Journal of Obsessive-Compulsive and Related Disorders 1 (2012) 189–195

Contents lists available at SciVerse ScienceDirect

Journal of Obsessive-Compulsive and Related Disorders

2211-36

http://d

n Corr

E-m

jessicag

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

The relationship between body dysmorphic concerns and the effects of imagesuppression: Implications for models of body dysmorphic disorder

Melissa Onden-Lim n, Jessica R. Grisham

School of Psychology, University of New South Wales, Sydney, NSW 2052, Australia

a r t i c l e i n f o

Article history:

Received 12 December 2011

Received in revised form

2 May 2012

Accepted 18 May 2012Available online 26 May 2012

Keywords:

Body dysmorphic disorder

Body image

Thought suppression

Imagery

49/$ - see front matter & 2012 Elsevier Ltd. A

x.doi.org/10.1016/j.jocrd.2012.05.001

esponding author. Tel.: þ612 938 53196; fax

ail addresses: mondenlim@psy.unsw.edu.au (

@unsw.edu.au (J.R. Grisham).

a b s t r a c t

A key clinical feature of body dysmorphic disorder (BDD) is the use of avoidance behaviours to

minimise discomfort associated with perceived defects in appearance. While overt avoidance, such as

avoidance of social situations, has been well-documented (e.g., Phillips, 2005), covert avoidance, such

as image suppression, has not been explored. This study investigated the role of suppression of negative

self-imagery in the maintenance of dysmorphic concern. Undergraduate participants completed a

thought suppression paradigm with an experimentally-constructed negative self-image as the target.

Dysmorphic concern was associated with how distressing and vivid an appearance-related intrusion

was, although it did not affect whether participants engaged in suppression of the intrusion.

Instructions to suppress the image led to reduced intrusion frequency and discomfort but it did not

affect the quality (e.g., vividness) of intrusions. In addition, participants high in dysmorphic concern

were more likely to internalise distorted appearance-related information and be disgusted by that

information. Implications for models of body dysmorphic disorder are discussed.

& 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Body dysmorphic disorder (BDD) is a disorder characterised bya distressing or impairing preoccupation with a perceived defectin appearance that is often accompanied by repetitive or compul-sive behaviour directed toward the defect. BDD shares manysimilarities with OCD in clinical features, demographic features,psychiatric comorbidity and similar treatment response to ser-otonin reuptake inhibitor medication (e.g., Phillips et al., 2007).For this reason BDD has widely been considered to be an OCDspectrum disorder.

BDD patients have been found to engage in a number ofbehavioural avoidance strategies intended to reduce anxiety andother negative emotional states associated with negative judge-ments of their own appearance (e.g., Didie, Kelly, & Phillips, 2010).While these overt avoidance patterns, such as mirror avoidance andsocial withdrawal, have been well-documented, more covert meth-ods of avoidance have not yet been investigated. Thought suppres-sion is one such method by which individuals with BDD mayattempt to avoid internal reminders of their appearance flaws.

BDD patients have been shown to report more vivid andnegative intrusive appearance-related images that are viewed

ll rights reserved.

: þ612 938 5364.

M. Onden-Lim),

from an observer (third-person) perspective compared to controls(Osman, Cooper, Hackmann, & Veale, 2004). In addition, preli-minary data from our laboratory suggests that body dysmorphicdisorder patients and undergraduates high in dysmorphic concernengage in more suppression of appearance-related imagery thananxious control participants and undergraduates lower in dys-morphic concern (Onden-Lim & Grisham, Unpublished). We positthat suppression of these appearance-related images may con-tribute to the maintenance of the disorder.

A large body of research on thought suppression suggests thatefforts to suppress unwanted cognitions are likely to be ineffec-tive and ultimately maladaptive, since unwanted thoughts canparadoxically increase after a period of suppression (Abramowitz,Tolin, & Street, 2001; Wenzlaff & Wegner, 2000). Furthermore,increases in target thought frequency are often accompanied bynegative emotions such as frustration and distress, which in turnfuel further suppression efforts and other avoidance behaviours(Purdon, 1999; Salkovskis & Campbell, 1994). These mechanismshave been shown to play a major role in the maintenance of otherdisorders (e.g., Purdon, 1999; Purdon, Rowa & Antony, 2005).

Much of this previous thought suppression research, however,has failed to distinguish between cognitions that are primarilyverbal in nature and those that are imagery-based. Semantic andimagery-based cognitions differ qualitatively (Paivio, 1971) and intheir relation with memories and emotions (Holmes & Mathews,2010; Holmes, Mathews, Mackintosh, & Dalgeish, 2008). Theverbal thought-imagery distinction may be particularly important

M. Onden-Lim, J.R. Grisham / Journal of Obsessive-Compulsive and Related Disorders 1 (2012) 189–195190

in BDD since appraisals about the self are often based on internalperceptual representations of actual and ideal selves (Veale &Neziroglu, 2010). Furthermore, imagery appears to be moreclosely tied to the affective system than verbal thoughts(Holmes & Mathews, 2010). Thus, the overarching aim of thecurrent study was to conduct a preliminary investigation of therole of suppression of negative self-images in the maintenanceof BDD.

