A quantitative analysis of facial emotion recognition in obsessive–compulsive disorder

Alexander Robert Daros a, Konstantine K. Zakzanis a,n, Neil Alexander Rector b,c

a Department of Psychology, University of Toronto Scarborough, Toronto, Ontario, Canadab Sunnybrook Research Institute, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canadac Department of Psychiatry, University of Toronto, Toronto, Ontario, Canada

a r t i c l e i n f o

Article history:Received 12 March 2013Received in revised form24 November 2013Accepted 29 November 2013Available online 7 December 2013

Keywords:Affect perceptionEmotion recognitionDisgustAngerEmotion processing

a b s t r a c t

Obsessive–Compulsive Disorder (OCD) is characterized by persistent and unwanted obsessions generallyaccompanied by ritualistic behaviors or compulsions. Previous research proposed specific disgust facialemotion recognition deficits in patients with OCD. This research however, remains largely inconsistent.Therefore, the results of 10 studies contrasting facial emotion recognition accuracy in patients with OCD(n¼221) and non-psychiatric controls (n¼224) were quantitatively reviewed and synthesized usingmeta-analytic techniques. Patients with OCD were less accurate than controls in recognizing emotionalfacial expressions. Patients were also less accurate in recognizing negative emotions as a whole; however,this was largely due to significant differences in disgust and anger recognition specifically. The results ofthis study suggest that patients with OCD have difficulty recognizing specific negative emotions infaces and may misclassify emotional expressions due to symptom characteristics within the disorder. Thecontribution of state-related emotion perception biases to these findings requires further clarification.

Crown Copyright & 2013 Published by Elsevier Ireland Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5142. Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515

2.1. Meta-analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5152.2. Literature search and inclusion criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5162.3. Data extraction and transformations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516

3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5163.1. Study characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5163.2. Demographic and clinical characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5163.3. Emotion recognition accuracy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5173.4. Associations with demographics, clinical characteristics, and testing format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517

4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520

1. Introduction

Obsessive–Compulsive Disorder (OCD) is characterized by persis-tent and unwanted obsessions generally accompanied by ritualistic

behaviors or compulsions performed to reduce the anxiety ofobsessions. The disorder affects between 2% and 3% of adults and1% of children and adolescents (Flament et al., 1988; Karno et al.,1988). OCD can be a chronic illness when untreated and as many as50% of adult cases developed during childhood (Rasmussen andEisen, 1990; Karno and Golding, 1991). According to theWorld HealthOrganization, OCD is one of the top 10 causes of disability (Murrayand Lopez, 1996), demonstrating its serious impact on quality of life.The symptoms experienced in OCD can be heterogeneous, ranging

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/psychres

Psychiatry Research

0165-1781/$ - see front matter Crown Copyright & 2013 Published by Elsevier Ireland Ltd. All rights reserved.http://dx.doi.org/10.1016/j.psychres.2013.11.029

n Correspondence to: Department of Psychology, University of Toronto Scarbor-ough, 1265 Military Trail, Toronto, Ontario, Canada M1C 1A4.Tel.: þ1 416 287 7424; fax: þ1 416 287 7642.

E-mail address: zakzanis@utsc.utoronto.ca (K.K. Zakzanis).

Psychiatry Research 215 (2014) 514–521

from fears of contamination, repeated checking, time-consumingand distressing ordering of objects, and distressing intrusive fearspertaining to harming oneself to name but a few.

Within the past two decades, the appraisal and interpretationof “disgusting” objects and situations by persons with OCDhas been intensely studied. Disgust differs from other negativeemotions in that it has evolved from a basic evolutionary sense inavoiding distaste and contamination to one that includes higherorder constructs such as moral reasoning and interpersonal rejec-tion (Rozin and Fallon, 1987). Cognitive appraisals that derive froman excessive experience of disgust may increase perceptions ofrejection towards objects and situational triggers and therebyincrease anxiety to cues that do not normally elicit anxiety.Individuals with OCD demonstrate heightened disgust sensitivityto symptom-specific stimuli and experience a higher degree ofdistress associated with these stimuli compared to healthy indivi-duals (Mataix-Cols et al., 2004, 2008). Disgust sensitivity has beenhypothesized to contribute to the development and maintenanceof anxiety to contamination-based stressors in the form of obses-sions and compulsions (Woody and Teachman, 2000; Olatunjiet al., 2010) because of its association with a broad range of OCDsymptoms, including rumination and checking (Mancini et al.,2001), and doubting, checking, and slowness (Schienle et al.,2003). There is also evidence that disgust sensitivity may bemalleable with exposure-based interventions, although less read-ily than anxiety during the therapeutic process (Smits et al., 2002;McKay, 2006; Cougle et al., 2007; Olatunji et al., 2009). Therefore,how individuals with OCD experience disgust in relation to objectsand cues remains a potentially important vulnerability factorassociated with the disorder.

Given the salience of stimuli related to the emotion of disgustin OCD, it would be hypothesized that they should demonstratevigilance or selective attention toward such stimuli. Research hassupported an attentional bias towards threat in many fear-basedanxiety disorders (e.g., panic disorder, post-traumatic stressdisorder, and social phobia), although studies attempting toreplicate this effect in OCD have been mixed (Williams et al.,1997; Armstrong and Olatunji, 2012). Some researchers havesuggested that attentional bias to threat may only occur in asubset of patients with OCD, namely those with contamination-based fear symptoms that clearly also involve the experience ofdisgust. A preliminary analog study demonstrated that individualshigh in contamination-based fears demonstrate excessive visualfixation on emotional facial expressions of disgust and fear(Armstrong et al., 2010). One might hypothesize that individualswith OCD, or at least a subset of them, would demonstrateenhanced recognition of disgust facial expressions. The emotionrecognition literature suggests, that there is a reduction in thecorrect classification of disgust facial expressions, however.

The earliest study on the topic, by Sprengelmeyer et al. (1997),reported that all patients with OCD exhibited a marked reductionin the ability to recognize disgust from facial expressions. Severalsubsequent studies failed to clearly detect disgust recognitiondifficulties (Buhlmann et al., 2004; Parker et al., 2004; Allenet al., 2006; Montagne et al., 2008; Bozikas et al., 2009; Jhunget al., 2010). Yet, more recent research has again demonstrated thepresence of disgust facial emotion recognition in OCD (Corcoranet al., 2008; Amiri et al., 2012; Lochner et al., 2012; Rector et al.,2012). Given these equivocal findings, research has focused onmoderator variables that may explain why disgust recognitiondeficits were found in some studies but not others, including thepossibility that the problem exists only in those with severesymptom presentations of OCD (Parker et al., 2004). There havealso been attempts to test the specificity of this finding to patientswith other anxiety disorders. For example, Corcoran et al. (2008)demonstrated that patients with OCD recognized disgust facial

expressions less often than patients with panic disorder, but foundthat only one-third of patients demonstrated this deficit. Rectoret al. (2012) compared groups of patients with OCD, generalizedsocial phobia, and panic disorder and again found a specificreduced accuracy for disgust expression recognition in patientswith OCD only. Furthermore, this study demonstrated thatpatients with OCD, who were treatment responders to cognitive-behavioral therapy, no longer demonstrated difficulty in recogniz-ing disgust and supported a hypothesis that these difficultieswould be state-dependent.

