iir

gsscrfcsaC

R

Dysfunctional Reward Circuitry inObsessive-Compulsive DisorderMartijn Figee, Matthijs Vink, Femke de Geus, Nienke Vulink, Dick J. Veltman, Herman Westenberg, andDamiaan Denys

Background: Obsessive-compulsive disorder (OCD) is primarily conceived as an anxiety disorder but has features resembling addictivebehavior. Patients with OCD may develop dependency upon compulsive behaviors because of the rewarding effects following reduction ofobsession-induced anxiety. Reward processing is critically dependent on ventral striatal-orbitofrontal circuitry and brain imaging studies inOCD have consistently shown abnormal activation within this circuitry. This is the first functional imaging study to investigate explicitlyreward circuitry in OCD.

Methods: Brain activity during reward anticipation and receipt was compared between 18 OCD patients and 19 healthy control subjects,using a monetary incentive delay task and functional magnetic resonance imaging. Reward processing was compared between OCDpatients with predominantly contamination fear and patients with predominantly high-risk assessment.

Results: Obsessive-compulsive disorder patients showed attenuated reward anticipation activity in the nucleus accumbens comparedwith healthy control subjects. Reduced activity of the nucleus accumbens was more pronounced in OCD patients with contamination fearthan in patients with high-risk assessment. Brain activity during reward receipt was similar between patients and control subjects. A hinttoward more dysfunctional reward processing was found in treatment-resistant OCD patients who subsequently were successfully treatedwith deep brain stimulation of the nucleus accumbens.

Conclusions: Obsessive-compulsive disorder patients may be less able to make beneficial choices because of altered nucleus accumbensactivation when anticipating rewards. This finding supports the conceptualization of OCD as a disorder of reward processing and behavioral

addiction.

idrjpcs(ns1ipi

rtpwrsorwsaclbpntr

Key Words: Behavioral addiction, deep brain stimulation, imag-ing, nucleus accumbens, obsessive-compulsive disorder, rewardprocessing

O bsessive-compulsive disorder (OCD) is a chronic and dis-abling disease with an estimated prevalence between 1%and 3% (1,2). Obsessive-compulsive disorder is character-

zed by the presence of recurrent and anxiety-provoking thoughts,mages, or impulses (obsessions), typically followed by repetitiveitualistic behaviors (compulsions) to relieve anxiety.

The anxiety symptoms and inadequate fear responses patho-nomonic for OCD may result from inadequate dorsal prefrontal-triatal control of the amygdala (3). However, obsessive-compulsiveymptoms are not always anxiety-driven and may also be related toognitive and behavioral inflexibility, reflected by impairments inesponse inhibition and attentional set shifting (4) because of dys-unctional frontal-striatal circuitry (5,6). Alternatively, OCD has beenonceptualized as a disorder of behavioral addiction, with obses-ions and compulsions being related to loss of voluntary controlnd a dependency on repetitious, self-defeating behavior (7–9).ompulsions could be viewed as addictive because of their reward-

From the Department of Psychiatry (MF, NV, DJV, DD), Academic MedicalCenter, Amsterdam; Department of Psychiatry (MV, FdG, HW), RudolfMagnus Institute of Neuroscience, University Medical Center Utrecht,Utrecht; and The Netherlands Institute for Neuroscience (DD), an Insti-tute of the Royal Netherlands Academy of Arts and Sciences, Amster-dam, The Netherlands.

Authors MF and MV contributed equally to this work.Address correspondence to Martijn Figee, M.D., University of Amsterdam,

Academic Medical Center, Department of Psychiatry, Meibergdreef 5,Postbox 75867, 1105 AZ Amsterdam, The Netherlands; E-mail: m.figee@amc.nl.

ceceived Jun 8, 2010; revised Dec 8, 2010; accepted Dec 8, 2010.

0006-3223/$36.00doi:10.1016/j.biopsych.2010.12.003

ng effects following reduction of obsession-induced anxiety. Ad-ictive behavior is associated with defective processing of natural

ewards. Likewise, OCD patients were found to be impaired in ad-usting their behavior following monetary incentives (10). Rewardrocessing is critically dependent on ventral striatal-orbitofrontalircuitry (11), and resting-state imaging studies have consistentlyhown abnormal metabolism in striatum and orbitofrontal cortexOFC) in OCD (12). Moreover, recent studies have shown that theucleus accumbens (NAcc), as part of the ventral striatum, is auccessful target for deep brain stimulation in OCD treatment (13–5). Studying reward processing and its neuroanatomical correlates

n OCD might therefore be a fruitful approach to unravel its patho-hysiology. However, there is surprisingly little research capitaliz-

ng on this idea.In the present study, we sought to investigate the neural basis of

eward system function in OCD by comparing brain activation be-ween OCD patients and healthy control subjects during rewardrocessing using a robust monetary incentive delay paradigm,hich allowed modeling of both reward anticipation and outcome

elative to neutral events. We employed a very rapid three-dimen-ional sequence for acquiring whole-brain functional magnetic res-nance images, thereby increasing sensitivity of our design. Prior

esearch in healthy humans showed distinct activation patternsithin the frontal-striatal network during reward anticipation ver-

us outcome: ventral striatum activation for reward anticipationnd OFC or ventromedial frontal cortex activation for reward out-ome (16,17). In OCD patients, obsessive-compulsive behaviors are