In order to accomplish this aim, we induced negative self-images in an undergraduate population using a distorted photo-graph of each participant. First, we examined the associationbetween level of dysmorphic concern and reaction to the negativeself-image prior to any explicit instructions. Second, we investi-gated the effect of experimental instructions to suppress theinduced negative self-imagery.

We hypothesised that dysmorphic concern would predictmore frequent intrusions of negative self-imagery, more effortsto suppress that imagery, and more discomfort experiencedduring intrusions. We also predicted that higher dysmorphicconcern would be associated with more vivid, longer intrusionsand more use of the observer perspective in line with Osmanet al.’s (2004) study, as well as more disgust toward and reportedbelievability of the negative self-imagery. Disgust and believabil-ity of an induced negative self-image were introduced as addi-tional variables based on BDD case reports of high levels ofreported disgust toward appearance and the tendency for patientsto use distorted non-mirror reflecting surfaces for feedback aboutthe appearance (Phillips, 2005; Veale & Riley, 2001). We predictedthat the above associations would remain significant controllingfor general psychopathology and baseline mood.

Consistent with previous thought suppression research (e.g.,Abramowitz et al., 2001), we hypothesised that suppression ofnegative self-imagery would result in more frequent intrusions inthe post-suppression period, and more discomfort experiencedduring intrusions both during and after the suppression period.We also predicted that suppression would decrease the durationof intrusions, since they would be terminated earlier, but notaffect their quality (vividness and perspective).

2. Method

2.1. Participants

Ninety-two first year undergraduate students (67 female) were recruited from

the University of New South Wales and participated in exchange for course credit.

Mean age of the students was 19.92 (SD¼2.61) years and mean years in education

was 14.19 (SD¼1.41). The majority of the sample was Asian (n¼39) or Anglo-

saxon (n¼32), with the remaining participants describing themselves as mixed or

other race (n¼22). Two participants who met diagnostic criteria for BDD were

excluded for ethical reasons.

2.2. Materials

2.2.1. Distorted photograph

A portrait photograph of the participant displaying a neutral expression was

digitally altered to include a prominent dry white wart in the nasal region of the

face (lateral side or tip of nose, depending on facial structure). The modified

photograph [17�23 cm] was presented to the participant against a black back-

ground on-screen.

2.2.2. Body Dysmorphic Disorder Diagnostic Module for Adults (BDD Diagnostic

Module; Phillips, 2005)

This is a brief semi-structured clinician-administered diagnostic measure that

maps onto the DSM-IV diagnostic criteria for BDD. The BDD Diagnostic Module has

excellent interrator reliability (kappa¼ .96; Phillips, 2005).

2.2.3. Body Image Concern Inventory (BICI; Littleton, Axsom & Pury, 2005)

The BICI is a 19-item self-report measure of dysmorphic concern. That is,

concern about, and preoccupation with, a perceived defect in appearance,

excessive checking or camouflaging of the defect, social avoidance and reassur-

ance seeking. It has good internal consistency in college and multiethnic commu-

nity samples (Cronbach’s alpha¼ .93–.94), good concurrent validity (r¼ .83, BDDE-

SR; r¼ .60, BDD Y-BOCS), reasonable convergent validity with eating disorder

measures (r¼ .62), and has excellent sensitivity (96%) (Littleton et al., 2005;

Littleton & Breitkopf, 2008). The BICI was also selected because it was developed

using an undergraduate population (Littleton et al., 2005). In the current study

Cronbach’s alpha was .91.

2.2.4. 21-Item Depression Anxiety Stress Scales (DASS21;

Lovibond & Lovibond, 1995)

The DASS21 consists of three 7-item self report scales taken from the full

version of the Depression Anxiety Stress Scales. The subscales show good

concurrent validity, excellent construct validity and excellent reliability (Cron-

bach’s alphas¼ .94 for depression, .87 for anxiety, and .91 for stress; Antony,

Bieling, Cox, Enns, & Swinson, 1998; Clara, Cox, & Enns, 2001). Furthermore, each

of these subscales haven been shown to tap a more general dimension of

psychological distress (Henry & Crawford, 2005). In this study Cronbach’s alphas

for the depression, anxiety and stress subscales were. 83, .78, and .83, respectively.

2.2.5. Monitor forms

There were four self-report forms in total. These contained questions asking

participants to rate their current mood (10-point scale), discomfort experienced

when the intrusions occurred (11-point scale), and effort to suppress intrusions

during the Monitor/Suppression periods (10-point scale). Participants also were

asked to report how many times intrusions about the target story or image

occurred; how many times intrusions related to the target story or image

occurred; and to report the vividness (10-point scale), perspective (as if from

one’s own eyes vs. as if from a third person point of view; 7-point scale), average

duration of intrusions (s), believability of the modified photo (10-point scale), and

how disgusting they found the modified photo (10-point scale).

2.3. Procedure

Fig. 1 shows the flow of participants through each stage of the experiment.