Disgust sensitivity may influence emotion recognition withinOCD. If an individual with the disorder experiences disgust to avariety of situations when others do not, he or she will undoubt-edly experience disgust more often than others and may in turn,modulate learned associations between objects that elicit disgustand reactions to them (Berle and Phillips, 2006). Therefore, toreconcile the disgust sensitivity and emotion recognition litera-ture, it is important to consider how both of these variablesinteract when responding to emotional facial expressions withinOCD. Individuals with OCD may become aroused when viewingfacial expressions of disgust to the point that attentional control isaffected. Participants with poor attentional control could thereforehave difficulty shifting their attention away from disgustingstimuli, which in turn, potentiates affective arousal or preventsdown-regulation of arousal (Adams and Lohr, 2012). Alternatively,vigilance and maintenance of attention on the stimulus couldincrease the state of anxiety experienced by the individual, causingcognitive interference and thereby impairing accurate perception(Armstrong and Olatunji, 2012). Either of these instances couldaffect the process of ascribing the correct emotion to a facialexpression. Finally, the recognition of disgusted facial expressionsis possible through the anterior insula (Fusar-Poli et al., 2009),an area that has been found to be abnormally hyperactive inOCD and correlated with symptom provocation (Husted et al.,2006; Mataix-Cols et al., 2008), suggesting underlying links topathophysiology.

The focus of this meta-analysis seeks to quantitatively synthe-size existing research on facial emotion recognition abilities inpatients with OCD to determine whether statistically reliableimpairments are present and the extent of their magnitude. It ishypothesized that patients with OCD, compared with healthycontrols, will demonstrate reduced accuracy in recognizing emo-tional facial expressions, particularly facial expressions of disgust.This meta-analytic investigation will also help to determine thedegree of specificity of emotional processing, i.e., are all negativeemotions poorly recognized in OCD or is the effect specificto disgust stimuli? We also aim to examine the relations betweenpotentially important moderator variables including age, gender,current mood, severity of OCD, symptom subtype of OCD(e.g., contamination-based symptoms), psychiatric co-morbidity(e.g., depression), medication use, and magnitude of effect repre-senting impaired emotion recognition in the disorder.

2. Materials and methods

2.1. Meta-analysis

Comprehensive Meta-analysis (CMA) Version 2.2.057 (Borenstein et al., 2010)was used to conduct the meta-analysis using a random effects modeling approach.We employed standard meta-analytic techniques to our review of the literature(e.g., Hedges and Olkin, 1985; Rosenthal, 1991). Briefly, the analysis of magnitudewas indexed with the effect size estimate d, the difference between two groupmeans calibrated in pooled standard deviation units (Cohen, 1988). Individual studyresults and relevant moderator variables were abstracted, quantified andassembled into a database for statistical analysis. Moderator variables were thencorrelated with effect size differences in order to test relationships that mayinfluence the magnitude of the effect. Depending on the nature of the moderating

A.R. Daros et al. / Psychiatry Research 215 (2014) 514–521 515

variable, we utilized Pearson's and point-biserial (rpb) to compute correlationalanalyses. Heterogeneity of effect sizes across studies was examined using estimatesof the standard deviation of effect size across studies. The effect sizes were alsotransformed into an overlap percentage (OL%) to articulate the meaningfulness ofan effect size (Zakzanis, 2001). In the present context, the OL% statistic representsthe degree of overlap between patients with OCD and participants in the controlgroup. To ascertain how robust our findings were, we also employed Orwin's (1983)Fail-Safe N (Nfs) formula, to determine the robustness of our findings. Nfs estimatesthe number of additional hypothetical studies needed to overturn the obtainedmean effect size to a small and typically meaningless effect size (i.e., d¼0.10).We calculated this value to assess the possibility of error due to a publication biasof statistically significant studies (i.e., Type I publication bias error; Hedges andOlkin, 1985).

It should be noted that meta-analytic studies are not entirely uncontroversial(Hunter and Schmidt, 1990). Since studies with larger sample sizes have morestatistical power than studies with smaller sample sizes, computations of meaneffect size must be weighted accordingly. The analysis of moderating variables wasdone without attempting to weight the various studies according to sample size(Van Horn and McManus, 1992). This method can therefore seem underpowered,due to the reliance on the number of studies rather than the number of participantsin the actual studies. Finally, given that specific analyses could be based on a subsetof the studies included in the larger meta-analysis, we specified the number ofstudies that contributed to each analysis (N¼number of studies) in parentheses.

2.2. Literature search and inclusion criteria

Studies were selected for inclusion by way of a computerized literature searchof PubMed, PsycInfo, and Google Scholar databases from 1987 (corresponding withthe publication of DSM-III-R) to December 2012. The following keywords wereused in various combinations: obsessive–compulsive disorder, facial emotionrecognition, facial affect recognition, emotion recognition, emotion identification,affect recognition, and affect identification. In addition, a thorough manual searchwas performed using cross-references from original articles and reviews.

Eligible studies were those that included standardized tests of facial emotionrecognition in patients with DSM-III-R or DSM-IV diagnosed OCD and healthycontrol participants. One study was included without the use of a healthy controlgroup (Rector et al., 2012). In this study, a group of patients with OCD who hadsuccessfully completed an intensive 16-week course of cognitive-behavioral ther-apy and had statistically significant symptom reduction (pre-post treatment,d¼2.37) was used as the control group. Furthermore, there were no significantdifferences in this post-treatment OCD group compared to the normative data onthe task used in the study. Standardized facial emotion recognition tasks weredefined as tests that required ascribing a qualitative label, usually from a limitednumber of choices, to the picture of a facial expression. The response format for allstudies was computerized, however the number of emotions utilized within aparticular study for the forced-choice trials could vary. Studies in which twoemotions were blended in one picture or presented successive approximations ofemotional intensities were included only if they included stimuli at 100% intensity.Although task descriptions varied, it is commonly advised that participants performthe task as fast and accurately as possible. However, none of the currently includedstudies reported reaction time data, relying exclusively on accuracy of emotionrecognition and therefore the emphasis in these studies may have, in fact, been onaccuracy. The search was not exclusively limited to English language publications.Studies included in the meta-analysis are marked with an asterisk (n) in theReferences section.