ikely to be associated with reward circuitry hyperactivity, reflectedy findings of increased OFC-striatal activity at rest and in symptomrovocation studies (12,18), at the expense of its responsiveness toatural rewards. Therefore, we expect to find decreased OFC-stria-

al activity during reward processing, more specifically, decreasedeward anticipatory activation in the nucleus accumbens and de-

reased activation of the OFC related to reward feedback. In addi-

BIOL PSYCHIATRY 2011;69:867–874© 2011 Society of Biological Psychiatry

DShottsar

R

botosstfcappdtwrrlsdtj

l(

868 BIOL PSYCHIATRY 2011;69:867–874 M. Figee et al.

w

tion, we aimed to explore brain activation patterns in two distinctOCD subdimensions. We recruited a group of OCD patients suffer-ing from obsessive fear of contamination and washing compulsionsand a group of OCD patients with obsessive high-risk assessmentand checking compulsions. Previous research suggests that OCDwith predominant contamination fear symptoms may be more re-lated to dysfunctional brain circuits for emotion processing (19).Because of its association with limbic regions, the reward systemmight thus be more dysfunctional in OCD patients with contamina-tion fear symptoms. In addition, symptom provocation studieshave consistently shown that disgust and OCD washing symptomsare related to hyperactivation of the insula (19,20), an area that isalso involved in processing of personally rewarding stimuli (21).Therefore, decreased reward responsiveness of the insula is ex-pected in contamination fear OCD.

Methods and Materials

SubjectsWe included 18 patients with a primary diagnosis of OCD (13

female patients; mean age 35 years) and 19 healthy control subjects(13 female control subjects; mean age 34 years) (Table 1). All sub-jects were right-handed. Patients were recruited from the outpa-tient clinic for anxiety disorders at our university hospital. All pa-tients consented to participate in this study and signed an informedconsent form. The study was approved by the Medical Ethical Re-view Committee of our hospital. Diagnosis of patients was con-firmed by the Mini International Neuropsychiatric Interview (22,23)according to DSM-IV criteria. Symptom severity was assessed usingthe Yale-Brown Obsessive-Compulsive Scale (Y-BOCS) (24,25). Weonly included patients predominantly suffering from contamina-tion obsessions with washing and cleaning compulsions on the onehand and patients suffering from high-risk assessment obsessions,such as fear for burglary, disasters, etc., with checking compulsionson the other hand. The Y-BOCS symptom checklist was used inaddition to a psychiatric interview with an experienced psychiatristto validate the presence of OCD subdimensions. To rate depressionand anxiety severity, we used Hamilton Rating Scales (26): Hamilton

Table 1. Demographic and Clinical Characteristics of 18 OCD Patients, 19 H

OCD (n � 18) Control Subjects (n � 1

Gender, Male/Female, Number 5/13 6/13

Age, Mean (SD), Years 34 (8.3) 32 (6.6)

Education, Mean (SD), Years 13 (2.37) 14.6 (1.9)

Illness Duration, Mean (SD), Years 19.7 (7.2)

Y-BOCS Obsessions, Mean (SD) 14.9 (3.6)

Y-BOCS Compulsions, Mean (SD) 15.2 (3.7)

Y-BOCS Total Score, Mean (SD) 29.1 (8.5)

HAM-A Score, Mean (SD) 17.4 (8.2)

HAM-D Score, Mean (SD) 16.8 (6.6)

Medication, Yes/No, Number 9/18

HAM-A, Hamilton Anxiety Rating Scale; HAM-D, Hamilton Depression Rating SCompulsive Scale.

ww.sobp.org/journal

epression Rating Scale (HAM-D) and Hamilton Anxiety Ratingcale (HAM-A). Exclusion criteria were the presence of severe alco-ol or substance abuse (including nicotine) and major neurologicalr other medical disorders. Nine patients were medication-free at

he time of investigation. For the remaining patients, use of variousypes and doses of psychotropic drugs were reported: selectiveerotonin reuptake inhibitors (n � 5, dose 20 – 60 mg), tricyclicntidepressants (n � 3; dose 125–225 mg), and combined norad-energic and serotonergic antidepressant (n � 1; dose 45 mg).

eward TaskThe monetary reward task (Supplement 1, see also 27) was

ased on the monetary incentive delay task (16). The task consistedf 72 trials, each lasting 6 seconds on average (range 3–10 sec). At

he beginning of each trial, a cue was presented for 500 millisec-nds signaling a potentially rewarding (a circle) or nonrewarding (aquare) trial. Following this cue, a target was presented to whichubjects had to respond. Feedback on performance was given (ei-her a reward or no reward). Subjects were instructed to respond asast as possible to the target (by pressing a button) irrespective ofue type. The target was displayed for a time period equal to theverage response speed of the subject during training trials. Tenractice trials were presented before the experiment. From theseractice data, the shortest reaction time to the target was used toetermine the individual time limit allowed for responses to the

arget during the task. That is, in case of a reward cue, subjects couldin €2 when they responded within the time limit. Subjects were

ewarded in only 50% of the reward trials, allowing for maximaleward uncertainty (28). This was achieved by increasing the timeimit by 200 milliseconds in half of the rewarding trials to make sureubjects would be fast enough to win the trial. The time limit wasecreased by 200 milliseconds in the other half of the rewarding

rials to make sure subjects would miss the reward. Hence, all sub-ects earned the same amount (€36).

