SVPXV

Bsd

Maoc

Rbcrds

Ci

Kft

WaeGsRdCsepse2

F

A

R

0d

ex Differences in Neural Responses to Disgustingisual Stimuli: Implications for Disgust-Relatedsychiatric Disorders

avier Caseras, David Mataix-Cols, Suk Kyoon An, Natalia S. Lawrence, Anne Speckens,incent Giampietro, Michael J. Brammer, and Mary L. Phillips

ackground: A majority of patients with disgust-related psychiatric disorders such as animal phobias and contamination-related obses-ive-compulsive disorder are women. The aim of this functional magnetic resonance imaging (fMRI) study was to examine possible sexifferences in neural responses to disgust-inducing stimuli that might help explain this female predominance.

ethods: Thirty-four healthy adult volunteers (17 women, all right-handed) were scanned while viewing alternating blocks of disgustingnd neutral pictures from the International Affective Picture System. Using a partially-silent fMRI sequence, the participants rated their levelf discomfort after each block of pictures. Skin conductance responses (SCR) were measured throughout the experiment. All participantsompleted the Disgust Scale.

esults: Both women and men reported greater subjective discomfort and showed more SCR fluctuations during the disgusting picturelocks than during the neutral picture blocks. Women and men also demonstrated a similar pattern of brain response to disgustingompared with neutral pictures, showing activation in the anterior insula, ventrolateral and dorsolateral prefrontal cortices, and visualegions. Compared with men, women had significantly higher disgust sensitivity scores, experienced more subjective discomfort, andemonstrated greater activity in left ventrolateral prefrontal regions. However, these differences were no longer significant when disgustensitivity scores were controlled for.

onclusions: In healthy adult volunteers, there are significant sex-related differences in brain responses to disgusting stimuli that are

rrevocably linked to greater disgust sensitivity scores in women. The implications for disgust-related psychiatric disorders are discussed.

ey Words: Animal phobias, blood-injury phobias, contaminationears, disgust, emotion, obsessive-compulsive disorder, sex, symp-om dimensions, washing

omen and men differ in their emotional perceptionand experience. It has been consistently reported thatwomen are more expressive of emotions (e.g., Kring

nd Gordon 1998), tend to score higher on scales related tomotional experiences such as neuroticism (e.g., Goodwin andotlib 2004), and have increased risk of suffering from depres-ion and most anxiety disorders (e.g., Wilhelm et al. 1997).ecent research in animals and humans suggests marked sexifferences on brain structure and function (for a review seeahill 2006). For instance, morphometric studies have shownexual dimorphism in several brain structures implicated inmotional processing, such as the cingulate and ventrolateralrefrontal cortices (larger in women) and the medial temporaltructures, including the amygdala (larger in men) (Goldsteint al. 2001; Good et al. 2001; Paus et al. 1996; Pujol et al.002). These differences are hypothesized to be associated

rom the Departments of Psychological Medicine (XC, DM-C, SKA, NSL,MLP), Psychology (DM-C, AS), and Biostatistics (VG, MJB), King’s CollegeLondon, Institute of Psychiatry, United Kingdom; Department of Psychi-atry and Forensic Medicine (XC), Institute of Neurosciences, Autono-mous University of Barcelona, Catalonia, Spain; and the Department ofPsychiatry (MLP), University of Pittsburgh, Pittsburgh, Pennsylvania.

ddress reprint requests to David Mataix-Cols, Ph.D., Division of Psycholog-ical Medicine, King’s College London, Institute of Psychiatry, PO 69 DeCrespigny Park Rd., London SE5 8AF United Kingdom; E-mail: d.mataix@iop.kcl.ac.uk.

eceived August 31, 2006; revised October 27, 2006; accepted October 31,

2006.

006-3223/07/$32.00oi:10.1016/j.biopsych.2006.10.030

with sex steroid activity early in development (Goldstein et al.2001).

Functional neuroimaging studies of sex differences in emo-tion have been less conclusive perhaps owing to methodologicaldifferences between studies (a wide variety of emotional taskshave been used) and the complexity and diversity of theemotions under study. In response to positive emotional stimuli,men have been found to show significantly greater activationthan women in several emotional brain regions, such as visualcortex and amygdala (Hamann et al. 2004; Sabatinelli et al. 2004;Wrase et al. 2003), but less consistent results emerged regardingaffectively negative stimuli. For instance, Wrase et al. (2003)found that women had greater responses than men in theanterior cingulate cortex (Brodmann’s area [BA] 24/33) duringthe presentation of affectively negative pictures, whereas othersfound greater amygdala activations to negative stimuli in men(Schienle et al. 2005; but see McClure et al. 2004). A recentmeta-analysis of 65 positron emission tomography and func-tional magnetic resonance imaging (fMRI) studies (Wager et al.2003) suggests that the relationship between emotional valence(positive, negative), sex, and laterality is more complex thanpreviously thought and might be region-specific. They found thatwomen did not show more frequent activation overall, butwomen more frequently activated midline limbic structures,including the subcallosal anterior cingulate, thalamus, midbrain,and cerebellum, whereas men showed more activation in leftinferior frontal and posterior cortices.

In addition, it is plausible that different negative emotions(disgust, fear, sadness, anger) might show different patterns ofsexual dimorphism. For instance, Schienle et al. (2005) foundgreater bilateral amygdala and left fusiform gyrus activations inmen when presented with fear-inducing pictures but no be-

tween-sex differences in brain responses to disgust-inducing

BIOL PSYCHIATRY 2007;62:464–471© 2007 Society of Biological Psychiatry

pf

rhbase1m2dp1aiebestic2t

M

P

hwc(d(2CPs

M

(shs(Tt

S

hPmcpnwtvl

X. Caseras et al. BIOL PSYCHIATRY 2007;62:464–471 465

ictures. Clearly, more research is needed to substantiate theseindings.