Participants signed statements of informed consent and were tested individually

by one female graduate level researcher. The experimenter first administered the

BDD Diagnostic Module. If participants met criteria for BDD they were debriefed

and the experiment was terminated. The remaining participants stood next to a

white backdrop and the experimenter took a picture of the participant’s face and

neck (neutral expression). Participants then completed the BICI and reported their

demographic information. BICI scores were collected at the start of the experiment

for matching purposes. Meanwhile, the researcher uploaded and digitally altered

the image to include a prominent crusty wart on nasal region of the face using the

programme Paint.net. Next, the researcher guided participants through a brief

practice visualisation task of cutting a lemon to engage participants in the use of

imagery. This procedure has been adopted in previous studies (e.g. Holmes,

Mathews, Dalgeish, & Mackintosh, 2006).

Afterward, the researcher asked participants take part in another visualisation

task and that they should ‘‘try to imagine the scenario as vividly as possible’’. The

researcher read aloud the following passage:

‘‘Imagine you are entering the university food court. You are on your way to a

group meeting over lunch with 4 other students. You feel quite hot and flustered

because you’ve been hurrying in order to get there on time. As you enter you

notice that it is very crowded and noisy with the sound of people chatting away.

The wonderful smells of food in the air make your stomach rumble and your

mouth water. After a short look around you spot your group members sitting at a

table at the back. You head over to greet them. As you say hi, you notice two of

them looking at you strangely. One of them asks you, ‘‘What’s that on your nose?’’

You run your hand across your face and feel a large pea-sized lump. It feels dry,

rough and bumpy. Hurriedly, you head to the bathroom. You look into the mirror

and this is what you seey’’

The researcher immediately presented the participants’ own modified images

(target image) back to the participants on a computer screen for 10 s, a period

intended to be sufficient for participants to engage with the image without

becoming bored. If a participant looked away from the screen during this time, the

researcher gently directed his or her gaze back to the image. At the end of the

viewing time, all participants were instructed to monitor their thoughts for 3 min

and press the space bar if they experienced a thought about the story or target

image during this time (Baseline, see Fig. 1). The key press procedure was utilised

in this study because it is less intrusive than other means (e.g., verbalising one’s

stream of consciousness) and may therefore be less likely to affect response rates

and incur problems of participants wanting to appear compliant with the

instructions (Purdon, 2004).

Monitor instructions. ‘‘During the next three minutes, you will be asked to

monitor your thoughts. During this time you may think about anything you like,

Allocated to Monitor group (n = 45) ♦ Monitored cognitions for 3 min

♦ Completed ratings of discomfort experienced during intrusions and effort to suppress intrusions

Allocated to Suppress group (n = 45) ♦ Suppressed cognitions for 3 min

♦ Completed ratings of discomfort experienced during intrusions and effort to suppress intrusions

♦ Monitored cognitions for 3 min ♦ Completed ratings of discomfort

experienced during intrusions, effort to suppress intrusions, and current mood

Baseline

Experimental phase

Picture taken and modified (n = 90)

Demographic information and BICI

Practice and target guided imagery tasks

Shown modified photograph for 10 s

Informed consent (n = 92)

♦ Monitored cognitions for 3 min

♦ Completed ratings of discomfort experienced during intrusions and effort to suppress intrusions

♦ Recorded the average duration (s) ofintrusions, and ratings of vividness, perspective, believability, disgust

♦ Completed DASS21

Post

Group allocation

Preparation phase

Excluded ♦ Met diagnostic criteria for BDD (n = 2)

Fig. 1. Flow of participants through each stage of the experiment.

M. Onden-Lim, J.R. Grisham / Journal of Obsessive-Compulsive and Related Disorders 1 (2012) 189–195 191

including thoughts related to the story and the image of yourself with a wart on

your face. If one of these thoughts or a similar thought should come into your

mind, please press the space bar, and continue to let all thoughts flow into your

mind. It is important that you try not to control your thoughts but simply allow all

thoughts to come into your mind freely. Anytime you have a thought about the

story, or the picture of yourself with the wart, please press the space bar. Please

continue in the same way for the full three minutes.’’

After this time the participants completed ratings of discomfort experienced

during intrusions, effort to suppress intrusions and indicated their current mood.

Next, during the Experimental Phase (see Fig. 1), the researcher repeated the same

instructions as before to half of the participants. These participants comprised the

Monitor Group. For the other half, the Suppression Group, the experimenter read

aloud the following Suppression Instructions:

Suppress instructions.‘‘During the next three minutes, try as hard as you can to

suppress all thoughts related to the story you just heard, including thoughts

related to the image of yourself with the wart on your face. If one of these thoughts

or a similar thought should come into your mind, please press the space bar and

try to remove it from your mind immediately. It is important that you try as hard

as you can to remove any thoughts about the story or image of yourself from your

mind whenever they occur. Anytime you have a thought about the story, or the

picture of yourself with the wart, please press the space bar. Please continue this

way for the full three minutes.’’