2.3. Data extraction and transformations

Relevant data extracted for this meta-analysis included demographic charac-teristics (age, gender), current psychotropic use, diagnostic comorbidity, symptomrating scales, and emotion recognition task information (test stimuli used, testingformat, number of stimuli presented, recognition accuracy). For studies that did notreport these data, the corresponding author of each study was contacted to requestthe relevant information wherever possible. Performance data for the emotionrecognition tasks were converted to percent accuracy, defined as the ratio of correctresponses to the total number of stimuli. The mean and standard deviation ofaccuracy scores in seven emotion categories (where available) were extracted fromprimary studies: happy, angry, sad, fear, disgust, and surprise. Accuracy scorescollapsing across negative emotions and all emotions were then created. Groupmeans and standard deviations were converted to effect sizes (Cohen's d) with thefollowing interpretations: small (0.2), medium (0.5) and large (Z0.8), although weacknowledge that the interpretation of effect sizes is dependent on context ratherthan these broadly based heuristics (see Zakzanis, 2001). Moderation analysis wasconducted using IBM SPSS Statistics Version 20.0 (IBM, Chicago).

3. Results

3.1. Study characteristics

Our systematic review of the research literature garnered 13potential studies; however, three primary studies were excludedbecause they did not measure emotion recognition at appropriateintensities and/or did not present data that could be appropriatelyextracted (Kornreich et al., 2001; Aigner et al., 2007; Montagne et al.,2008). Therefore, 10 primary studies met inclusion criteria with atotal of 221 patients with OCD and 224 healthy individuals (Table 1).Two variations of emotion recognition were commonly operationa-lized, where the first simply displayed static images of facial emo-tional expressions at 100% intensity. The second variation involvedmorphing the static images into more ambiguous trials but alsoincluded responses to full threshold, static images. For the ambig-uous trials, the intensity of the facial expression is altered from itsoriginal 100% into a combinatorial blend with another emotionalfacial expression. It is important to note that this could affect theaccuracy indices during these trials since emotional expressions withlower intensities (or combinations) are harder to distinguish. There-fore, the approach taken in the our meta-analysis was to run anoverall accuracy analysis independent of task design but thensubsequently separate the results by variation of task. Two studiesreported the accuracy results for both static and morphing taskvariations (Sprengelmeyer et al., 1997; Jhung et al., 2010) andtherefore both sets of data were entered into the analyses withoutduplicating the demographic or clinical characteristics.

3.2. Demographic and clinical characteristics

The mean age of patients with OCD was 30.38 (S.D.¼7.64)years and they were mostly female (53.7%), while healthy controlswere on average 30.86 (S.D.¼8.83) years old and also primarilyfemale (54.1%). The two groups did not differ by age, t(18)¼0.13,P¼0.90, or gender, t(18)¼0.05, P¼0.96. The sample of patientswith OCD (N¼3, M¼111.6, S.D.¼4.94) did not differ on tests ofintelligence from healthy controls (N¼2, M¼116.79, S.D.¼3.27),t(3)¼1.28, P¼0.29. Data regarding psychotropic medications atthe time of the study was largely incomplete. Approximately 59%of patients with OCD were medicated at the time of the testing;

Table 1Characteristics of included studies.

Study Stimuli set Sample sizes Stimulus type

Allen et al. (2006) POFA OCD¼11, HC¼19 MorphAmiri et al. (2012) POFA OCD¼20, HC¼20 StaticBozikas et al. (2009) KAMT OCD¼25, HC¼25 StaticBuhlmann et al. (2004) POFA OCD¼20, HC¼20 StaticCorcoran et al. (2008) POFA OCD¼40, HC¼36 StaticJhung et al. (2010) POFA OCD¼41, HC¼37 BothLochner et al. (2012) ERT OCD¼20, HC¼20 StaticParker et al. (2004) POFA OCD¼15, HC¼15 MorphRector et al. (2012) POFA OCD¼20, HC¼11b StaticSprengelmeyer et al. (1997) POFA OCD¼12, HC¼18 Botha

Notes: POFA, Pictures of Facial Affect (Ekman and Friesen, 1976); KAMT, Kinney AffectMatching Test (Kinney et al., 1995); ERT, Emotion Recognition Task (Montagne et al.,2007). Static refers to tasks presenting stimuli at 100% intensity only while Morphrefers to tasks that used both stimuli at 100% intensity and other stimuli at differingintensities.

a Only standard deviations from the control group were available, therefore forthis study Hedges' g statistics were calculated to obtain effect sizes.

b Post-treatment group of OCD patients, symptom reduction statisticallysignificant with effect size, d¼2.37.

A.R. Daros et al. / Psychiatry Research 215 (2014) 514–521516

however this statistic was based on only three studies (Bozikaset al., 2009; Lochner et al., 2012; Rector et al., 2012). The mostcommon psychotropic medication was antidepressants, of whichtwo studies reported over 50% of patients taking at the time of thestudy. Based on five studies, the symptom dimensions endorsed bypatients with OCD included: contamination/washing (45.2%),checking/ordering (62.1%), obsessional (i.e., aggressive, sexual,religious; 47.6%) and hoarding (18.2%). Patients could rate multiplesymptoms categories and therefore the totals do not add up to100%. Only four studies included self-report measures of depres-sion severity at the time of testing, indicating that patientsgenerally fell within the mild- to-moderately depressed range(Buhlmann et al., 2004; Corcoran et al., 2008; Jhung et al., 2010;Lochner et al., 2012). Most studies excluded patients with exten-sive comorbidity, especially psychotic and mood disorders otherthan depression. Comorbidity of MDD with OCD in the studysample was not commonly reported (N¼5; M¼12.6%, S.D.¼12.03)and approximately 30% of the patient sample had at least oneother co-morbid anxiety disorder at the time of testing.

3.3. Emotion recognition accuracy

All meta-analytic results for emotion recognition accuracy arepresented in Table 2. Considering all basic emotions, patientswith OCD were significantly less accurate than controls in facialemotion recognition, and this effect size difference fell within themedium range (d¼�0.55; N¼11; 95% CI¼�0.92 to �0.19,z¼�2.99, P¼0.03; OL%¼64; Nfs¼50). When dividing the studiesby task, the static variation produced a larger effect size differencecompared to morphing variation. The ability to recognize negativeemotions in general was tested empirically by first calculating

Cohen's d values for each negative emotion (anger, sadness,disgust, and fear) across each study. The standardized d valueswere then averaged for the particular study and then meta-analyzed in a separate data file. Patients with OCD were lessaccurate in recognizing negative emotions as a whole and thisdifference was statistically significant (d¼�0.34; N¼10; 95%CI �0.56 to �0.11, z¼�2.87, Po0.01; OL%¼77; Nfs¼24). Analysisof the individual emotion categories revealed that patients withOCD had the greatest difficulty recognizing disgust (d¼�0.59;N¼11; 95% CI¼�1.06 to �0.11, z¼�2.43, P¼0.02; OL%¼62;Nfs¼54) and anger (d¼�0.36; N¼10; 95% CI¼�0.67 to -0.05,z¼�2.25, P¼0.02; OL%¼76; Nfs¼26). Once again, static taskvariations produced larger effect sizes when breaking down thestudies into the two tasks. There was also a marginally significanteffect between patients with OCD and healthy controls for therecognition of sadness, with patients performing more poorly(d¼�0.31; N¼10; 95% CI¼�0.62 to 0.00, z¼�1.99, P¼0.05;OL%¼78; Nfs¼21). Upon dividing the studies by task designhowever, neither the static nor morphing variations produced asignificant result for sadness individually. Differences betweenpatients with OCD and controls did not reach statisticalsignificance for facial expressions of fear, surprise, or happiness(see Table 2).