The task was designed in such a way that the blood oxygenationevel-dependent (BOLD) signal in response to anticipation of rewardi.e., time between cue and target) could be modeled independently

y Control Subjects, and OCD Subtypes

Statistic Contamination Fear High-Risk Assessment Statistic

�2 � .064p � .800

2/5 3/8 �2 � .004p � .952

t � 1.01p � .371

40.0 (7.5) 30.6 (7.1) t � 2.69p � .016

t � �2.29p � .028

11.6 (2.4) 14 (1.8) t � 2.35p � .45

18 (6.2) 20.7 (7.9) t � .77p � .54

30.2 (11.7) 28.5 (6.5) t � .37p � .72

14.3 (7.4) 14.2 (3.9) t � .48p � .96

15.8 (4.5) 14.3 (3.1) t � .82p � .43

17.0 (8.5) 17.7 (8.7) t � .13p � .90

17.1 (6.9) 16.7 (7.6) t � .08p � .94

4/7 5/11 �21 � .059p � .808

ealth

9)

cale; OCD, obsessive-compulsive disorder; Y-BOCS, Yale-Brown Obsessive-

vIstcat

D

r(KiiwmFaglttTbnaffsab[1mngorfy

feKsi

R

D

pscws(cwdo

fTsn.o

B

fpcHiFgwsmm

N

Tiiwwjt(at(ttprdcpN

M. Figee et al. BIOL PSYCHIATRY 2011;69:867–874 869

from that of actual reward (i.e., feedback). This enabled us to differen-tiate between anticipation and outcome, which reportedly inducesBOLD responses in different parts of the reward network (29). To re-duce colinearity between anticipation of reward and actual reward,anticipation time and intertrial interval were varied (3–10 sec, mean 6sec; and 0–30 sec, mean 4.2 sec, respectively). Only one level of rewardwas used, and no loss trials were included, to obtain maximum powerwithin a relatively short time period.

Image AcquisitionBrain imaging data were collected on a 1.5T Philips ACS-NT

scanner (Philips Medical Systems, Best, The Netherlands) with fastgradients (PT6000). The head was held in place with a strap andpadding. Structural and functional images were acquired in trans-verse orientation from the same section of the brain. The area thatwas scanned ranged from roughly Z � �20 to Z � �60, tiltedsomewhat anterior. The primary motor cortex and primary visualcortex were not included. For functional scans, a navigated three-dimensional principle of echo shifting with a train of observationpulse sequence (30) was used with the following parameters: echotime 29.23 milliseconds; repetition time 19.23 milliseconds; flip an-gle 9°; matrix 40 � 64; 20 slices; field of view 160 � 256 � 80 mm;

oxel size 4 mm isotropic; scan duration 1 second per 20-slice volume.mmediately after the functional scans, an additional principle of echohifting with a train of observation scan of the same volume of brainissue was acquired with a high (30°) flip angle (FA30) for the imageo-registration routine. Finally, a T1-weighted structural image wascquired for anatomical registration purposes. A total of 1008 func-ional images were acquired for each subject.

ata Preprocessing and AnalysisPreprocessing and analysis of individual functional magnetic

esonance imaging time series data were performed with SPM2Wellcome Department of Imaging Neuroscience, London, Unitedingdom). First, all functional images were realigned to the FA30

mage. Next, the structural image was co-registered to the FA30mage. The structural image was then registered to the T1-

eighted Montreal Neurological Institute standard brain. The nor-alization parameters were then applied to all functional scans.

inally, all functional scans were smoothed with an 8 mm full-widtht half maximum Gaussian kernel. For each individual subject, re-ression coefficients for each voxel were obtained from a general

inear model regression analysis, using a factor matrix that con-ained factors representing event-related changes time-locked tohe anticipation of neutral (36 events) and reward (36 events) trials.he duration for these events was, on average, 6 seconds (varyingetween 3 and 10 seconds). The third factor modeled hemody-amic responses during responding to the target (72 events). Threedditional factors described the brain response during neutraleedback (36 events), feedback of rewarded targets (18 events), andeedback of missed targets (18 events). To correct for drifts in theignal, a high-pass filter with a cutoff frequency of .006 Hz waspplied to the data. The task was designed such that the overlapetween factors of interest was minimized (variance inflation factor

VIF] of neutral anticipation � 1.24; VIF of reward anticipation �.21; VIF of feedback of rewarded targets � 1.60; VIF of feedback ofissed targets � 1.70). Factors for all six conditions (anticipation of

eutral trials, anticipation of rewarding trials, response to the tar-et, neutral feedback, feedback of rewarded targets, and feedbackf missed targets) were convolved with a canonical hemodynamic

esponse function. Group activation maps were generated for eachactor using MULTISTAT (31). This constitutes a mixed-effects anal-

sis. Group results were tested for significance (p � .05, corrected (

or multiple comparisons) resulting in a critical t value of 4.5 forvery voxel. We used Gaussian random field theory as proposed byiebel et al. (32). In the patient group, additional correlation analy-es were performed between BOLD responses on reward process-ng and severity scores on the Y-BOCS, HAM-A, and HAM-D.

esults

emographic and Clinical CharacteristicsTable 1 summarizes demographic and clinical characteristics for

atients and healthy control subjects and for OCD subgroups. Notatistical significant differences were found between patients andontrol subjects for age and gender. The average years of educationere slightly lower for patients compared with healthy control

ubjects. Mean total Y-BOCS score for the patient group was 29.6SD 7.3), indicating severe OCD. Two patients were diagnosed withomorbid major depressive disorder, four patients were diagnosedith additional disorders on Axis 1 (social phobia, dysthymic disor-er, and hypochondria), and two patients were diagnosed withbsessive-compulsive personality disorder.