After 3 min, all participants completed ratings of discomfort experienced

during intrusions and effort to suppress intrusions. Then the researcher read

aloud the Monitor Instructions again to all participants and participants pressed

the space bar whenever they had an intrusion about the story or target image

(Post, see Fig. 1). After 3 min, the participants repeated ratings of discomfort

during intrusions and effort to suppress intrusions. They also recorded the average

duration (s) of intrusions and provided ratings of intrusion vividness, perspective,

believability of the modified photograph and disgust experienced in viewing the

modified photograph. Upon completion of the above tasks, participants completed

the DASS21 and were debriefed about the aims of the study.

3. Results

3.1. Assumptions

Outliers were identified and excluded prior to conductingstatistical procedures. Outliers were defined as cases with stan-dardised scores in excess of 3.29 (po .001, two-tailed test), asrecommended by Tabachnick and Fidell (2007). Five outliers wereremoved on this basis. To meet the assumption of normality, a logtransformation was applied to the intrusion duration, relatedimage, intrusion frequency, discomfort and suppression effortdata. Assumptions for multivariate normality, linearity, andhomogeneity of variances and covariances were met. Mauchly’stests were conducted on suppression effort, intrusion frequencyand discomfort, and indicated that the assumption of sphericity

M. Onden-Lim, J.R. Grisham / Journal of Obsessive-Compulsive and Related Disorders 1 (2012) 189–195192

had not been violated (respectively, w2¼5.48, p¼ .07; w2

¼2.22,p¼ .33; w2

¼4.01, p¼ .13).

3.2. Group assignment and manipulation checks

Participants were matched across groups for BICI scores andtime of day. Each group contained 45 participants, and the meanBICI scores for the Suppress and Monitor groups were 43.3(SD¼11.9) and 43.2 (SD¼12.01), respectively. DASS21 scores forthe Suppress and Monitor groups were 12.6 (SD¼9.0) and 13.0(SD¼8.8), respectively.

Independent samples t-tests and w2 were used to examinewhether any group differences existed in sex, age, years ofeducation, time of testing, baseline mood, and scores on the BICIand DASS21. There were no significant group differences on anyof these variables. Further, there was no significant differencebetween the groups in how realistic participants perceived theactual distorted computerised photo to be, t(88)¼�1.01, p¼ .32.Mean believability scores for the Suppress and Monitor groupswere 6.48 (SD¼2.29) and 5.96 (SD¼2.61) out of 10, respectively.

Participant efforts to suppress were also analysed to ensurethe condition instructions were followed. A one way 3 (time)�2(group) repeated measures multivariate analysis of variance (RM-MANOVA) was conducted to investigate the impact of suppres-sion and monitoring (control) on suppression effort, intrusionfrequency and discomfort experienced during intrusions. Resultsfor intrusion frequency and discomfort are presented later inSection 3.7. There was a main effect of time, Wilks’ L¼ .27, F(6,74)¼32.70, po .001, and a significant time by group interaction,Wilks’ L¼ .32, F(6, 74)¼25.95, po .001. Follow-up pairwisecomparisons confirmed that the Suppression group (M¼7.77,

Table 2Hierarchical multiple regression analyses predicting frequency, suppression effort an

(Baseline).

Frequency Su

Predictor DR2 b DR

Step 1 .10n .14

Baseline mood .22n

DASS21 .27n

Step 2 o .01 � .03 .02

BICI

Total R2 .10n .16

n po .05.nn po .01.

Table 1Zero order correlations.

Variable 1 2 3 4 5 6

1. BICI – .40nn� .03 .09 .28nn .3

2. DASS21 – � .17 .23n .29nn .2

3. Baseline mood – .05 � .06 �

4. Intrusion frequency – .28nn .2

5. Suppression effort – .3

6. Discomfort –

7. Vividness

8. Perspective

9. Duration

10. Believability

11. Disgust

12. Related intrusion

n po .05.nn po .01.

SD¼2.56) attempted to suppress their intrusions more than theMonitor group (M¼3.02, SD¼2.09) during the Experimental Phase,

F(1, 84)¼67.83, po .01. As expected, there were no groupdifferences in suppression effort at Pre, F(1, 84)¼2.28, p¼ .14, orPost, F(1, 84)¼3.32, p¼ .07.

3.3. The relationship between dysmorphic concern and intrusion

frequency, suppression efforts and discomfort prior to manipulation

(baseline)

We hypothesised that dysmorphic concern would predicthigher frequency, greater spontaneous efforts to suppress andmore discomfort during intrusions at Baseline, controlling forbaseline mood and general psychopathology (depression, anxietyand stress symptoms). Eighty-seven (95%) participants reportedexperiencing intrusions of the target image at Baseline. Data fromthe two groups were collapsed since both received identicalprocedures at this time point.

Separate hierarchical multiple regression analyses (MRAs)were used to test whether dysmorphic concern predicted intru-sion frequency, suppression effort, and discomfort, controlling forbaseline mood and DASS21 scores. For each dependent variable,baseline mood and DASS21 scores were entered as predictors inthe first step and BICI was entered in the second step (see Table 1for zero order correlations and Table 2 for results of the MRAs).