This study investigated sex differences in brain activation inesponse to aversive pictures rated as highly disgusting byealthy individuals. Disgust is a basic emotion that can be elicitedy a wide range of stimuli, such as food, body products, certainnimals, poor hygiene, and also more abstract or moral concernsuch as “inappropriate” sexuality, violations of the normal bodynvelope (gore, surgery, deformity), and death (Rozin and Fallon987; Rozin et al. 2000). Women systematically score higher thanen on disgust sensitivity scales (Haidt et al. 1994; Olatunji et al.005; Rozin et al. 1999, 2000). Furthermore, the importance ofisgust sensitivity in certain anxiety disorders with a clear femaleredominance, such as animal and blood-injury phobias (Davey994; de Jong et al. 2000; Matchett and Davey 1991; Mulkens etl. 1996; Olatunji et al. 2006; Sawchuk et al. 2000) and contam-nation-related obsessive-compulsive disorder (OCD) (Lawrencet al. in press; Olatunji et al. in press; Woody and Tolin 2002), haseen well established. On the basis of this evidence we hypoth-sized that compared with men, women would: 1) have highercores on a disgust sensitivity scale, and 2) show greater activa-ion in regions previously associated with disgust (anteriornsula, striatum, associative visual cortex, ventrolateral prefrontalortex [PFC], anterior cingulate gyrus) (Phillips et al. 1997, 1998,004; Sprengelmeyer et al. 1998; Wicker et al. 2003) in responseo aversive visual stimuli rated as highly disgusting.

ethods and Materials

articipantsThirty-four healthy adult volunteers (17 women, all right-

anded) with no history of neurological or psychiatric disorderere included in the present study. Some had participated as

ontrol subjects in an OCD study conducted in our laboratoryMataix-Cols et al. 2004), but none had been exposed to theisgusting stimuli before. The mean age was similar for womenmean � 29.88, SD � 9.04, range 19–51) and men [mean �9.06, SD � 7.30, range 18–45; t (32) � .29, p � .1]. The Ethicsommittee (Research) of the Maudsley Hospital and Institute ofsychiatry approved the study protocol, and all participantsigned an informed consent form before their participation.

easuresSensitivity to disgust was measured with the Disgust Scale

DS; Haidt et al. 1994). The DS is a widely used 32-itemelf-administered questionnaire for which reliability and validityave been well established in numerous studies. Depression andtate anxiety were evaluated with the Beck Depression InventoryBDI; Beck and Steer 1987) and the state subscale of the Staterait Anxiety Inventory (STAI-S; Spielberger et al. 1983), respec-ively.

timuli and ParadigmFifty color pictures of scenes rated as highly disgusting by

ealthy subjects were obtained from the International Affectiveicture System (IAPS; Lang et al. 1997). These included scenes ofutilated bodies/wounds (64%), small animals (snakes, spiders,

ockroaches, and rats; 24%), and body products (12%). Fiftyictures of neutral or mildly positive scenes (e.g., furniture,ature scenes, urban landscapes, household items, pets, families)ere also selected from the IAPS and used as control stimuli. All

he pictures were selected after an independent group of healthyolunteers (not included in this study) had rated an originally

arger pool of pictures according to their level of visual complex-

ity, disgust, and anxiety on a 0–3 scale (0 � nil; 3 � high).Pictures that were too simple/complex were excluded. The final100 stimuli were well matched regarding visual complexity and,as intended, the aversive pictures induced more disgust andanxiety than the neutral pictures (data not shown).

All subjects participated in a 6-min fMRI experiment in whichthey viewed 10 20-sec alternating blocks of disgusting andneutral pictures in a counterbalanced order. Each block con-sisted of 10 pictures of the same valence presented for 1,950 msand with an inter-stimulus interval of 50 ms. Before the presen-tation of each set of pictures, subjects were played a pre-recorded voice file by means of high-fidelity pneumatic head-phones, instructing them to imagine being in a particularsituation while looking at the scenes they were about to see (e.g.,“Imagine that you must come into contact with what’s shown inthe pictures”). After each set of pictures, another pre-recordedsound file of the question “How anxious do you feel?” wasplayed, and the subjects rated their subjective anxiety on a 0 (noanxiety) to 8 (extreme anxiety) scale.

Skin Conductance Response Data AcquisitionSkin conductance responses (SCRs) were measured continu-

ously during the fMRI session with silver chloride electrodesattached to the medial phalanges of the first and second fingersof the non-dominant hand. The SCR data were analyzed withlocally developed software (SCAnalyse version 4; J.Dalton, KCL,London). The SCR fluctuations from the running baseline with aminimum rise time of 500 ms and a minimum amplitude of .01microSiemen were automatically detected with our software. Thenumber of SCR fluctuations in each 20-sec block of pictureviewing was measured. Where there were no fluctuations, 0 wasrecorded for that block. The total number of fluctuations wasthen calculated separately for blocks of neutral and disgustingpictures.

MRI AcquisitionGradient echo echoplanar (EPI) images were acquired on a

GE Signa 1.5-T Neuro-optimized MR system (General Electric,Milwaukee, Wisconsin) at the Maudsley Hospital, London. Onehundred T2*-weighted whole-brain volumes depicting bloodoxygen level dependent (BOLD) contrast (Ogawa et al. 1990)and consisting of 16 slices oriented according to the bicommis-sural plane (thickness 7 mm, .7 mm gap) were acquired over 6min (repetition time [TR] � 2.0 sec; echo time [TE] � 40 msec;field of view � 24 cm; flip angle � 70; 64 � 64 matrix). This EPIdataset provided almost complete brain coverage.