Seven patients had predominantly symptoms of contaminationear and 11 had high-risk assessment and checking symptoms.here was no statistically significant difference between symptomubtypes in gender, Y-BOCS, HAM-A, or HAM-D scores. There wereo differences between symptom subtypes in medication use (�2

1 �059, p � .808). The contamination fear group was significantlylder than the group with high-risk assessment symptoms.

ehaviorReaction times in anticipation of reward trials were significantly

aster than in neutral trials when no reward was expected in bothatients and control subjects. Patients were generally slower thanontrol subjects [408 vs. 349 msec; F (1,34) � 13.07, p � .001].owever, patients and control subjects both reacted faster in antic-

pation to reward compared with no reward [385 vs. 418 msec;(1,16) � 15.49, p � .001] and this did not differ between theroups (F � 1). Reaction times did not differ between OCD patientsith contamination fear and patients with high-risk assessment

ymptoms [426 vs. 390 msec; F (1,16) � 1.78, p � .200]. As wasodeled, patients and control subjects all won the same amount ofoney (€36).

euroimagingNeuroimaging results are presented in Figures 1, 2, and 3 and

able 2. For each group, a whole-brain analysis was performed todentify activity during reward anticipation and receipt by contrast-ng activity during anticipation or receipt of reward versus no re-

ard. Reward feedback was defined as the activation associatedith successful versus neutral responses. Both healthy control sub-

ects and OCD patients showed reward-related activity in the ven-ral striatum, putamen, thalamus, insula, and several frontal areasinferior, medial, and superior frontal gyrus), indicating that the taskctivated reward-related frontostriatal areas. Compared with con-rol subjects, OCD patients showed reduced activation of the NAccbilateral) during anticipation of monetary gain (Figure 2). In addi-ion, OCD patients showed less reward anticipation activation ofhe left insula. Reduced activation of the NAcc and insula was moreronounced in OCD with contamination fears compared with high-

isk assessment OCD patients. Blood oxygenation level-dependentifferences between patient groups remained significant after in-luding age as a covariate. Upon receiving rewarding feedback,atients and control subjects showed comparable activation of theAcc, OFC, and other regions belonging to the reward network

Figure 3).

www.sobp.org/journal

tc

D

wata

ole-b

870 BIOL PSYCHIATRY 2011;69:867–874 M. Figee et al.

w

Correlation AnalysesNo statistically significant correlations were found between

BOLD responses during reward anticipation and symptom severityratings on the Y-BOCS, HAM-D, and HAM-A in the patient group.However, post hoc regions of interest analyses revealed that thelargest reduction in NAcc activity during reward anticipation wasfound in a subgroup of nine patients. These patients suffered froman extremely severe, treatment-resistant form of OCD and weresubsequently treated with deep brain stimulation of the NAcc. Dif-ferences between the deep brain stimulation prone patients and

Figure 1. Brain activation during reward anticipation in healthy control subjhigh-risk assessment OCD versus patients with contamination fear OCD. Wh

Figure 2. Brain activation during reward outcome in healthy control subjects

assessment OCD vs. patients with contamination fear OCD. Reward outcome wasWhole-brain group results thresholded at p � .05 corrected.

ww.sobp.org/journal

he normal treatment responders did not reach statistical signifi-ance (p � .122), presumably because of the small sample size.

iscussion

This functional imaging study is unique in examining both re-ard anticipation and receipt in a sample of OCD patients. By

dapting a monetary incentive delay task, we were able to focus onhese different aspects of reward processing separately. Behavior-lly, all subjects reacted significantly faster when a reward was

nd patients with obsessive-compulsive disorder (OCD) and in patients withrain group results thresholded at p � .05 corrected.

bsessive-compulsive disorder (OCD) patients and in patients with high-risk

ects a

and o

defined as the activation associated with successful vs. neutral responses.

lcwbOvrowbsiofwirwa

mFrciifvttsdm

aha(s[ttectstccr

ivdbtrisesirrsdotpaidmrb

otaact

sbwata

M. Figee et al. BIOL PSYCHIATRY 2011;69:867–874 871

expected, but OCD patients reacted significantly slower thanmatched healthy control subjects in anticipation of rewards. Com-pared with healthy control subjects, OCD patients showed greatlyattenuated reward anticipation activity in the bilateral NAcc. Re-duced brain activity of the NAcc was more pronounced in OCDpatients with contamination fear, compared with patients withhigh-risk assessment symptoms. The difference between both pa-tient groups in the absence of differences in performance speedindicates that blunted NAcc activity is primarily accounted for bydefective reward anticipation and not by motor impairments. Re-ward outcome was related to activation in various regions of thebrain reward network, including the NAcc and the OFC, but this didnot differ between patients and control subjects.