In regards to frequency, the model as a whole predicted 9.7% ofthe total variance, F (3, 81)¼2.90, p¼ .04. In the final model, onlybaseline mood (B¼1.02, SE B ¼ .51, b¼ .22, t¼2.01, p¼ .047) andDASS21 scores (B¼ .19, SE B¼ .08, b¼ .27, t¼2.28, p¼ .03) signifi-cantly predicted frequency at Baseline. In analyses of suppressioneffort, the final model predicted 16.2% of variance, F (3, 81)¼5.21,

d discomfort from baseline mood, DASS21, and BICI prior to the manipulation

ppression Effort Discomfort

2 b DR2 b

nn .05

.22n o .01

.25n .08

.10nn .34nn

.17

nn .14nn

7 8 9 10 11 12

7nn .34nn o .01 .12 .35nn .48nn .07

1n .05 � .08 .06 .26n .27n .07

.37nn .01 � .15 � .04 � .13 � .37nn� .04

8nn .11 o .01 .06 .04 .17 .24n

4nn .34nn .14 � .05 .24n .26n .26n

.26n .08 .05 .21n .64nn .04

– � .13 .11 .32nn .33nn .34nn

– � .05 � .09 .10 � .06

– � .02 � .19 .06

– .27n .02

– .10

–

Table 3Hierarchical Multiple Regression Analyses Predicting Intrusion Vividness, Perspective, Duration, Believability and Disgust.

Vividness Perspective Duration Believability Disgust

Predictor DR2 b DR2 b DR2 b DR2 b DR2 b

Step 1 .04 .01 o .01 .07 .08n

Baseline mood .17 � .06 � .02 � .08 � .17

DASS21 � .07 � .10 o .01 .13 .06

Step 2 .10n .01 .08 .09n .19n

Group o .01 .10 – .11 .12

BICI .35nn .04 .26n .30n .46nn

.12

Total R2 .15n .02 .08 .15nn .28nn

n po .05.nn po .01.

M. Onden-Lim, J.R. Grisham / Journal of Obsessive-Compulsive and Related Disorders 1 (2012) 189–195 193

po .01. Similar to frequency, only baseline mood (B¼ .39, SE

B¼ .19, b¼ .22, t¼2.09, p¼ .04) and DASS21 scores (B¼ .07, SE

B¼ .03, b¼ .25, t¼2.20, p¼ .03) significantly predicted suppressioneffort at Baseline. Finally, in regards to discomfort experiencedduring intrusions, the final model predicted 14.3% of the totalvariance, F (3, 81)¼4.52, po .01. Only BICI scores significantlypredicted discomfort (B¼ .08, SE B¼ .03, b¼ .34, t¼3.01, po .01).

3.4. The relationship between dysmorphic concern, suppression and

the quality of intrusions

Several t-tests were conducted to examine the hypotheses thatintrusions would be shorter in the Suppression group comparedto the Monitor group but no different in vividness or perspective.Consistent with hypotheses, there was no difference between thegroups in terms of intrusion vividness, t (84)¼� .17, p¼ .87, orperspective, t (81)¼� .87, p¼ .39. Intrusion duration was signifi-cantly shorter in the Suppression group compared to the Monitorgroup, t (84)¼2.54, p¼ .01.

Separate hierarchical MRAs were conducted to investigatewhether dysmorphic concern and experimentally instructed sup-pression predicts intrusion vividness, duration and perspective,controlling for general psychopathology, baseline mood (seeTable 3 for results of the MRAs). For each dependent variable,baseline mood and DASS21 were entered as predictors in the firststep, and group and BICI scores were entered as predictors in thesecond step.

For vividness, the final model accounted for 14.8% of thevariance, F (4, 77)¼3.33, p¼ .01. Only BICI uniquely predictedvividness (B¼ .06, SE B¼ .02, b¼ .35, t¼3.05, po .01). Regardingperspective, the final model failed to significantly account for thevariance, F (4, 75)¼ .36, p¼ .84. Similarly, the final model did notaccount for the variance in intrusion duration, F (4, 77)¼1.66,p¼ .17; however, group membership did uniquely predict dura-tion (B¼� .11, SE B¼ .05, b¼� .26, t¼�2.34, p¼ .02).

3.5. The relationship between dysmorphic concern and

interpretation of the distorted photo

As confirmed in the manipulation checks, there were nosignificant group differences in how realistic (believable) partici-pants perceived the actual distorted computerised photo was. Inaddition, there was no group difference in how disgustingparticipants thought the distorted photo was, t (81)¼�1.05,p¼ .30.

Separate hierarchical MRAs were applied to investigate thehypotheses that dysmorphic concern predicts how realistic/believable the distorted photo was and how disgusted they wereby it, over and above baseline mood and general psychopathology

(see Table 3 for results of the MRAs). For both analyses, baselinemood and DASS21 scores were entered in the first step, and groupand BICI scores were entered in the second step.

In the analysis looking at believability, the final model pre-dicted 15.4% of the total variance, F (4, 81)¼3.69, po .01. OnlyBICI uniquely predicted believability (B¼ .06, SE B¼ .02, b¼ .30,t¼2.65, p¼ .01). In regard to disgust, the model accounted for27.7% of the total variance, F (4, 76)¼7.27, po .01). Similarly, onlyBICI (B¼ .08, SE B¼ .02, b¼ .46, t¼4.32, po .01) uniquelyaccounted for the variance in disgust.