In each 20-sec stimulus presentation block, subjects viewedeither 10 disgusting or 10 neutral pictures. Ten whole-brainvolumes were acquired during each block. Each block wasfollowed by: 1) an 8-sec period of complete silence during whichsubjects were asked to rate their level of anxiety, and 2) a further8-sec period during which the subjects listened to a sound filecontaining instructions pertinent to the next stimulus block. Four“dummy-volumes” were excited during this 8-sec period withexactly the same radio frequency envelope and gradient sliceselection parameter, with the same repetition time of 2 sec toallow the magnetization to reach an equilibrium amplitudebefore the next period of data acquisition. The frequency-encoding gradient was turned off during this period to minimizeacoustic noise and ensure that the instructions were heard clearlyby the subjects. The four dummy volumes before and the fourvolumes after each block were modeled in the analyses and

treated as a nuisance variable.

www.sobp.org/journal

s4r3a

S

nup

dmrsmnpmcshupssmdpdtobtcsbdc

teelriPaa

pa

wimcTrd

466 BIOL PSYCHIATRY 2007;62:464–471 X. Caseras et al.

w

Individual brain activation maps were co-registered to atructural scan with the following acquisition parameters: TE �0 msec, TR � 3000 msec, field of view � 24 cm, imageesolution � 128 � 128, number of slices � 43, slice thickness �.0 mm, inter-slice gap � .3 mm, and number of signalverages � 8.

tatistical AnalysesQuestionnaires. The scores in the DS, BDI, and STAI-S were

ormally distributed (Kolmogorov-Smirnov test, all three p val-es �.1); therefore, mean comparisons between genders wereerformed by means of the Student t test.

fMRI Data. The fMRI data were analyzed with softwareeveloped at the Institute of Psychiatry (XBAM), with a nonpara-etric approach to minimize assumptions. Data were first cor-

ected for subject motion (Bullmore et al. 1999) and thenmoothed with a Gaussian filter (full width at half maximal 7.2m) chosen to improve signal-to-noise ratio over the spatialeighborhood of each voxel. Responses to the experimentalaradigm were then detected by time-series analysis with a linearodel in which each component of the experimental design was

onvolved separately with two � variate functions (peak re-ponses at 4 and 8 sec, respectively) to permit variability in theemodynamic delay. The method of Friman et al. (2003) wassed to constrain model fits to those deemed physiologicallylausible. After computation of the model fit, a goodness of fittatistic was computed. This consisted of the ratio of the sum ofquares of deviations from the mean image intensity due to theodel (over the whole time series) to the sum of squares ofeviations due to the residuals (SSQratio). This addresses theroblem inherent in the use of the F statistic that the residualegrees of freedom are often unknown in fMRI time series owingo the presence of colored noise in the signal. After computationf the observed SSQratio at each voxel, the data were permutedy the wavelet-based method described and extensively charac-erized in Bullmore et al. (2001), which permits the data-drivenalculation of the null distribution of SSQratios under the as-umption of no experimentally determined response. This distri-ution can then be used to threshold the activation maps at anyesired type I error rate. In addition to the SSQratio, the % BOLDhange was also calculated from the model fit at each voxel.

The detection of activated regions was extended from voxelo cluster level with the method described in detail by Bullmoret al. (2001). The observed and randomized SSQratio data forach individual were transformed into standard space of Ta-airach and Tournoux (1988), and group maps of activatedegions were computed with the median observed and random-zed SSQ ratio data as described by Brammer et al. (1997).ermutation methods and median statistics were employed tollow exact computation of p values with minimal assumptionsnd the minimization of outlier effects.

Comparisons of responses between men and women wereerformed by fitting the data at each intracerebral voxel at whichll subjects have non-zero data with a linear model of the type:

Y � a � bX � e

here Y is the vector of BOLD effect sizes for each individual, Xs the contrast matrix for the between-group contrast, a is theean effect across all individuals in the two groups, b is the

omputed group difference, and e is a vector of residual errors.he model is fitted by minimizing the sum of absolute deviationsather than the sums of squares to reduce outlier effects. The null

istribution of b is computed by permuting data between groups

ww.sobp.org/journal

(assuming the null hypothesis of no effect of group membership)and refitting the aforementioned model. Group difference mapsare computed as described previously at voxel or cluster level byappropriate thresholding of the null distribution of b. In thisstudy, BOLD effect maps were used to compute significant groupdifferences rather than standardized measures such as SSQratio,F, or t, because these contain explicit noise components (errorSSQ or error variance), raising the possibility that group differ-ences resulting from SSQratio, F, or t comparisons could reflectdifferences in noise rather then signal.

Results

Questionnaire DataAs predicted, women had significantly higher disgust sensi-

tivity scores than men did. However, the two groups did notdiffer in their scores on the BDI or the STAI-S (Table 1).

Subjective Discomfort Scores During the ExperimentBoth men and women rated the disgusting pictures as more

anxiety-provoking than the neutral pictures [paired t test for ment (16) � 8.67; women t (16) � 16.72; both p � .001], but womenrated the disgusting pictures as significantly more anxiety-pro-voking than men did [women mean � 6.3, SD � 1.5; menmean � 4.8, SD � 2.4; t (32) � 2.21, p � .03]. The neutralpictures were rated similarly by the two genders [women mean �.9, SD � 1; men mean � .6, SD � .7; t (32) � .92, p � .36].