Reward processing in OCD has only been sparsely investigatedpreviously. Nevertheless, Remijnse et al. (33) found in a reversal-earning task reduced OFC activity during monetary reward out-ome. Our study confirmed an association between monetary re-ard feedback and OFC activity but failed to detect a differenceetween patients and control subjects. It could be speculated thatCD patients in our study had primarily difficulties in estimating thealue of a potential rewarding situation because of reduced NAccesponsiveness, whereas subsequent appreciation and evaluationf rewards in the OFC was intact. Several other prefrontal areasere activated, especially during reward feedback. Studies by To-ler et al. (11), among others, have indicated that prefrontal areasubserve the individual evaluation of rewards, more specifically thentegration of expected value and risk. Obsessive-compulsive dis-rder patients who suffered predominantly from contamination

ear showed reduced ventral striatal activation when comparedith patients with predominantly high-risk assessment and check-

ng symptoms. This is in agreement with Rauch et al. (34), whoeported an inverse relationship between striatal activation andashing but not checking symptoms during implicit learning. Also,study by Mataix-Cols et al. (35) suggests that OCD with contami-

nation fear may be characterized by dysfunctional brain circuitsinvolved in emotion processing, whereas OCD with checking symp-toms might be associated with regions that are important for motorand attentional functions. The reward network, which is linked to

Figure 3. Explorative overview of brain activation during reward anticipa-tion � neutral anticipation in healthy control subjects (black bars), obses-ive-compulsive disorder (OCD) patients with high-risk assessment (grayars), and OCD patients with contamination fear (white bars). The regionsere defined using the contrast healthy control subjects � OCD patients

nd are therefore only intended to provide information concerning activa-ion levels in the OCD subgroups on an explorative basis. NAcc, nucleusccumbens.

limbic regions for emotion processing, may thus be more compro- a

ised in OCD with predominant contamination fear symptoms.urthermore, patients with contamination fear showed reducedeward anticipation activity of the left insula, a region that is impli-ated in emotion perception and processing of personally reward-

ng stimuli (21) but also in the integration of gustatory and olfactorynputs. Previous functional magnetic resonance imaging studiesound increased activation of the left insula in OCD patients wheniewing aversive pictures (35) and in OCD patients with contamina-ion fear when viewing pictures depicting washing or contamina-ion (20). The insula may thus be excessively activated during ob-essive-compulsive symptoms, particularly in association toisgust-related symptoms, compromising its recruitment for nor-al reward processing.

The NAcc, as part of the ventral striatum, has been implicated asbrain region that is critically involved in reward processing. In

ealthy humans, the ventral striatum is activated particularly innticipation of a reward and in proportion with its expected value11,29). Obsessive-compulsive disorder has been associated withtructural and functional abnormalities of the striatum (reviewed in18]). Within this circuit, the NAcc has been successfully used as aarget for deep brain stimulation in the treatment of severely ill,herapy-refractory OCD patients (13,15). Interestingly, in the pres-nt study, we found a hint toward more dysfunctional reward pro-essing in a subgroup of severely ill, treatment-resistant OCD pa-ients who subsequently were successfully treated with deep braintimulation of the NAcc. Together, our findings suggest an impor-ant role for the NAcc in the pathophysiology of OCD. Obsessive-ompulsive disorder patients may be less able to make beneficialhoices because of defective NAcc activation when anticipatingewards.

Our results match up remarkably with the findings of functionalmaging studies in addiction disorders. Blunted reactivity of theentral striatum during anticipation of monetary gain was found inetoxified alcoholics (36), nicotine smokers (27,37,38), and canna-is smokers (27,37,38). Drug-related stimuli, however, increase ac-

ivity of the reward circuitry in drug addicts (39). Likewise, bluntedesponsiveness of the reward circuitry in our study is paralleled byncreased activity in response to OCD-provoking stimuli in previoustudies (reviewed by [18]). Of interest, a recent case study showedfficacy of deep brain stimulation in the NAcc in a patient with OCD,evere nicotine addiction, and eating problems (40). The NAcc ismportant for focusing on potential alerting and rewarding envi-onmental stimuli that can be used for modulation of behavior byeinforcement learning (40). Therefore, the NAcc may be less re-ponsive when recruited during conventional reward processingue to its bias toward drugs of abuse in addiction, as well as towardbsessions and compulsions in OCD, supporting the conceptualiza-

ion of OCD as a disorder of behavioral addiction (8). Althoughopulation studies show relatively little comorbidity between OCDnd substance abuse, dysfunctional brain reward circuitry underly-

ng both disorders may explain some shared phenomena, such as aependency on repetitious, self-defeating behavior that becomesore difficult to control over time. Because OCD patients are al-

eady fully engaged in reinforcing compulsive behaviors, they maye less prone to develop substance abuse.

The present study has a number of potential limitations. Nine ofur 18 OCD subjects were using medication. Post hoc, we tested for

he effects of medication in the OCD group by comparing thectivation between medicated and medication-free patients usingwhole-brain, two-sample t test, which did not show any signifi-

ant differences in our regions of interest. Even upon lowering thehreshold to p � .001, not corrected for multiple comparisons, no

ctivation related to medication was found. We take this result to

www.sobp.org/journal

Table 2. Maximum t Value and Talairach Coordinates of Brain Regions That Show Altered Reward Activity in Healthy Control Subjects Versus OCD Patients and in High-Risk Assessment OCD VersusContamination Fear OCD

Control Group (n � 19) OCD Group (n � 18) Control Subjects � OCDHigh-Risk Assessment (n � 11) �

Contamination Fear (n � 7)