We further tested the specificity of impairments in recognizingdisgust in patients with OCD by comparing the mean effect sizefor disgust with negative emotions and their aggregate mean(i.e., anger, sad, fear, and an average of the three). The reducedability for recognizing disgust in patients with OCD was signifi-cantly greater than their decrement in recognizing other negativeemotions, t(9)¼2.63, P¼0.03. However, reduced ability for recog-nizing disgust in patients with OCD was equivalent to decrementsin recognizing anger, t(9)¼1.64, P¼0.16, and emotion recognitiondecrements overall, t(10)¼0.58, P¼0.57.

3.4. Associations with demographics, clinical characteristics, andtesting format

The most significant meta-analytic findings were found foroverall emotion recognition as well as specific anger and disgustrecognition. Therefore, correlational analyses were undertakenusing these three significant results only. For patients with OCD,age was not significantly related to overall accuracy on theemotion recognition task, N¼11, r¼�0.22, P¼0.51. On specificemotions, increasing age appeared to be moderately associatedwith poorer disgust recognition, N¼11, r¼�0.45, P¼0.16, but notanger recognition, N¼10, r¼�0.03, P¼0.94. Gender was alsounrelated to the primary research findings on emotion recognitionaccuracy (rso0.42, Ps40.20). Unexpectedly, the severity ofOCD as measured by the Yale-Brown Obsessive–Compulsive Scale(YBOCS; Goodman et al., 1989) was not significantly associatedwith overall emotion recognition (N¼6, r¼0.33, P¼0.67).Additionally, illness severity was also unrelated to recognition ofanger (N¼5, r¼0.66, P¼0.22) and the recognition of disgust(N¼6, r¼�0.22, P¼0.68) when examined individually.

The relationship between depressed mood at the time oftesting and emotion recognition was also examined. Mean scoreson the Beck Depression Inventory (BDI; Beck et al., 1996) andthe Montgomery-Asberg Depression Rating Scale (MADRS;Montgomery and Asberg, 1979) were divided by their maximumscore in order to equate the scales. Self-reported depressionseverity was neither significantly associated with poorer emotionrecognition ability (N¼4, r¼�0.11, P¼0.88), nor anger or disgustrecognition when examined individually (rso0.27, Ps40.72).Rates of comorbid depression were not associated with emotionrecognition (N¼4; r¼0.14, P¼0.86). The association betweenemotion recognition and psychotropic medications could not be

Table 2Effect sizes comparing patients with OCD and healthy controls on facial emotionrecognition.

Emotion N d SE 95% CI z p OL (%)a

Overall 11 �0.56 0.185 �0.92/�0.19 �2.99 0.03 64Static 7 �0.77 0.233 �1.23/�0.32 �3.32 0.01 53Morph 4 �0.14 0.191 �0.51/0.24 �0.71 0.48 90

Negative 11 �0.34 0.117 �0.56/�0.11 �2.87 o0.01 77Static 7 �0.37 0.152 �0.68/�0.07 �2.45 0.01 76Morph 4 �0.28 0.213 �0.70/0.14 �1.32 0.19 80

Disgust 11 �0.59 0.241 �1.06/�0.11 �2.43 0.02 62Static 7 �0.61 0.152 �0.90/�0.33 �4.23 o0.01 61Morph 4 �0.56 0.682 �1.91/0.77 �0.83 0.41 64

Anger 10 �0.36 0.158 �0.67/�0.05 �2.25 0.02 76Static 6 �0.35 0.246 �0.83/�0.13 �1.48 0.17 76Morph 4 �0.38 0.167 �0.71/�0.06 �2.29 0.02 76

Sadness 10 �0.31 0.158 �0.62/0.00 �1.99 0.05 78Static 6 �0.43 0.248 �0.92/0.05 �1.75 0.08 70Morph 4 �0.15 0.157 �0.46/0.17 �0.97 0.33 87

Fear 10 �0.09 0.094 �0.27/0.10 �0.95 0.36 92Static 6 �0.19 0.118 �0.42/0.04 �1.56 0.11 85Morph 4 0.09 0.156 �0.22/0.39 0.54 0.59 92

Surprise 7 �0.12 0.274 �0.66/0.42 �0.46 0.66 90Static 4 �0.25 0.336 �0.91/0.40 �0.76 0.45 81Morph 3 0.07 0.537 �0.99/1.12 0.12 0.90 95

Happyb 6 �0.34 0.231 �0.78/0.10 �1.51 0.13 77

Notes: Negative values represent poorer scores on emotion recognition tasks forpatients with OCD. Static and Morph refer to two specific paradigm designs used toassess emotion recognition.

a Values are approximations based on Zakzanis (2001).b Too few studies to be further divided by task.

A.R. Daros et al. / Psychiatry Research 215 (2014) 514–521 517

assessed because too few primary studies reporting medicationuse at the time of the study. Employing the POFA stimulus setcompared to other sets had no clear relationship with overallrecognition accuracy or disgust recognition when examinedindividually (N¼11, rpbo�0.25, Ps40.46). Studies that employedstatic emotion recognition paradigms were moderately associatedwith larger effect size differences between groups, N¼11,rpb¼�0.50, P¼0.12. Finally, symptom subtypes as self-reportedby patient participants with OCD were correlated with overallperformance as well as accuracy for anger and disgust facialexpressions. For overall performance and disgust recognition,poorer accuracy was not significant associated with the percentof patients acknowledging symptoms of contamination, checking,or obsessional thoughts (rso�0.25, Ps40.71). For anger recogni-tion, there was a significant association between poorer accuracyand the percent of patients acknowledging obsessional thoughts(r¼�0.97, P¼0.01). It is imperative to interpret this and our othercorrelative results with caution however, as very few studiescontributed data to some of these analyses.

4. Discussion

The present study capitalized on the statistical power affordedby meta-analytic techniques to systematically determine whetherpatients with OCD demonstrate deficits in recognizing facial dis-plays of emotion. Based on a combined sample of 221 OCDpatients and 224 healthy controls, there was an observed deficitin overall emotion recognition accuracy for OCD patients whencollapsing across six basic emotions (happy, angry, sad, disgust,fear, and surprise). The effect size difference for this comparisonfell within the medium range based on Cohen's (1988) conven-tions for interpreting the magnitude of effect sizes. The overlappercentage for this finding is roughly 60%, which means thatroughly 40% of individuals with OCD experience a significantreduction in the ability to recognize facial emotions accurately.This result suggests that there is a general deficiency in theprocessing of emotional facial expressions in patients with OCD,a result found in other forms of psychopathology such as MDD andschizophrenia. Although the severity of OCD, as measured by theYBOCS, was not associated with poorer performance for recogniz-ing facial expressions of disgust in the present study, it remains animportant question in the literature requiring further study(Rector et al., 2012). In addition, further examination of otherindices of symptom severity in the disorder may provide strongerrelationships to difficulties in recognizing emotional facial expres-sions. Emotion recognition difficulties in patients were also inde-pendent of diagnostic co-morbidity with MDD, although theseanalyses, in addition to those based on mood symptom ratingscales were limited because these data were reported too infre-quently in the primary studies. Mood-congruent biases in MDDare associated with significant reductions in recognizing facialexpression of happiness (Bourke et al., 2010). Given that patientswith OCD in the present study did not demonstrate deficits inrecognizing happiness, this finding may represent an importantdistinction between patients with OCD and those with MDD alone,although more research is needed to determine the specificity ofthese results within various clinical populations.

