11, 263]287.

Slosman, D.O., Magistretti, P., 1997. Biological basis of cere-bral perfusion SPECT. In: De Deyn, P.P., Dierckx, R.A.,

Ž .Alavi, A., Pickut, B.A. Eds. , SPECT in Neurology andPsychiatry. John Libbey, London, pp. 513]519.

Spitzer, R.L., Devlin, M., Walsh, B.T., Hasin, D., Wing, R.,Marcus, M., Stunkard, A., Wadden, T., Yanovski, S., Agras,S., Mitchell, J., Nonas, C., 1992. Binge eating disorder: amultisite trial for the diagnostic criteria. InternationalJournal of Eating Disorders 11, 191]203.

Spitzer, R.L., Yanovski, S., Wadden, T., Wing, R., Marcus,M.D., Stunkard, A., Devlin, M., Mitchell, J., Hasin, D.,Horne, R.L., 1993. Binge eating disorder: its further valida-tion in a multisite study. International Journal of EatingDisorders 13, 137]153.


Tataranni, P.A., Gautier, J.F., Chen, K., Uecker, A., Bandy,D., Salbe, A.D., Pratley, R.E., Lawson, M, Reiman, E.M.,Ravussin, E., 1999. Neuroanatomical correlates of hungerand satiation in humans using positron emission tomogra-phy. Proceedings of the National Academy of Sciences ofthe USA 96, 4569]4574.

Uusitupa, M.I.J., Karhunen, L., Rissanen, A., Franssila-Kal-lunki, A., Niskanen, L., Kervinen, K., Kesaniemi, Y.A.,¨1996. Apolipoprotein E phenotype modifies metabolic andhemodynamic abnormalities related to central obesity inwomen. American Journal of Clinical Nutrition 64, 131]136.

Walsh, B.T., Devlin, M.J., 1998. Eating disorders: progress andproblems. Science 280, 1387]1390.

Watanabe, M., 1996. Reward expectancy in primate prefrontalneurons. Nature 382, 629]632.

Wu, J.C., Hagman, J., Buchsbaum, M.S., Blinder, B., Derrfler,M., Tai, W.Y., Hazlett, E., Sicotte, N., 1990. Greater leftcerebral hemispheric metabolism in bulimia assessed bypositron emission tomography. American Journal of Psy-chiatry 147, 309]312.

Regional cerebral blood flow during exposure to food in obese binge eating women

Documents