3.6. Enhancement and rebound effects for intrusion frequency and

discomfort

A one way 3 (time)�2 (group) RM-MANOVA was conducted toinvestigate the impact of suppression and monitoring (control) onsuppression effort, intrusion frequency and discomfort experiencedduring intrusions. There was a main effect of time, Wilks’ L¼ .27, F

(6, 74)¼32.70, po .01, and a significant time by group interaction,Wilks’ L¼ .32, F (6, 74)¼25.95, po .01. Results for suppressioneffort (manipulation check) are presented in Section 3.2.

Looking at intrusion frequency, there was a main effect of time, F

(2, 158)¼60.20, po .01, such that the frequency of intrusionsdecreased over time. There was also a significant time by groupinteraction, F (2, 158)¼11.12, po .01. Follow-up orthogonal simpleinteractions showed that from Baseline to the Experimental Phase

there was a greater decline in intrusion frequency in the Suppressgroup relative to the Monitor group, F (1, 79)¼23.88, po .01. On theother hand, there was no interaction in intrusion frequency betweenthe Experimental Phase and Post, F (1, 79)¼1.66, p¼ .20.

There was a main effect of time for discomfort experiencedduring intrusions, F (2, 158)¼52.65, po .01, such that levels ofdiscomfort decreased over time. There was also a significant time bygroup interaction, F (2, 158)¼3.74, p¼ .03. Similar to the intrusionfrequency data, follow-up orthogonal simple interactions revealed agreater decline in discomfort in the Suppress group relative to theMonitor group from Baseline to the Experimental Phase, F (1,79)¼8.66, po .01, but no interaction in discomfort between theExperimental Phase and Post, F (1, 79)¼ .17, p¼ .68.

4. Discussion

The purpose of this study was to investigate the potential roleof suppression of appearance-related imagery in body dys-morphic concerns, via naturalistic and experimental methodsusing an undergraduate population. The primary objective wasto investigate the degree to which dysmorphic concern predictedsuppression effort, intrusion frequency, and discomfort

M. Onden-Lim, J.R. Grisham / Journal of Obsessive-Compulsive and Related Disorders 1 (2012) 189–195194

experienced during intrusions, as well as the vividness, durationand perspective of those intrusions. The second aim was toinvestigate the degree to which dysmorphic concern predictedhow believable and disgusting the distorted photograph wasperceived to be. The final objective was to experimentallyexamine the effect of suppression on discomfort and intrusionfrequency, as well as intrusion vividness, perspective andduration.

We found that dysmorphic concern predicted vividness anddiscomfort experienced during intrusions but not suppression effort,frequency, perspective or duration. Dysmorphic concern also pre-dicted the degree to which participants viewed the distortedphotograph as believable and how disgusted they were by it. Thefindings are consistent with Osman et al. (2004), who found thatBDD patients experience more vivid and distressing imagery thancommunity controls. Our results also suggest that this vivid anddistressing imagery may in part a result of the greater likelihood ofindividuals with high levels of dysmorphic concern to internalisedistorted appearance-related information (because they view it asrealistic) and experience disgust toward it. Interestingly, dysmorphicconcern was not associated with the perspective taken duringintrusions and duration, thereby failing to support the hypothesisthat higher dysmorphic concern is associated with use of theobserver perspective, and longer intrusion duration.

Experimentally-instructed suppression reduced intrusion fre-quency, discomfort and duration but did not affect intrusionvividness or perspective. Together with the above results, thesefindings suggest that while dysmorphic concern affects howdistressing and vivid an appearance-related intrusion will be, itdoes not affect intrusion frequency or whether one engages insuppression to reduce distress associated with that intrusion.However, suppression of a negative self-image may reduce intru-sion frequency and discomfort in the short term, although it willnot affect the quality (vividness and perspective) of intrusions.The finding that intrusions were significantly shorter in theSuppression group suggests that this group terminated intrusionsearlier than the Monitor group, and is consistent with lesserintrusion discomfort in this group.

These findings have interesting implications for the role ofcovert avoidance in BDD. They show that suppression of appear-ance-related imagery reduces intrusion frequency and discomfort,at least in the short term. If image suppression is negativelyreinforced by this decrease in discomfort, it may function as asafety behaviour for those high in dysmorphic concern. Thefindings are also consistent with models of body image andBDD (e.g., Cash, 2011; Veale & Neziroglu, 2010) that predict thatgreater body image investment and negative evaluation leads tonegative appraisals of appearance-related information, which inturn leads to more distress experienced during intrusions. How-ever, it does not support the models’ prediction that negativeappraisals lead to more use of thought suppression (or, covertavoidance). Our study found that only baseline mood and generalpsychopathology uniquely predicted intrusion frequency andsuppression effort. In view of the models, this suggests intrusionfrequency and suppression effort may not be due to investment inand negative evaluations of body image but rather stem frommore general factors. One reason why mood may be important isthat cognitions consistent with current mood states (e.g., negativeself-images and negative mood) may be more difficult to managethan thoughts inconsistent with current mood state (e.g.,Reynolds & Salkovskis, 1992; Wenzlaff, Wegner, and Klein, 1991).