Given the arguments for specific disgust recognition impair-ments in OCD, an analysis of this specificity was undertaken.Results revealed that patients with OCD were significantly worseon negative emotions as a group (disgust, anger, fear, and sadness)when compared to healthy controls. While this result suggests thatpatients with OCD may have difficulties recognizing negativeemotions more generally, the results of the individual emotionsare more compelling as two negative emotions (fear and sadness)

did not achieve significance on an individual examination. Thesefindings suggest that patients with OCD have the most difficultyrecognizing facial expressions of disgust, followed by expressionsof anger, and these difficulties contribute to the overall decre-ments in recognizing emotional expressions in general.

Deficits in recognizing disgust were most pronounced in OCDwhen paradigms were static in nature (i.e., presenting facialexpressions at 100% intensity). However, patients with OCD maydemonstrate improved accuracy to more ambiguous disgust facialexpressions due to sensitivity in perceiving subtle cues of disgust.Indeed, Jhung et al. (2010) reported that patients with OCDwere significantly more likely to perceive disgust and less likelyto perceive anger in ambiguous facial expressions of disgust(i.e., those lower than 100% full intensity expressions). A similarstudy by Montagne et al. (2008) found equivocal results betweenpatients with OCD and healthy controls. This study combinedemotion recognition results for ambiguous and non-ambiguous(i.e., static) stimuli, however. Disgust has become an increasinglyimportant emotion in OCD, particularly because of the increasedimportance placed on the distinction between processes of fearand disgust in the disorder. Patients with OCD usually describetheir feared object or situation as “disgusting” rather than “frigh-tening” (Rozin and Fallon, 1987) suggesting an intimate connectionbetween these affective states within the disorder. An argumentexists as to whether attentional biases to threat are present inOCD, given that this finding has not been consistent in research onsamples of OCD patients, but appears to be a general finding inother anxiety disorders (see Armstrong et al., 2010 for review).The altered cognitive processing of disgust-related informationappears to be relatively specific to OCD as an anxiety disorder,with the possible exception of certain types of phobias (e.g., spiderand blood–injection–injury; Olatunji et al., 2010). Altered cogni-tive processing may have treatment implications, as there isevidence to suggest that disgust is less malleable to exposure-type treatment sessions, while at the same time, anxiety toobsessive stimuli tends to significantly decrease (McKay, 2006;Olatunji et al., 2009).

The underlying pathophysiology of the disorder may contributeto the misappraisal of disgust facial expressions. Abnormalitiesin fronto-striatal regions in patients with OCD, particularly theorbitofrontal cortex, caudate nucleus, and putamen, may beinvolved in mediating emotional characteristics of disgust viaconnections with the insular cortex (McGuire, 1995; Abbruzzeseet al., 1997; Husted et al., 2006). Functional imaging studies of theneural responses to facial expressions of disgust have also foundthat the insula is heavily involved in the perception of disgust-related information, including facial expressions of disgust(Phillips et al., 1997; Fusar-Poli et al., 2009). Patients with OCDalso appear to have difficulty identifying repugnant olfactorystimuli, another function of the insula, in terms of threshold andalso discrimination (Barnett et al., 1999; Segalas et al., 2011).Therefore, given that the neurological substrates of obsessive andcompulsions may compromise neural pathways involved in theperception of disgust-related information, alteration in perceivingand experiencing disgust could be expected in OCD (Husted et al.,2006). This may arise because individuals with OCD fail to developaccurate connections between disgust-related cues and evocationsof disgust, given the chronic course of the disorder that oftenarises in childhood or even adolescence. If individuals with OCDexperience disgust when others do not, they may display inap-propriate facial expressions or misinterpret others when experi-encing disgust, leading to a weakened learned associationbetween emotional experiences of disgust and the correspondingfacial expression displayed by others (Sprengelmeyer et al., 1997;Berle and Phillips, 2006; Husted et al., 2006). However, alterationsin emotion recognition may be state-dependent in OCD due to the

A.R. Daros et al. / Psychiatry Research 215 (2014) 514–521518

fact that normalization of hyperactivities within the orbitofrontalcortex, caudate nucleus, and insular cortex in patients occur inresponse to psychotropic medication and behavior therapy(Baxter et al., 1992; Schwartz et al., 1996; Saxena et al., 2002;Lazaro et al., 2008). As Mataix-Cols et al. (2008) report, insularactivation is still associated with symptom provocation and ispositively correlated with disgust sensitivity in OCD. However,aberrant functional connectivity of these pathophysiologicalregions of OCD may be responsible for both the exaggerateddisgust sensitivity and diminished recognition of facial expressionsof disgust.

Patients with OCD also exhibited significant difficulty labelingfacial expressions of anger. Anger may be less understood as anemotion connected to OCD; however, given the insight requiredfor the diagnosis of the condition, patients understand that theirobsessions and compulsions are excessive and unreasonable andan underlying frustration could prompt increased experiences ofanger. Indeed, there is evidence for increased experiences andinternal expression of anger in patients with OCD compared tohealthy controls (Whiteside and Abramowitz, 2005). Familydynamics may also be disrupted, as uncooperative family mem-bers of patients with OCD tend to increase both their anger andanxiety (Calvocoressi et al., 1995). Alternatively, expressed angerby the patient may also increase familial accommodation andsupport of the disorder, thereby maintaining or exacerbatingsymptoms of OCD. Moscovitch et al. (2008) found that patientswith panic disorder, social phobia, and OCD all had elevated angerexperiences compared to healthy controls. This finding disap-peared when depression was co-varied out of the analyses andtherefore the relationship between OCD and anger may resultfrom comorbid depression or symptoms of general distress(i.e., negative affect) seen across all of the anxiety disorders(Brown et al., 1998). There is evidence that recognizing expressionsof anger is also related to activation in the insula (Fusar-Poli et al.,2009), albeit less selectively than disgust. Increased experiences ofanger within OCD could lead to an internalization process that failsto detect external displays of anger appropriately. Sensitivity inrecognizing angry facial expressions may be present in OCD due tothe fact that these individuals frequently experience anger.An alternative hypothesis may revolve around the fact thatpatients with OCD are overly aroused by facial expressions ofanger due to patients' own realization of the problems theirsymptoms cause for others. Displays of anger may cause auto-nomic arousal and impede performance on recognizing theemotion.