MNI Coordinates MNI Coordinates MNI Coordinates MNI Coordinates

Region R/L x y z Vox Max t BA x y z Vox Max t BA x y z Vox Max t BA x y z Vox Max t BA

Reward AnticipationNucleus accumbens L �12 16 4 44 13.49 �8 0 12 68 9.19 �8 12 0 2 9.14 �16 8 8 11 6.74

L �8 20 4 65 8.61 �16 16 �8 7 7.81L �16 16 �4 20 7.41R 4 8 8 144 10.34 8 4 8 3 5.34 16 20 �4 15 7.69 8 16 0 31 8.01

Putamen L �20 20 0 59 8.94 �24 8 4 11 6.18 �28 24 0 10 6.80Inferior frontal gyrus L �44 24 0 60 10.42 47 �44 24 0 40 8.53 47 �48 16 �4 13 7.09 47 �52 12 0 32 8.19 22

L �56 12 8 49 9.21 44 �60 12 16 2 6.05 44 �52 12 12 6 5.62 44Thalamus L �8 �4 16 87 9.97Anterior cingulate

gyrus 0 12 44 36 9.06 32 8 8 44 1 5.43 32 �8 12 44 6 6.22 320 8 40 1 5.25 24

Reward FeedbackNucleus accumbens R 12 8 �4 48 10.78 8 12 4 102 13.07

L �32 �4 12 64 11.16Medial frontal gyrus L �4 48 28 21 6.76 9

R 4 60 36 3 5.77 9R 4 52 44 5 5.27 9

Superior frontal gyrus L �40 �24 24 49 11.33 13 �44 �24 20 74 12.78 13L �56 �24 20 86 10.48 40 �56 �16 36 54 7.28 4

Thalamus L �12 �12 �8 51 11.89 �4 �20 8 128 12.92Orbitofrontal cortex 0 48 0 134 11.92 32 0 48 4 214 15.85 32Precuneus L �20 �60 16 50 10.60 30 �12 �60 24 186 12.06 31 �12 �44 8 38 6.59

R 8 �60 20 57 11.0 31 �16 �40 8 46 10.87 �16 �56 12 32 10.82 308 �60 24 84 10.45 31 4 �56 12 36 9.65 23

Insula L �36 28 0 33 9.53 47 �36 24 8 10 10.60 45�44 20 4 41 9.08 45

Anterior cingulategyrus

L �4 24 28 48 9.73 24 �4 16 40 15 9.12 32 �4 28 40 17 6.92 6

0 36 28 18 8.86 32Medial frontal gyrus 0 52 28 113 10.56 9 0 48 8 24 6.04 32

4 52 8 13 7.29 100 56 12 28 8.04 10

Whole-brain group comparisons thresholded at p � .05 corrected.BA, Brodmann area; L, left; Max t, maximum t value; MNI, Montreal Neurological Institute; OCD, obsessive-compulsive disorder; R, right; Vox, number of voxels.

872B

IOL

PSY

CH

IATRY

2011;69:867–874

M.Figee

eta

l.

ww

w.so

bp

.org

/jou

rnal

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

M. Figee et al. BIOL PSYCHIATRY 2011;69:867–874 873

indicate that medication use is not likely to explain the significantdifferences observed between healthy control subjects and OCDpatients. Also, whereas the nine medicated subjects used differenttypes of drugs in various, often subtherapeutic doses, medicationuse was similar in the high-risk assessment and contamination feargroups.

Attenuated reward anticipation has been demonstrated in de-pression (41). Although our mean HAM-D score of 16.8 � 6.6 issuggestive for minor to moderate depressive symptoms, only twopatients met DSM-IV criteria for a comorbid diagnosis of depressivedisorder. The HAM-D scores were nevertheless measured in all pa-tients as part of our routine screening. We therefore believe thesescores reflect depressive features overlapping with or secondary toOCD. Moreover, depression scores were not related to BOLDchanges and group differences remained significant even whenexcluding the patients with a diagnosis of depressive disorder.

Further, we included more female than male subjects in thisstudy, whereas recent research has indicated stronger striatal acti-vation in healthy men than in women if money was to be expected(42). However, gender ratios did not differ between patients andhealthy control subjects. Patients were slightly less educated thenhealthy control subjects (mean difference 1.6 years of education),most likely because of the illness process. Finally, within the patientgroup, the contamination fear group was significantly older thanthe group with high-risk assessment symptoms; however, NAccactivation remained significantly smaller in the contamination feargroup when age was included as a covariate.

The study also has several notable strengths. The patient groupwas highly relevant for studying OCD pathophysiology becausemean Y-BOCS score was of high severity (29.1, SD 8.5). We em-ployed a robust reward paradigm coupled with a very rapid imageacquisition method, revealing activation differences that survivedwhole-brain correction for multiple comparisons, both betweengroups and within our OCD group. Also, our main difference inBOLD activation was found in a focus that is anatomically compara-ble with the activation focus that was found by Knutson et al. (16) inhealthy subjects anticipating monetary gain.

In conclusion, the present study suggests that OCD is associatedwith reduced NAcc activity during reward anticipation, especially incontamination fear. This finding is consistent with a proposed rolefor the NAcc, as part of the frontal-striatal network, in the pathogen-esis of OCD and supports its conceptualization as a disorder ofreward processing and behavioral addiction. Reward processingand its neural underpinnings in compulsivity may be importantavenues for future research. Nucleus accumbens activity duringreward anticipation in OCD patients might be explored as a poten-tial marker for treatment response. Also, studying changes in brainactivity that are related to deep brain stimulation of the NAcc inOCD patients might further elucidate its role in the disease.