The failure to find increases in intrusion frequency during andafter experimental instructions to suppress is inconsistent withmuch of the literature on thought suppression of personallyrelevant cognitions (e.g., Abramowitz et al., 2001; Purdon,1999). One possible reason for this divergence may be due to

methodological differences. In the present study, the target storyand modified photograph were presented once only before thefirst monitoring period (Baseline), whereas in other thoughtsuppression studies the target thought is presented duringinstructions at the beginning of each monitor/suppress period(e.g., ‘‘you may think about a white bear, but you don’t have to’’;Tolin, Abramowitz, Przeworski, & Foa, 2002, pp. 1261). Thus, astime progressed, the target story and image may have faded suchthat any immediate and post-suppression rebound effects couldnot be observed. Evidence in support of this assertion is thefinding that intrusion frequency decreased over time for bothgroups.

To counter fading of the target story and image, future studiesmay wish to incorporate a target story and image reminder (e.g.,re-presentation of the modified photo) prior to the commence-ment of each monitor/suppress period. Furthermore, investiga-tions into level of suppression success, level of personal andemotional relevance (in this case, dysmorphic concern and mood),and multiple periods of suppression may be important in experi-mental investigations, as these have been shown to modulate theemergence of immediate and post-suppression rebound effects(e.g., Corcoran & Woody, 2009; Reynolds & Salkovskis, 1992;Williams & Moulds, 2007). Should rebound effects fail to beobserved following such manipulations, this may indicate thatthe effects of image suppression in body image preoccupation donot follow standard rebound patterns as observed in other statescharacterised by obsessive or recurrent cognitions (e.g., OCD andother obsessive-compulsive spectrum disorders). However, giventhat this study was conducted on a non-clinical sample, it isnecessary to replicate findings within a clinical BDD sample. Asalluded to above, rebound effects may be evident in this popula-tion since target images (e.g., images of one’s deformed appear-ance) are likely to be of greater personal and emotional relevancein BDD populations than in non-clinical ones.

One limitation of the present study is that the target story hasovert implications for negative social evaluation (e.g., being in abusy food court). As a result it is difficult to separate whetherintrusions represented incidence of negative self-imagery orincidence of fears of negative social evaluation that are indepen-dent of dysmorphic concern. Given that we were interested inimplications for models of BDD we attempted to address theconcerns that are typical to this disorder. One key feature in thephenomenology of BDD is the high degree of social evaluativeconcern and shame surrounding the perceived defect (e.g., Veale& Neziroglu, 2010). Social concerns are also a key component ofdysmorphic concern (Castle, Molton, Hoffman, Preston, & Phillips2004). As this was the first study to address negative imagerywith application to BDD models, we elected to include a socialcomponent. However, this was at the cost of being unable toseparate out social evaluative concerns that were independent ofdysmorphic concern. Future studies may wish to modify andcompare target visualisations with and without social compo-nents in order to isolate their relative effect. In addition, futurestudies may also wish to investigate the actual content ofthoughts and images being suppressed in order to allow identi-fication of cognitions focused on body dysmorphic tendencies andthose better characterised by social anxiety.

5. Conclusion and future directions

In sum, the present study demonstrated a possible role forsuppression of appearance-related imagery in BDD, and providedsome initial evidence as to the nature of internal distortedappearance-related imagery. A noteworthy contribution of thisstudy is the use of a novel, standardised facial disfigurement

M. Onden-Lim, J.R. Grisham / Journal of Obsessive-Compulsive and Related Disorders 1 (2012) 189–195 195

added to participants’ photographs to serve as a component of thesuppressed imagery. This approach permitted greater experimen-tal control over the target concern because participants did nothave direct (only related) prior experiences with that concern.However, in order to increase applicability of the concern toparticipants, it would be useful for future studies to complementsuch studies by using idiosyncratic concerns as part of thesuppressed imagery.

The current study also extended the suppression literaturethrough the investigation of more qualitative factors, includingintrusion vividness, perspective and duration. Future researchshould investigate the effect of longer-term suppression anddelineate what individual characteristics distinguish whether ornot one engages in suppression as a strategy to respond tonegative thoughts and images.

Acknowledgements

This research was supported by an Australian PostgraduateAward of the Australian Federal Government awarded to the firstauthor.

References

Abramowitz, J. S., Tolin, D. F., & Street, G. P. (2001). Paradoxical effects of thoughtsuppression: A meta-analysis of controlled studies. Clinical Psychology Review,21, 683–703.

Antony, M. M., Bieling, P. J., Cox, B. J., Enns, M. W., & Swinson, R. P. (1998).Psychometric properties of the 42-item and 21-item versions of the Depres-sion Anxiety Stress Scales in clinical groups and a community sample.Psychological Assessment, 10, 176–181.

Cash, T. F. (2011). Cognitive-behavioral perspectives on body image. In T. F. Cash, &L. Smolak (Eds.), Body Image: A Handbook of Science, Practice and Prevention (pp.39–47). New York, NY: Guilford Press.