Skin Conductance Responses During the ExperimentSkin conductance data were available from 18 participants.

Fourteen of these showed at least 1 significant fluctuation duringthe experiment; the remaining participants (3 women, 1 man)showed no fluctuations and were therefore excluded from theanalysis. A paired samples t test revealed a significantly greaternumber of SCR fluctuations during the disgust (mean � 5.7, SD �7) than the neutral (mean � 3.4, SD � 5.1) blocks of pictureviewing [t (13) � 2.9, p � .01]. However, there were no signifi-cant differences (p � .1) in SCR between the 9 women and 5 menfor whom data were available.

Sex Differences in Brain ActivationAs seen in Figure 1, the contrast Disgust � Neutral showed

similar significant activations in women and men in the followingregions: bilateral cerebellum; bilateral occipito-parietal regions(BA 7/17/18/19/40), including the fusiform gyrus (BA 18/19);bilateral occipito-temporal regions (BA 19/38/42, left BA 22); andbilateral dorsolateral prefrontal regions (BA 8/9/44/45/46, ex-tending toward the pre-central (BA 4/6) gyrus), ventrolateral, andorbital prefrontal regions (BA 10/11/47), bilateral insula, thala-mus, striatum, and brain stem. Only men significantly activated acluster in the bilateral medial PFC (BA 8/9/10) extending to the

Table 1. Comparisons of Questionnaire Scores Between Women and Men

Scale(score range)

Women (n � 17) Men (n � 17) Student t(df � 32)Mean (SD) Mean (SD)

DS (0–100) 55.17 (15.10) 37.29 (16.59) 3.29a

BDI (0–63) 4.29 (3.60) 2.71 (3.44) 1.31STAI-S (20–80) 32.29 (6.35) 31.53 (10.38) .26

DS, Disgust Scale; BDI, Beck Depression Inventory; STAI-S, State subscaleof the State- Trait Anxiety Inventory.

ap � .005.

right dorsal anterior cingulate cortex (BA 24/32).

mg(vitn

FI2

TBP

W

M

7c

X. Caseras et al. BIOL PSYCHIATRY 2007;62:464–471 467

The analysis of variance comparing the Disgust � Neutralaps between sexes revealed that women had significantlyreater activation in a large cluster in the left ventrolateral PFCpeak activation x � �30, y � 37, z � �18; cluster size 110oxels; clusterwise p value � .0021), including the middle/nferior frontal (BA11/47) and orbital (BA11) gyri and extendingo the temporal pole (BA38/21) (Table 2 and Figure 2). Men didot activate any regions significantly more than women.

igure 1. Significant blood oxygen level dependent responses in 17 womennternational Affective Picture System (IAPS) pictures superimposed on a h000). The left side of the brain appears on the left side of the image.

able 2. Brain Regions Showing Significant Differences in BOLD Signaletween Women and Men in Response to Disgusting Versus Neutralictures

BA x y zNumberof Voxels

omen � MenLeft middle/inferior frontal gyrus 11/47 �29 37 �18 49Left orbital gyrus 11 �25 37 �24 34Left inferior frontal gyrus 47 �40 26 �13 7Left superior frontal gyrus 11 �25 56 �18 4Left superior temporal gyrus 38 �40 15 �29 4Left middle temporal gyrus 21 �43 7 �24 4en � WomenNone

The listed brain regions and coordinates represent maxima in each-mm slice of the large three-dimensional ventrolateral prefrontal cortexluster (x � � 30, y � 37, z � �18) that separated women from men.

BOLD, blood oxygen level dependent; BA, Brodmann’s area.

To facilitate the comparability of these findings with otherstudies, we reanalyzed our data with a t test rather than anonparametric permutation-based test. Using a functional re-gion-of-interest approach, as is commonly performed in otherpackages such as AFNI (analysis of functional neuro-images)(e.g., Lawrence et al. 2002), we extracted the mean effect sizevalues of the left ventrolateral prefrontal cortical response to theaversive versus neutral scenes in women and men and enteredthem into SPSS (SPSS, Chicago, Illinois). A t test comparing thetwo groups revealed a statistically significant difference in mag-nitude of activation in this cluster [t (32) � 3.75, p � .001],showing greater magnitude of activation in women, supportingthe results of the nonparametric test.

Because on average women showed higher disgust sensitivitythan men and this personality trait could potentially drive theaforementioned differences in brain activation, we repeated theanalyses with disgust sensitivity scores as a covariate. The observedbetween-sex differences in subjective anxiety [F (1,31) � 1.664,p � .207] and brain activation were no longer significant.However because disgust sensitivity is so intimately related tofemale sex, removal of the variance in brain activation asso-ciated with disgust sensitivity would possibly remove most ofthe variance in brain activation associated with sex (see Millerand Chapman 2001 for a discussion of the uses and misuses ofanalysis of covariance). In an attempt to avoid this problem,we also repeated the analyses in a sub-sample of men (n � 10)

panel) and 17 men (bottom panel) in response to disgusting versus neutralsolution anatomical template with the MRIcro software (Rorden and Brett

(topigh-re

and women (n � 10) who were carefully matched for DS

www.sobp.org/journal

srsma

rlw

Fct�

468 BIOL PSYCHIATRY 2007;62:464–471 X. Caseras et al.

w

cores (in both groups, mean DS � 48 and SD � 13). Theesults showed that women and men no longer differed onubjective discomfort scores [women mean � 6.2, SD � 1.2;en mean � 5.0, SD � 2.8; t (18) � 1.26, p � .22] or in brain

ctivation, even with more relaxed statistical thresholds.Finally, we extracted the mean percentage change in BOLD

esponse (Disgust � Neutral) from the functionally-derived ventro-ateral PFC cluster for each individual (n � 34) and correlated itith their disgust sensitivity scores. As seen in Figure 3, there was

igure 3. Significant correlation between Disgust Scale scores and %hange in blood oxygen level dependent (BOLD) response (Disgust � Neu-ral) in the left ventrolateral prefrontal cortex cluster (x � �30, y � 37, z �

18) in the entire study sample (n � 34).

ww.sobp.org/journal

a significant positive correlation between activation in this clusterand DS scores (Spearman � � .38, p � .03).