All authors report no biomedical financial interests or potentialconflicts of interest.

Supplementary material cited in this article is available online.

1. Ruscio AM, Stein DJ, Chiu WT, Kessler RC (2010): The epidemiology ofobsessive-compulsive disorder in the National Comorbidity Survey Rep-lication. Mol Psychiatry 15:53– 63.

2. Fullana MA, Mataix-Cols D, Caspi A, Harrington H, Grisham JR, Moffitt TE,Poulton R (2009): Obsessions and compulsions in the community: Prev-alence, interference, help-seeking, developmental stability, and co-oc-curring psychiatric conditions. Am J Psychiatry 166:329 –336.

3. van den Heuvel OA, Veltman DJ, Groenewegen HJ, Dolan RJ, Cath DC,Boellaard R, et al. (2004): Amygdala activity in obsessive-compulsive

disorder with contamination fear: A study with oxygen-15 waterpositron emission tomography. Psychiatry Res 132:225–237.

4. Chamberlain SR, Blackwell AD, Fineberg NA, Robbins TW, Sahakian BJ(2005): The neuropsychology of obsessive compulsive disorder: Theimportance of failures in cognitive and behavioural inhibition as candi-date endophenotypic markers. Neurosci Biobehav Rev 29:399 – 419.

5. Chamberlain SR, Menzies L, Hampshire A, Suckling J, Fineberg NA, delCampo N, et al. (2008): Orbitofrontal dysfunction in patients with obses-sive-compulsive disorder and their unaffected relatives. Science 321:421– 422.

6. Page LA, Rubia K, Deeley Q, Daly E, Toal F, Mataix-Cols D, et al. (2009): Afunctional magnetic resonance imaging study of inhibitory control inobsessive-compulsive disorder. Psychiatry Res 174:202–209.

7. Denys D, Zohar J, Westenberg HG (2004): The role of dopamine inobsessive-compulsive disorder: Preclinical and clinical evidence. J ClinPsychiatry 65(suppl 14):11–17.

8. Holden C (2001): “Behavioral” addictions: Do they exist? Science 294:980 –982.

9. Denys DA (2004): On Certainty: Studies in Obsessive Compulsive Disorder[dissertation]. Amsterdam: Utrecht University.

0. Nielen MM, den Boer JA, Smid HG (2009): Patients with obsessive-com-pulsive disorder are impaired in associative learning based on externalfeedback. Psychol Med 39:1519 –1526.

1. Tobler PN, Christopoulos GI, O’Doherty JP, Dolan RJ, Schultz W (2009):Risk-dependent reward value signal in human prefrontal cortex. ProcNatl Acad Sci U S A 106:7185–7190.

2. Whiteside SP, Port JD, Abramowitz JS (2004): A meta-analysis of func-tional neuroimaging in obsessive-compulsive disorder. Psychiatry Res132:69 –79.

3. Denys D, Mantione M, Figee M, van den Munckhof P, Koerselman F,Westenberg H, et al. (2010): Deep brain stimulation of the nucleus ac-cumbens for treatment-refractory obsessive-compulsive disorder. ArchGen Psychiatry 67:1061–1068.

4. Huff W, Lenartz D, Schormann M, Lee SH, Kuhn J, Koulousakis A, et al.(2010): Unilateral deep brain stimulation of the nucleus accumbens inpatients with treatment-resistant obsessive-compulsive disorder: Out-comes after one year. Clin Neurol Neurosurg 112:137–143.

5. Sturm V, Lenartz D, Koulousakis A, Treuer H, Herholz K, Klein JC,Klosterkotter J (2003): The nucleus accumbens: A target for deep brainstimulation in obsessive-compulsive- and anxiety-disorders. J ChemNeuroanat 26:293–299.

6. Knutson B, Fong GW, Adams CM, Varner JL, Hommer D (2001): Dissoci-ation of reward anticipation and outcome with event-related fMRI. Neu-roreport 12:3683–3687.

7. Schultz W, Tremblay L, Hollerman JR (2000): Reward processing in pri-mate orbitofrontal cortex and basal ganglia. Cereb Cortex 10:272–284.

8. Menzies L, Chamberlain SR, Laird AR, Thelen SM, Sahakian BJ, BullmoreET (2008): Integrating evidence from neuroimaging and neuropsycho-logical studies of obsessive-compulsive disorder: The orbitofronto-stri-atal model revisited. Neurosci Biobehav Rev 32:525–549.

9. Mataix-Cols D, Wooderson S, Lawrence N, Brammer MJ, Speckens A,Phillips ML (2004): Distinct neural correlates of washing, checking, andhoarding symptom dimensions in obsessive-compulsive disorder. ArchGen Psychiatry 61:564 –576.

0. Phillips ML, Marks IM, Senior C, Lythgoe D, O’Dwyer AM, Meehan O, et al.(2000): A differential neural response in obsessive-compulsive disorderpatients with washing compared with checking symptoms to disgust.Psychol Med 30:1037–1050.