It is important to recognize the limitations of our synthesis ofthe research literature. To begin, many of the primary studies didnot report important clinical and demographic information suchas comorbidity rates, self-report measures of mood, and medica-tion usage. Given that mood and anxiety disorders are prevalentin patients with OCD, the contribution of mood state to emotionperception biases is important consider in this population.Medication usage is equally important to address, given that theuse of antidepressants can improve detection of facial expressionsin patients with depression (Tranter et al., 2009) and healthyindividuals (Harmer et al., 2003; Browning et al., 2007). Interest-ingly, benzodiazepines also appear to selectively impair therecognition of facial expressions displaying anger in healthyvolunteers (Blair and Curran, 1999). Treatment information is alsoimportant to note, as there is evidence for the reversal of impair-ments in the recognition of disgust in OCD following cognitive-behavioral therapy (Rector et al., 2012) and acute escitalopramadministration (Lochner et al., 2012). Longitudinal designs withindividuals diagnosed with OCD may improve both reliability ofthe difficulties in recognizing emotions and provide importanttrajectories of improvements during the course of psychological

and/or pharmacological treatment. Reaction time data may alsoshed further light on hypotheses with regard to moderatorvariables, such that their relationships may become clearer withrespect to emotion recognition accuracy (e.g., speed-accuracytrade-offs due to medication or symptoms of depression).

Meta-analyses themselves are not without their own limitationsand criticisms. Consideration of publication bias was addressed bycalculating fail-safe N values for each meta-analytic statistic, whichrepresents the number of studies that would be required to nullifythe results of significance testing. Inspection of the results suggeststhat our significant findings are robust and unlikely to be stronglyaffected by publication biases, particularly because several publishedstudies included in this meta-analysis contained seemingly “null”results. Nevertheless, additional unpublished studies of null results(i.e., no statistical differences but possible group differences) withrespect to facial emotional recognition in patients with OCD mayexist. The statistical power of this meta-analysis was limited by theamount of information available and while attempts were made togather all pertinent information from all of the primary studies, thiswas not always possible. Our moderation analyses should beconsidered preliminary and exploratory, considering that the avail-able information to conduct such analyses was limited. Futurestudies should plan to assess patient comorbidity, current mood,clinical severity, and current psychotropic use. In addition, research-ers should attempt to recruit patients without a history of psycho-logical or pharmacological treatment to determine the state-dependent nature of emotion recognition deficits.

Despite these limitations, this meta-analysis provides animportant quantitative synthesis of facial emotion recog-nition ability of patients with OCD. Even though recognizingemotional facial expressions is thought to be a universal ability(Ekman, 1992), many psychiatric conditions have been asso-ciated with difficulties in the accurate recognition of emotions.The synthesis of data from primary sources led to the discoverythat patients with OCD demonstrate significantly reduced emo-tion recognition capacities compared to healthy controls. Thiseffect was accompanied by a similar effect for negative emo-tions when grouped as a category and, on an individual basis,disgust and anger. The results pertaining to anger contestedarguments of a specific impairment in the recognition of disgustfacial expressions (Sprengelmeyer et al., 1997). Secondly, ourresults also demonstrated that static paradigms provided morespecificity for emotion recognition deficits in patients with OCDas these studies produced more consistent effect sizes. Coupledwith the plethora of research demonstrating that patients withOCD have difficulty with tasks related to orbitofrontal cortexand striatal areas of the brain, a difficulty for recognizing facialexpressions may also be due to hyperactivation of the frontal-striatal circuit underlying the pathophysiology of OCD. Finally,several new questions emerged from the synthesis of researchon this topic. First, do individuals diagnosed with any anxietydisorder exhibit a significant difference in the perception ofanger, given that increased expressed anger appears to be acommon expression across the anxiety disorders (Moscovitchet al., 2008)? Second, do generalized impairments for therecognition of negative emotions exist in all anxiety disorders,as is sometimes suggested by results in comparative samples(Corcoran et al., 2008; Rector et al., 2012)? Finally, is there aspecific pattern of facial emotion recognition impairments thatmay improve diagnostic validity of mood and anxiety disor-ders? In order to move forward with facial emotion recognitionresearch in mood and anxiety disorders, our findings emphasizethe need for reliable reporting of potential moderating variablesand improved design of emotion recognition tasks in whichboth emotion recognition and sensitivity to emotional facialexpressions are tested.

A.R. Daros et al. / Psychiatry Research 215 (2014) 514–521 519

Acknowledgments

We gratefully acknowledge Maryam Sharif-Razi and her family,who provided written Farsi translations.

Referencesn

Abbruzzese, M., Ferri, S., Scarone, S., 1997. The selective breakdown of frontalfunctions in patients with obsessive–compulsive disorder and in patients withschizophrenia: a double dissociation experimental finding. Neuropsychologia35, 907–912.

Adams, T.G., Lohr, J.M., 2012. Disgust mediates the relation between attentionalshifting and contamination aversion. Journal of Behavior Therapy and Experi-mental Psychiatry 43, 975–980.

Aigner, M., Sachs, G., Bruckmuller, E., Winklbaur, B., Zitterl, W., Kryspin-Exner, I.,Gur, R., Katschnig, H., 2007. Cognitive and emotion recognition deficits inobsessive–compulsive disorder. Psychiatry Research 149, 121–128.

(n) Allen, J.L., Abbott, M.J., Rapee, R.M., Coltheart, M., 2006. Ew gross! Recognition ofexpressions of disgust by children with obsessive–compulβsive disorder.Behaviour Change 23, 239–249.

(n) Amiri, A., Ghasempour, A., Fahimi, S., Abolghasemi, A., Akbari, E., Agh, A.,Fakhari, A., 2012. Recognition of facial expression of emotion in patients withobsessive–compulsive disorder and average people. Armaghane-danesh, YasujUniversity of Medical Sciences Journal 17, 30–38.

Armstrong, T., Olatunji, B.O., 2012. Eye tracking of attention in the affectivedisorders: a meta-analytic review and synthesis. Clinical Psychology Review32, 704–723.

Armstrong, T., Olatunji, B.O., Sarawgi, S., Simmons, C., 2010. Orienting and main-tenance of gaze in contamination fear: biases for disgust and fear cues.Behaviour Research and Therapy 48, 402–408.

Barnett, R., Maruff, P., Purcell, R., Wainwright, K., Kyrios, M., Brewer, W., Pantelis, C.,1999. Impairment of olfactory identification in obsessive–compulsive disorder.Psychological Medicine 29, 1227–1233.

Baxter Jr., L.R., Schwartz, J.M., Bergman, K.S., Szuba, M.P., Guze, B.H., Mazziotta, J.C.,Alazraki, A., Selin, C.E., Ferng, H.K., Munford, P., et al., 1992. Caudate glucosemetabolic rate changes with both drug and behavior therapy for obsessive–compulsive disorder. Archives of General Psychiatry 49, 681–689.

Beck, A.T., Steer, R.A., Brown, G.K., 1996. Beck Depression Inventory, 2nd ed.Psychological Corporation, San Antonio, TX.