Discussion

This study compared the brain responses of healthy adultwomen and men to aversive disgusting stimuli. The mainstrengths of the study were the relatively large sample of womenand men and the use of robust, nonparametric methods of fMRIdata analysis. As in many previous studies (e.g., Haidt et al. 1994;Olatunji et al. 2005; Rozin et al. 1999), women had significantlyhigher disgust sensitivity scores than men. This was reflected ingreater subjective discomfort in women when they viewedaversive disgusting pictures. Furthermore, although both womenand men showed widespread activations in regions previouslyidentified as mediating emotions in general and disgust inparticular (i.e., anterior insula, ventrolateral PFC, striatum, andvisual associative regions), we report for the first time thatwomen and men show quantitatively distinct activations in leftventrolateral prefrontal regions, including BA47 and BA11, ex-tending to the temporal pole. These findings are in agreementwith morphometric studies that found sexual dimorphism (largerin women) in several emotion-related regions, including theventrolateral PFC (Goldstein et al. 2001; Good et al. 2001).

The specific role of the ventrolateral PFC in mediating re-sponses to disgusting stimuli in women versus men is unclear,but several complementary explanations are possible. Spren-gelmeyer et al. (1998) suggested that the left ventrolateral PFCserved as a common emotion processing “end point,” integratinginformation from distributed networks involved in emotionprocessing. This hypothesis is supported by studies in healthypopulations that found that this region is significantly activatedby a variety of emotional tasks involving different negativeemotions (e.g., McClure et al. 2004; Sprengelmeyer et al. 1998).

Figure 2. Brain regions showing greater activationin women than men in response to aversive/disgusting pictures (see Table 2 for coordinates). Thefunctional data are superimposed on a high-reso-lution anatomical template with the MRIcro soft-ware (Rorden and Brett 2000). This image is fordisplay purposes only and indicates the most rep-resentative slices. L, left; R, right. The box plot de-picts the % change in blood oxygen level depen-dent (BOLD) response (Disgust � Neutral) forwomen and men taken from the most activatedvoxel within the cluster (x � �30, y � 37, z � �18).

For instance, McClure et al. (2004) found that adult women

saloisan�wddsmatfiaSnsctaiudto(coi

rcmen(blddcdCso2

dc((lsabM2m

X. Caseras et al. BIOL PSYCHIATRY 2007;62:464–471 469

howed greater right lateral orbitofrontal cortex activation tongry versus neutral faces than did adult men. Another relatedine of research points to a role for the ventrolateral PFC andther brain areas implicated in disgust (such as the neighboringnsula and somatosensory cortex) in the monitoring of internaltates (Critchley et al. 2004). When healthy control subjects paidttention to their heartbeat, BOLD responses were increased in aetwork of regions, including a left ventrolateral PFC cluster (x �48, y � 32, z � 4). The increased ventrolateral PFC response inomen in this study might therefore suggest that healthy womenemonstrated increased attention to their bodily response to theisgust pictures. This is plausible, because women reportedignificantly more subjective anxiety during the experiment thanen did. Another possibility is that the left BA47 region is more

ctivated in women because they tried harder to regulate/inhibitheir emotional responses during the experiment. The inferiorrontal gyrus (extending to BA47) has been implicated in inhib-tory processes in numerous cognitive (e.g., Rubia et al. 2003)nd emotional paradigms (e.g., Harenski and Hamman 2006;hafritz et al. 2006). Monk et al. (2006) have recently reported aegative correlation between activation in this region (right-ided) in response to angry faces and anxiety scores in adoles-ents with generalized anxiety disorder, perhaps suggesting thathis region is involved in the suppression or regulation ofnxiety. However, we could not find such an inverse correlationn this study (data not shown). The current design does not allows to disentangle the precise mechanisms at play. Event-relatedesigns would be better suited for this endeavor, although theyend to be less sensitive to detect subtle activations in the regionsf interest. Emotionally-guided response inhibition paradigmse.g., Shafritz et al. 2006) could be useful to separate the neuralorrelates of disgust perception per se from the neural correlatesf the suppression of emotional responses during the experimentn women versus men.

A second important finding of this study was that when wee-analyzed the data with disgust sensitivity scores as a nuisanceovariate and in a sub-sample of 10 women and 10 men carefullyatched for disgust sensitivity scores, the between-sex differ-

nces in subjective anxiety and brain activation became nonsig-ificant. Furthermore, a correlation analysis in the entire samplen � 34) revealed a statistically significant positive correlationetween disgust sensitivity scores and activation in the ventro-ateral PFC cluster. It is therefore plausible that the observed sexifferences might be, at least in part, explained by increasedisgust sensitivity scores in women. Similarly, sex differences inontamination fears have been found to be mediated by higherisgust sensitivity in women than in men (Olatunji et al. 2005).learly, female sex and elevated disgust sensitivity are difficult toeparate, because women systematically score higher than menn this personality trait (e.g., Haidt et al. 1994; Olatunji et al.005; Rozin et al. 1999).