1. Enzi B, de Greck M, Prosch U, Tempelmann C, Northoff G (2009): Is ourself nothing but reward? Neuronal overlap and distinction betweenreward and personal relevance and its relation to human personality.PLoS One 4:e8429.

2. Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, etal. (1998): The Mini-International Neuropsychiatric Interview (M.I.N.I.):The development and validation of a structured diagnostic psychiatricinterview for DSM-IV and ICD-10. J Clin Psychiatry 59(suppl 20):22–33.

3. van Vliet IM, de Beurs E (2007): [The mini-international NeuropsychiatricInterview. A brief structured diagnostic psychiatric interview for DSM-IVen ICD-10 psychiatric disorders]. Tijdschr Psychiatr 49:393–397.

4. Goodman WK, Price LH, Rasmussen SA, Mazure C, Fleischmann RL, Hill

CL, et al. (1989): The Yale-Brown Obsessive Compulsive Scale. I. Devel-opment, use, and reliability. Arch Gen Psychiatry 46:1006 –1011.

www.sobp.org/journal

3

3

3

3

3

4

4

4

874 BIOL PSYCHIATRY 2011;69:867–874 M. Figee et al.

w

25. Goodman WK, Price LH, Rasmussen SA, Mazure C, Delgado P, HeningerGR, Charney DS (1989): The Yale-Brown Obsessive Compulsive Scale. II.Validity. Arch Gen Psychiatry 46:1012–1016.

26. Hamilton M (1960): A rating scale for depression. J Neurol NeurosurgPsychiatry 23:56 – 62.

27. van Hell HH, Vink M, Ossewaarde L, Jager G, Kahn RS, Ramsey NF (2010):Chronic effects of cannabis use on the human reward system: An fMRIstudy. Eur Neuropsychopharmacol 20:153–163.

28. Fiorillo CD, Tobler PN, Schultz W (2003): Discrete coding of reward proba-bility and uncertainty by dopamine neurons. Science 299:1898–1902.

29. Knutson B, Adams CM, Fong GW, Hommer D (2001): Anticipation ofincreasing monetary reward selectively recruits nucleus accumbens.J Neurosci 21:RC159.

30. van Gelderen P, Ramsey NF, Liu G, Duyn JH, Frank JA, Weinberger DR,Moonen CT (1995): Three-dimensional functional magnetic resonanceimaging of human brain on a clinical 1.5-T scanner. Proc Natl Acad SciU S A 92:6906 – 6910.

31. Worsley KJ, Liao CH, Aston J, Petre V, Duncan GH, Morales F, Evans AC(2002): A general statistical analysis for fMRI data. Neuroimage 15:1–15.

32. Kiebel SJ, Poline JB, Friston KJ, Holmes AP, Worsley KJ (1999): Robustsmoothness estimation in statistical parametric maps using standard-ized residuals from the general linear model. Neuroimage 10:756 –766.

33. Remijnse PL, Nielen MM, van Balkom AJ, Cath DC, van Oppen P, UylingsHB, Veltman DJ (2006): Reduced orbitofrontal-striatal activity on a rever-sal learning task in obsessive-compulsive disorder. Arch Gen Psychiatry63:1225–1236.

34. Rauch SL, Wedig MM, Wright CI, Martis B, McMullin KG, Shin LM, et al.(2007): Functional magnetic resonance imaging study of regional brainactivation during implicit sequence learning in obsessive-compulsive

disorder. Biol Psychiatry 61:330 –336.

ww.sobp.org/journal

5. Schienle A, Schafer A, Stark R, Walter B, Vaitl D (2005): Neural responsesof OCD patients towards disorder-relevant, generally disgust-inducingand fear-inducing pictures. Int J Psychophysiol 57:69 –77.

6. Wrase J, Schlagenhauf F, Kienast T, Wustenberg T, Bermpohl F, Kahnt T,et al. (2007): Dysfunction of reward processing correlates with alcoholcraving in detoxified alcoholics. Neuroimage 35:787–794.

7. Buhler M, Vollstadt-Klein S, Kobiella A, Budde H, Reed LJ, Braus DF, etal. (2009): Nicotine dependence is characterized by disordered re-ward processing in a network driving motivation. Biol Psychiatry67:745–752.

8. Martin-Soelch C, Missimer J, Leenders KL, Schultz W (2003): Neural ac-tivity related to the processing of increasing monetary reward in smok-ers and nonsmokers. Eur J Neurosci 18:680 – 688.

9. Diekhof EK, Falkai P, Gruber O (2008): Functional neuroimaging of re-ward processing and decision-making: A review of aberrant motiva-tional and affective processing in addiction and mood disorders. BrainRes Rev 59:164 –184.

0. Mantione M, van de Brink W, Schuurman PR, Denys D (2010): Smokingcessation and weight loss after chronic deep brain stimulation of thenucleus accumbens: Therapeutic and research implications: Case re-port. Neurosurgery 66:E218.

1. Pizzagalli DA, Holmes AJ, Dillon DG, Goetz EL, Birk JL, Bogdan R, et al.(2009): Reduced caudate and nucleus accumbens response to rewardsin unmedicated individuals with major depressive disorder. Am J Psychi-atry 166:702–710.

2. Spreckelmeyer KN, Krach S, Kohls G, Rademacher L, Irmak A, Konrad K, etal. (2009): Anticipation of monetary and social reward differently acti-vates mesolimbic brain structures in men and women. Soc Cogn Affect

Neurosci 4:158 –165.