Castle, D. J., Molton, M., Hoffman, K., Preston, N. J., & Phillips, K. A. (2004).Correlates of dysmorphic concern in people seeking cosmetic enhancement.Australian and New Zealand Journal of Psychiatry, 38, 439–444.

Clara, I. P., Cox, B. J., & Enns, M. W. (2001). Confirmatory factor analysis of theDepression-Anxiety-Stress Scales in depressed and anxious patients. Journal ofPsychopathology and Behavioral Assessment, 23, 61–67.

Corcoran, K. M., & Woody, S. R. (2009). Effects of suppression and appraisals onthought frequency and distress. Behaviour Research and Therapy, 47,1024–1031.

Didie, E. R., Kelly, M. M., & Phillips, K. A. (2010). Clinical features of bodydysmorphic disorder. Psychiatric Annals, 40, 310–316.

Henry, J. D., & Crawford, J. R. (2005). The short-form version of the DepressionAnxiety Stress Scales (DASS-21): Construct validity and normative data in alarge non-clinical sample. British Journal of Clinical Psychology, 44, 227–239.

Holmes, E. A., & Mathews, A. (2010). Mental imagery in emotion and emotionaldisorders. Clinical Psychology Review, 30, 349–362.

Holmes, E. A., Mathews, A., Dalgeish, T., & Mackintosh, B. (2006). Positiveinterpretation training: Effects of mental imagery versus verbal training onpositive mood. Behavior Therapy, 37, 237–247.

Holmes, E. A., Mathews, A., Mackintosh, B., & Dalgeish, T. (2008). The causal effectof mental imagery on emotion assessed using picture-word cues. Emotion, 8,

395–409.Littleton, H. L., Axsom, D., & Pury, C. L. S. (2005). Development of the body image

concern inventory. Behaviour Research and Therapy, 43, 229–241.Littleton, H., & Breitkopf, C. R. (2008). The body image concern inventory:

Validation in a multiethnic sample and initial development of a Spanish

language version. Body Image, 5, 381–388.Lovibond, S. H., & Lovibond, P. F. (1995). Manual for the depression anxiety stress

scales (2nd Ed). Sydney, Australia: Psychology Foundation.Onden-Lim, M., & Grisham, J.R. (2011). [Intrusive imagery experiences in a high

dysmorphic concern population].Unpublished raw data.Osman, S., Cooper, M., Hackmann, A., & Veale, D. (2004). Spontaneously occurring

images and early memories in people with body dysmorphic disorder.Memory, 12, 428–436.

Paivio, A. (1971). Imagery and verbal processes. New York: Holt, Rinehart, andWinston.

Phillips, K. A. (2005). The broken mirror: Understanding and treating body dys-

morphic disorder. New York: Oxford University Press.Phillips, K. A., Pinto, A., Menard, W., Eisen, J. L., Mancebo, M., & Rasmussen, S. A.

(2007). Obsessive-compulsive disorder versus body dysmorphic disorder: Acomparison study of two possibly related disorders. Depression and Anxiety, 24,

399–409.Purdon, C. (1999). Thought suppression and psychopathology. Behaviour Research

and Therapy, 37, 1029–1054.Purdon, C. (2004). Empirical investigations of thought suppression in OCD. Journal

of Behavior Therapy and Experimental Psychiatry, 35, 121–136.Purdon, C., Rowa, K., & Antony, M. M. (2005). Thought suppression and its effects

on thought frequency, appraisal and mood state in individuals with obsessive-compulsive disorder. Behaviour Research and Therapy, 43, 93–108.

Reynolds, M., & Salkovskis, P. M. (1992). Comparison of positive and negativeintrusive thoughts and experimental investigation of the differential effects ofmood. Behaviour Research and Therapy, 30, 273–281.

Salkovskis, P. M., & Campbell, P. (1994). Thought suppression induces intrusionin naturally occurring negative intrusive thoughts. Behaviour Research and

Therapy, 32, 1–8.Tabachnick, B. G., & Fidell, L. S. (2007). Using Multivariate Statistics (5th ed.).

Boston: Pearson Education.Tolin, D. F., Abramowitz, J. S., Przeworski, A., & Foa, E. B. (2002). Thought suppression in

obsessive-compulsive disorder. Behaviour Research and Therapy, 40, 1255–1274.Veale, D., & Neziroglu, F. (2010). Body dysmorphic disorder: A treatment manual.

Chichester: Wiley-Blackwell.Veale, D., & Riley, S. (2001). Mirror, mirror on the wall, who is the ugliest of them

all? The psychopathology of mirror gazing in body dysmorphic disorder.

Behaviour Research and Therapy, 39, 1381–1393.Wenzlaff, R. M., & Wegner, D. M. (2000). Thought suppression. Annual Review of

Psychology, 51, 59–91.Wenzlaff, R. M., Wegner, D. M., & Klein, S. B. (1991). The role of thought

suppression in the bonding of thought and mood. Journal of Personality and

Social Psychology, 60, 500–508.Williams, A. D., & Moulds, M. L. (2007). Investigation of the indulgence cycles

hypothesis of suppression on experimentally induced visual intrusions indysphoria. Behaviour Research and Therapy, 45, 2780–2788.