The current results have implications for several anxietyisorders of predominant female prevalence that are also asso-iated with elevated disgust sensitivity, such as specific phobiasparticularly the animal and blood-injury subtypes) and OCDparticularly its contamination/washing dimension). Epidemio-ogical and clinical studies show a clear female predominance inpecific phobia (see Marks and Mataix-Cols 2004), and thessociation between specific phobias and disgust sensitivity haseen well documented (Davey et al. 1994; de Jong et al. 2000;atchett and Davey 1991; Mulkens et al. 1996; Olatunji et al.006; Sawchuk et al. 2000). Although there is a similar female/

ale ratio in OCD as a whole (e.g., Rasmussen and Eisen 1992),

there is a clear excess of women among patients with contami-nation/washing symptoms (e.g., Denys et al. 2004; Hasler et al.2005; Mataix-Cols et al. 1999), and several studies found aspecific association between contamination/washing—but notother types of OCD—symptoms and elevated disgust sensitivity(e.g., Lawrence et al. in press; Olatunji et al. in press; Woody andTolin 2002).

The “disgust circuitry” has been implicated in several neuro-imaging studies of patients with specific phobias (e.g., Rauch etal. 1995) and contamination-related OCD. For instance, in arecent study (Lawrence et al. in press), eight OCD patients (sevenwomen) with prominent contamination/washing symptomsshowed enhanced responses to facial expressions of disgust inthe left ventrolateral PFC (x � �41, y � 34, z � 0). Similarly,Shapira et al. (2003) found that a group of eight OCD patients(five women) with predominantly contamination concernsshowed significantly enhanced activation relative to controlsubjects in response to generally disgusting pictures in a rightand a left ventrolateral PFC cluster (x � �43, y � 24, z � �11)that is close to the one identified here (x � �30, y � 37, z ��18). Another recent study found activation in a left ventrolateralPFC/orbitofrontal cortex cluster in a group of 10 OCD patients(6 women) with mixed symptoms in response to generallydisgusting versus neutral scenes (Schienle et al. 2005). Our group(Mataix-Cols et al. 2004) has previously reported strong correla-tions between washing symptoms in OCD and brain activation inresponse to washing-related pictures in the right ventrolateralPFC (BA 47). Taken together, these findings suggest a role for theventrolateral PFC in mediating enhanced responses to disgust-related visual stimuli in women that seems to be irrevocablylinked to their greater disgust sensitivity and could make themmore vulnerable to suffer from disgust-related psychiatric disor-ders.

Our results are at odds with a previous study that found nosex-related differences in brain activation in response to similardisgusting stimuli (Schienle et al. 2005). It is possible that thecontent of the stimuli in the two studies was somewhat different,the current study possibly containing more pictures of mutilatedbodies than Schienle’s study. Other methodological differences,such as methods of fMRI data analysis (i.e., use of parametric vs.nonparametric statistics), as well as other potentially confound-ing variables such as disgust sensitivity (men in our sample hadunusually low scores on disgust sensitivity; Jon Haidt, Ph.D.,personal communication August 2006), and position in the men-strual cycle might also contribute to these inconsistencies.

This study had several limitations. We did not acquire SCRdata from all the participants, and the numbers of women andmen might have been too small to detect any significant differ-ences in SCR, although others also failed to find such differences(Wrase et al. 2003). During the experiment, we asked theparticipants how anxious rather than how disgusted they felt. Wedid not control for our female participants’ position in themenstrual cycle, which has been reported to modulate the neuralresponses to affectively negative stimuli (Goldstein et al. 2005).Although the IAPS pictures were rated as highly disgusting, thesestimuli were never designed to induce particular types of emo-tion and therefore our results might reflect non-specific effects ofnegative emotion.

To summarize, the present results suggest that, in adultvolunteers, there are significant sex-related differences in brainresponses to disgusting stimuli that are intimately related to thewell known sex differences in disgust sensitivity. Disgust sensi-

tivity in complex interaction with biological sex could be an

www.sobp.org/journal

ip

WGt

dHti

B

B

B

B

C

C

D

D

d

F

G

G

G

G

H

H

H

H

K

L

470 BIOL PSYCHIATRY 2007;62:464–471 X. Caseras et al.

w

mportant diathesis or risk factor in the etiology of disgust-relatedsychiatric disorders.

This study was funded by Project Grant No. 064846 from theellcome Trust United Kingdom to MLP, DM-C, and AS and arant from the South London and Maudsley NHS Trust (SLaM)

o DM-C.We are grateful to Mrs. Sarah Wooderson for her help with

ata collection. We also thank Qazi Rahman, Ph.D., and Jonaidt, Ph.D., for their useful comments on an earlier version of

his manuscript. None of the authors have any conflicts ofnterest.

eck AT, Steer RA (1987): Beck Depression Inventory Manual. San Antonio,Texas: The Psychological Corporation.

rammer MJ, Bullmore ET, Simmons A, Williams SCR, Grasby PM, Howard RJ,et al. (1997): Generic brain activation mapping in fMRI: A nonparametricapproach. Magn Reson Imaging 15:763–770.

ullmore ET, Brammer MJ, Rabe-Hesketh S, Curtis VA, Morris RG, Williams SC,et al. (1999): Methods for diagnosis and treatment of stimulus-correlatedmotion in generic brain activation studies using fMRI. Hum Brain Mapp7:38 – 48.

ullmore ET, Long C, Suckling J, Fadili J, Calvert GA, Zelaya F, et al. (2001)Coloured noise and computational inference in neurophysiological(fMRI) time series analysis: Resampling methods in time and waveletdomains. Hum Brain Mapp 12:61–78.