Berle, D., Phillips, E.S., 2006. Disgust and obsessive–compulsive disorder: an update.Psychiatry 69, 228–238.

Blair, R.J., Curran, H.V., 1999. Selective impairment in the recognition of angerinduced by diazepam. Psychopharmacology (Berl) 147, 335–338.

Borenstein, M., Hedges, L., Higgins, J., Rothstein, H., 2010. Comprehensive Meta-analysis, version 2.2.057. Biostat, Englewood, NJ.

Bourke, C., Douglas, K., Porter, R., 2010. Processing of facial emotion expression inmajor depression: a review. The Australian and New Zealand Journal ofPsychiatry 44, 681–696.

(n) Bozikas, V.P., Kosmidis, M.H., Giannakou, M., Saitis, M., Fokas, K., Garyfallos, G.,2009. Emotion perception in obsessive–compulsive disorder. Journal of theInternational Neuropsychological Society 15, 148–153.

Brown, T.A., Chorpita, B.F., Barlow, D.H., 1998. Structural relationships amongdimensions of the DSM-IV anxiety and mood disorders and dimensions ofnegative affect, positive affect, and autonomic arousal. Journal of AbnormalPsychology 107, 179–192.

Browning, M., Reid, C., Cowen, P.J., Goodwin, G.M., Harmer, C.J., 2007. A single doseof citalopram increases fear recognition in healthy subjects. Journal of Psycho-pharmacology 21, 684–690.

(n) Buhlmann, U., McNally, R.J., Etcoff, N.L., Tuschen-Caffier, B., Wilhelm, S., 2004.Emotion recognition deficits in body dysmorphic disorder. Journal of Psychia-tric Research 38, 201–206.

Calvocoressi, L., Lewis, B., Harris, M., Trufan, S.J., Goodman, W.K., McDougle, C.J.,Price, L.H., 1995. Family accommodation in obsessive–compulsive disorder. TheAmerican Journal of Psychiatry 152, 441–443.

Cohen, J., 1988. Statistical Power Analysis for the Behavioral Sciences, 2nd ed.Lawrence Erlbaum Associates, Hillsdale, NJ.

(n) Corcoran, K.M., Woody, S.R., Tolin, D.F., 2008. Recognition of facial expressions inobsessive–compulsive disorder. Journal of Anxiety Disorders 22, 56–66.

Cougle, J.R., Salkovskis, P.M., Wahl, K., 2007. Perception of memory ability andconfidence in recollections in obsessive–compulsive checking. Journal ofAnxiety Disorders 21, 118–130.

Ekman, P., 1992. Are there basic emotions? Psychological Review 99, 550–553.Ekman, P., Friesen, W.V., 1976. Pictures of Facial Affect. Available from: ⟨http://www.

paulekman.com/research_cds.php⟩.Flament, M.F., Whitaker, A., Rapoport, J.L., Davies, M., Berg, C.Z., Kalikow, K., Sceery,

W., Shaffer, D., 1988. Obsessive compulsive disorder in adolescence: anepidemiological study. Journal of the American Academy of Child and Adoles-cent Psychiatry 27, 764–771.

Fusar-Poli, P., Placentino, A., Carletti, F., Landi, P., Allen, P., Surguladze, S., Benedetti,F., Abbamonte, M., Gasparotti, R., Barale, F., Perez, J., McGuire, P., Politi, P., 2009.

Functional atlas of emotional faces processing: a voxel-based meta-analysis of105 functional magnetic resonance imaging studies. Journal of Psychiatry andNeuroscience 34, 418–432.

Goodman, W.K., Price, L.H., Rasmussen, S.A., Mazure, C., Fleischmann, R.L., Hill, C.L.,Heninger, G.R., Charney, D.S., 1989. The Yale-Brown obsessive compulsive scale.I. Development, use, and reliability. Archives of General Psychiatry 46,1006–1011.

Harmer, C.J., Bhagwagar, Z., Perrett, D.I., Vollm, B.A., Cowen, P.J., Goodwin, G.M.,2003. Acute SSRI administration affects the processing of social cues in healthyvolunteers. Neuropsychopharmacology 28, 148–152.

Hedges, L.V., Olkin, I., 1985. Statistical Methods for Meta-analysis. Academic Press,New York, NY.

Hunter, J.E., Schmidt, F.L., 1990. Methods of Meta-analysis. Sage Publications,Newbury Park, CA.

Husted, D.S., Shapira, N.A., Goodman, W.K., 2006. The neurocircuitry of obsessive–compulsive disorder and disgust. Progress in Neuro-Psychopharmacology andBiological Psychiatry 30, 389–399.

(n) Jhung, K., Namkoong, K., Kang, J.I., Ha, R.Y., An, S.K., Kim, C.H., Kim, S.J., 2010.Perception bias of disgust in ambiguous facial expressions in obsessive–compulsive disorder. Psychiatry Research 178, 126–131.

Karno, M., Golding, J.M., 1991. Obsessive–compulsive disorder. In: Robins, L.N.,Regier, D.A. (Eds.), Psychiatric Disorders in America: The Epidemiologic Catch-ment Area Study. Free Press, New York, pp. 204–219.

Karno, M., Golding, J.M., Sorenson, S.B., Burnam, M.A., 1988. The epidemiology ofobsessive–compulsive disorder in five US communities. Archives of GeneralPsychiatry 45, 1094–1099.

Kinney, J.M., Fantie, B.D., Pascualvaca, D.M., Rodriguez, D., Mirsky, A.F., 1995.Comprehension of affect in children with pervasive developmental disorders:deficits in matching faces to cartoon situations. Journal of the InternationalNeuropsychological Society 1, 156–157.

Kornreich, C., Blairy, S., Philippot, P., Dan, B., Foisy, M., Hess, U., Le Bon, O., Pelc, I.,Verbanck, P., 2001. Impaired emotional facial expression recognition in alco-holism compared with obsessive–compulsive disorder and normal controls.Psychiatry Research 102, 235–248.

Lazaro, L., Caldu, X., Junque, C., Bargallo, N., Andres, S., Morer, A., Castro-Fornieles, J.,2008. Cerebral activation in children and adolescents with obsessive–compul-sive disorder before and after treatment: a functional MRI study. Journal ofPsychiatric Research 42, 1051–1059.

(n) Lochner, C., Simmons, C., Kidd, M., Chamberlain, S.R., Fineberg, N.A., van Honk, J.,Ipser, J., Stein, D.J., 2012. Differential effects of escitalopram challenge ondisgust processing in obsessive–compulsive disorder. Behavioral Brain Research226, 274–280.

Mancini, F., Gragnani, A., D’Olimpio, F., 2001. The connection between disgust andobsessions and compulsions in a non-clinical sample. Personality and Indivi-dual Differences 31, 1173–1180.

Mataix-Cols, D., An, S.K., Lawrence, N.S., Caseras, X., Speckens, A., Giampietro, V.,Brammer, M.J., Phillips, M.L., 2008. Individual differences in disgust sensitivitymodulate neural responses to aversive/disgusting stimuli. European Journal ofNeuroscience 27, 3050–3058.

Mataix-Cols, D., Wooderson, S., Lawrence, N., Brammer, M.J., Speckens, A., Phillips,M.L., 2004. Distinct neural correlates of washing, checking, and hoardingsymptoms dimensions in obsessive–compulsive disorder. Archives of GeneralPsychiatry 61, 564–576.

McGuire, P.K., 1995. The brain in obsessive–compulsive disorder. Journal ofNeurology, Neurosurgery, and Psychiatry 59, 457–459.

McKay, D., 2006. Treating disgust reactions in contamination-based obsessive–compulsive disorder. Journal of Behavior Therapy and Experimental Psychiatry37, 53–59.

Montagne, B., de Geus, F., Kessels, R.P., Denys, D., de Haan, E.H., Westenberg, H.G.,2008. Perception of facial expressions in obsessive–compulsive disorder: adimensional approach. European Psychiatry 23, 26–28.

Montagne, B., Kessels, R.P., de Haan, E.H., Perrett, D.I., 2007. The Emotion Recogni-tion Task: a paradigm to measure the perception of facial emotional expres-sions at different intensities. Perceptual Motor Skills 104, 589–598.

Montgomery, S.A., Asberg, M., 1979. A new depression scale designed to besensitive to change. The British Journal of Psychiatry 134, 382–389.

Moscovitch, D.A., McCabe, R.E., Antony, M.M., Rocca, L., Swinson, R.P., 2008. Angerexperience and expression across the anxiety disorders. Depression andAnxiety 25, 107–113.

Murray, C.J., Lopez, A.D., 1996. The Global Burden of Disease. World HealthOrganization, Harvard School of Public Health, World Bank, Geneva.

Olatunji, B.O., Cisler, J., McKay, D., Phillips, M.L., 2010. Is disgust associated withpsychopathology? Emerging research in the anxiety disorders. PsychiatryResearch 175, 1–10.

Olatunji, B.O., Wolitzky-Taylor, K.B., Willems, J., Lohr, J.M., Armstrong, T., 2009.Differential habituation of fear and disgust during repeated exposure to threat-relevant stimuli in contamination-based OCD: an analogue study. Journal ofAnxiety Disorders 23, 118–123.

Orwin, R.G., 1983. A fail-safe N for effect size in meta-analysis. Journal ofEducational Statistics 8, 157–159.

(n) Parker, H.A., McNally, R.J., Nakayama, K., Wilhelm, S., 2004. No disgustrecognition deficit in obsessive–compulsive disorder. Journal of BehaviorTherapy and Experimental Psychiatry 35, 183–192.

Phillips, M.L., Young, A.W., Senior, C., Brammer, M., Andrew, C., Calder, A.J.,Bullmore, E.T., Perrett, D.I., Rowland, D., Williams, S.C., Gray, J.A., David, A.S.,n Studies included in the meta-analysis are marked with an asterisk.

A.R. Daros et al. / Psychiatry Research 215 (2014) 514–521520

1997. A specific neural substrate for perceiving facial expressions of disgust.Nature 389, 495–498.

Rasmussen, S.A., Eisen, J.L., 1990. Epidemiology of obsessive compulsive disorder.The Journal of Clinical Psychiatry 51 (Suppl.), S10–S13. (discussion 14).

(n) Rector, N.A., Daros, A.R., Bradbury, C.L., Richter, M.A., 2012. Disgust recognitionin obsessive–compulsive disorder: diagnostic comparisons and posttreatmenteffects. Canandian Journal of Psychiatry 57, 177–183.

Rosenthal, R., 1991. Meta-analytic Procedures for Social Research. Sage, BeverlyHills, CA.

Rozin, P., Fallon, A., 1987. A perspective on disgust. Psychological Review 94, 23–41.Saxena, S., Brody, A.L., Ho, M.L., Alborzian, S., Maidment, K.M., Zohrabi, N., Ho, M.K.,

Huang, S.C., Wu, H.M., Baxter Jr., L.R., 2002. Differential cerebral metabolicchanges with paroxetine treatment of obsessive–compulsive disorder vs majordepression. Archives of General Psychiatry 59, 250–261.

Schienle, A., Schafer, A., Stark, R., Walter, B., Franz, M., Vaitl, D., 2003. Disgustsensitivity in psychiatric disorders: a questionnaire study. Journal of Nervousand Mental Disease 191, 831–834.

Schwartz, J.M., Stoessel, P.W., Baxter Jr., L.R., Martin, K.M., Phelps, M.E., 1996.Systematic changes in cerebral glucose metabolic rate after successful behaviormodification treatment of obsessive–compulsive disorder. Archives of GeneralPsychiatry 53, 109–113.

Segalas, C., Labad, J., Alonso, P., Real, E., Subira, M., Bueno, B., Jimenez-Murcia, S.,Menchon, J.M., 2011. Olfactory identification and discrimination in obsessive–compulsive disorder. Depression and Anxiety 28, 932–940.

Smits, J.A., Telch, M.J., Randall, P.K., 2002. An examination of the decline in fear anddisgust during exposure-based treatment. Behavior Research and Therapy 40,1243–1253.

(n) Sprengelmeyer, R., Young, A.W., Pundt, I., Sprengelmeyer, A., Calder, A.J., Berrios,G., Winkel, R., Vollmoeller, W., Kuhn, W., Sartory, G., Przuntek, H., 1997. Disgustimplicated in obsessive–compulsive disorder. Proceedings of the Royal SocietyB: Biological Sciences 264, 1767–1773.

Tranter, R., Bell, D., Gutting, P., Harmer, C., Healy, D., Anderson, I.M., 2009. The effectof serotonergic and noradrenergic antidepressants on face emotion processingin depressed patients. Journal of Affective Disorders 118, 87–93.

Van Horn, J.D., McManus, I.C., 1992. Ventricular Enlargement in Schizophrenia—ametaanalysis of studies of the Ventricle–Brain Ratio (Vbr). British Journal ofPsychiatry 160, 687–697.

Whiteside, S.P., Abramowitz, J.S., 2005. The expression of anger and its relationshipto symptoms and cognitions in obsessive–compulsive disorder. Depression andAnxiety 21, 106–111.

Williams, J.M., Watts, F.N., MacLeod, C., Matthews, A., 1997. Cognitive Psychologyand Emotional Disorders, 2nd ed. Wiley, Chichester, UK.

Woody, S.R., Teachman, B.A., 2000. Intersection of disgust and fear: normative andpathological views. Clinical Psychology: Science and Practice 7, 291–311.

Zakzanis, K.K., 2001. Statistics to tell the truth, the whole truth, and nothing but thetruth: formulae, illustrative numerical examples, and heuristic interpretation ofeffect size analyses for neuropsychological researchers. Archives of ClinicalNeuropsychology 16, 653–667.

A.R. Daros et al. / Psychiatry Research 215 (2014) 514–521 521