ahill L (2006): Why sex matters for neuroscience. Nat Rev Neurosci 7:477–484.

ritchley HD, Wiens S, Rotshtein P, Ohman A, Dolan RJ (2004): Neural sys-tems supporting interoceptive awareness. Nat Neurosci 7:189 –195.

avey GC (1994): Self-reported fears to common indigenous animals in anadult UK population: The role of disgust sensitivity. Br J Psychol 85:541–554.

enys D, de Geus F, van Megen HJ, Westenberg HG (2004): Use of factoranalysis to detect potential phenotypes in obsessive-compulsive disor-der. Psychiatry Res 128:273–280.

e Jong PJ, Vorage I, van den Hout MA (2000): Counterconditioning in thetreatment of spider phobia: Effects on disgust, fear and valence. BehavRes Ther 38:1055–1069.

riman O, Borga P, Lundberg P, Knutsson H (2003): Adaptive analysis of fMRIdata. Neuroimage 19:837– 845.

oldstein JM, Jerram M, Poldrack R, Ahern T, Kennedy DN, Seidman LJ,Makris N (2005): Hormonal cycle modulates arousal circuitry in womenusing functional magnetic resonance imaging. J Neurosci 25:9309–9316.

oldstein JM, Seidman LJ, Horton NJ, Makris N, Kennedy DN, Caviness VS Jr.,et al. (2001): Normal sexual dimorphism of the adult human brain as-sessed by in vivo magnetic resonance imaging. Cereb Cortex 11:490 –497.

ood CD, Johnsrude I, Ashburner J, Henson RN, Friston KJ, Frackowiak RS(2001): Cerebral asymmetry and the effects of sex and handedness onbrain structure: A voxel-based morphometric analysis of 465 normaladult human brains. Neuroimage 14:685–700.

oodwin RD, Gotlib IH (2004): Sex differences in depression: The role ofpersonality factors. Psychiatry Res 126:135–142.

aidt J, McCauley C, Rozin P (1994): Individual differences in sensitivity todisgust: A scale sampling seven domains of disgust elicitors. Pers IndivDiff 16:701–713.

amann S, Herman RA, Nolan CL, Wallen K (2004): Men and women differ inamygdala response to visual sexual stimuli. Nat Neurosci 7:411– 416.

arenski CL, Hamann S (2006): Neural correlates of regulating negativeemotions related to moral violations. Neuroimage 30:313–324.

asler G, LaSalle-Ricci VH, Ronquillo JG, Crawley SA, Cochran LW, Kazuba D,et al. (2005): Obsessive-compulsive disorder symptom dimensions showspecific relationships to psychiatric comorbidity. Psychiatry Res 135:121–132.

ring AM, Gordon AH (1998): Sex differences in emotion: Expression, expe-rience, and physiology. J Pers Soc Psychol 74:686 –703.

ang PJ, Bradley MM, Cuthbert BN (1997): International Affective PictureSystem (IAPS). New York: NIMH Center for Study of Emotion and Atten-

tion.

ww.sobp.org/journal

Lawrence NS, An SK, Mataix-Cols D, Ruths F, Speckens A, Phillips MA (inpress). Neural responses to facial expressions of disgust but not fearare modulated by washing symptoms in OCD. Biol Psychiatry [Epubahead of print].

Marks IM, Mataix-Cols D (2004): Diagnosis and classification of phobias: Areview. In: Maj M, Akiskal HS, Lopez-Ibor JJ, Okasha A, editors. Phobias.WPA Series, Evidence and Experience in Psychiatry, vol 7. London: JohnWiley & Sons, 1–32.

Mataix-Cols D, Rauch SL, Manzo P, Jenike MA, Baer L (1999): Use of factor-analyzed symptom dimensions to predict outcome with serotonin re-uptake inhibitors and placebo in the treatment of obsessive-compulsivedisorder. Am J Psychiatry 156:1409 –1416.

Mataix-Cols D, Wooderson S, Lawrence N, Brammer MJ, Speckens A, PhillipsML (2004): Distinct neural correlates of washing, checking, and hoardingsymptom dimensions in obsessive-compulsive disorder. Arch Gen Psy-chiatry 61:564 –576.

Matchett G, Davey GC (1991): A test of a disease-avoidance model of animalphobias. Behav Res Ther 29:91–94.

McClure EB, Monk CS, Nelson EE, Zarahn E, Leibenluft E, Bilder RM, et al.(2004): A developmental examination of sex differences in brain en-gagement during evaluation of threat. Biol Psychiatry 55:1047–1055.

Miller GA, Chapman JP (2001): Misunderstanding analysis of covariance.J Abn Psychol 110:40 – 48.

Monk CS, Nelson EE, McClure EB, Mogg K, Bradley BP, Leibenluft E, et al.(2006): Ventrolateral prefrontal cortex activation and attentional bias inresponse to angry faces in adolescents with generalized anxiety disor-der. Am J Psychiatry 163:1091–1097.

Mulkens SA, de Jong PJ, Merckelbach H (1996): Disgust and spider phobia.J Abnorm Psychol 105:464 – 468.

Ogawa S, Lee TM, Kay AR, Tank DW (1990): Brain magnetic resonance imag-ing with contrast dependent on blood oxygenation. Proc Natl Acad SciU S A 87:9868 –9872.

Olatunji BO, Lohr JM, Sawchuk CN, Tolin DF (2007): Multimodal assessmentof disgust in contamination-related obsessive-compulsive disorder. Be-hav Res Ther 45:263–276.

Olatunji BO, Sawchuk CN, Arrindell WA, Lohr JM (2005): Disgust sensitivity asa mediator of the sex differences in contamination fears. Pers Indiv Diff38:713–722.

Olatunji BO, Williams NL, Sawchuk CN, Lohr JM (2006): Disgust, anxiety andfainting symptoms associated with blood-injection-injury fears: A struc-tural model. J Anxiety Disord 20:23– 41.

Paus T, Otaky N, Caramanos Z, MacDonald D, Zijdenbos A, D’Avirro D, et al.(1996): In vivo morphometry of the intrasulcal gray matter in the humancingulate, paracingulate, and superior-rostral sulci: Hemispheric asym-metries, sex differences and probability maps. J Comp Neurol 376:664 –673.

Phillips ML, Williams LM, Heining M, Herba CM, Russell T, Andrew C, et al.(2004): Differential neural responses to overt and covert presenta-tions of facial expressions of fear and disgust. Neuroimage 21:1484 –1496.

Phillips ML, Young AW, Scott SK, Calder AJ, Andrew C, Giampietro V, et al.(1998): Neural responses to facial and vocal expressions of fear anddisgust. Proc Biol Sci 265:1809 –1817.

Phillips ML, Young AW, Senior C, Brammer M, Andrew C, Calder AJ, et al.(1997): A specific neural substrate for perceiving facial expressions ofdisgust. Nature 389:495– 498.

Pujol J, Lopez A, Deus J, Cardoner N, Vallejo J, Capdevila A, Paus T (2002):Anatomical variability of the anterior cingulate gyrus and basic dimen-sions of human personality. Neuroimage 15:847– 855.

Rasmussen SA, Eisen JL (1992): The epidemiology and clinical features ofobsessive compulsive disorder. Psychiatr Clin North Am 15:743–758.

Rauch SL, Savage CR, Alpert NM, Miguel EC, Baer L, Breiter HC, et al.(1995): Apositron emission tomographic study of simple phobic symptom prov-ocation. Arch Gen Psychiatry 1995 52:20 –28.

Rorden C, Brett M (2000): Stereotaxic display of brain lesions. Behav Neurol12:191–200.

Rozin P, Fallon AE (1987): A perspective on disgust. Psychol Rev 94:23– 41.Rozin P, Haidt J, McCauley C, Dunlop L, Ashmore M (1999): Individual differ-

ences in disgust sensitivity: Comparisons and evaluations of paper-and-pencil versus behavioural measures. J Res Pers 33:330 –351.

Rozin P, Haidt J, McCauley CR (2000): Disgust. In: Lewis M, Haviland JM,

editors. Handbook of Emotions. New York: Guilford Publications.

R

S

S

S

S

S

S

X. Caseras et al. BIOL PSYCHIATRY 2007;62:464–471 471

ubia K, Smith AB, Brammer MJ, Taylor E (2003): Right inferior prefrontalcortex mediates response inhibition while mesial prefrontal cortex isresponsible for error detection. NeuroImage 20:351–358.

abatinelli D, Flaisch T, Bradley MM, Fitzsimmons JR, Lang PJ (2004): Affec-tive picture perception: sex differences in visual cortex? Neuroreport15:1109 –1112.

awchuk CN, Lohr JM, Tolin DF, Lee TC, Kleinknecht RA (2000): Disgustsensitivity and contamination fears in spider and blood-injection-injuryphobias. Behav Res Ther 38(8):753–762.

chienle A, Schafer A, Stark R, Walter B, Vaitl D (2005): Relationship betweendisgust sensitivity, trait anxiety and brain activity during disgust induc-tion. Neuropsychobiology 51:86 –92.

hafritz KM, Collins SH, Blumberg HP (2006): The interaction of emotionaland cognitive neural systems in emotionally guided response inhibition.Neuroimage 31:468 – 475.

hapira NA, Liu Y, He AG, Bradley MM, Lessig MC, James GA, et al. (2003):Brain activation by disgust-inducing pictures in obsessive-compulsivedisorder. Biol Psychiatry 54:751–756.

pielberger CD, Gorsuch RL, Lushene RE, Vagg PR, Jacobs GA (1983): Manual

for the State-Trait Anxiety Inventory, Palo Alto, California: Consulting Psy-chologists Press.

Sprengelmeyer R, Rausch M, Eysel UT, Przuntek H (1998): Neural structuresassociated with recognition of facial expressions of basic emotions. ProcBiol Sci 265:1927–1931.

Talairach J, Tournoux P (1998): Co-Planar Stereotactic Atlas of the HumanBrain. Stuttgart, Germany: Thieme.

Wager TD, Phan KL, Liberzon I, Taylor SF (2003): Valence, sex, and lateraliza-tion of functional brain anatomy in emotion: A meta-analysis of findingsfrom neuroimaging. Neuroimage 19:513–531.

Wicker B, Keysers C, Plailly J, Royet JP, Gallese V, Rizzolatti G (2003): Both of usdisgusted in my insula: The common neural basis of seeing and feelingdisgust. Neuron 40:655– 664.

Wilhelm K, Parker G, Hadzi-Pavlovic D (1997): Fifteen years on: Evolvingideas in researching sex differences in depression. Psychol Med 27:875– 883.

Woody SR, Tolin DF (2002): The relationship between disgust sensitivity andavoidant behavior: Studies of clinical and nonclinical samples. J AnxietyDis 16:543–559.

Wrase J, Klein S, Gruesser SM, Hermann D, Flor H, Mann K, et al. (2003): Sexdifferences in the processing of standardized emotional visual stimuli in

humans: A functional magnetic resonance imaging study. Neurosci Lett348:41– 45.

www.sobp.org/journal