Journal of Abnormal Psychology

The Role of Aesthetic Sensitivity in Body Dysmorphic Disorder

Christina Lambrou, David Veale, and Glenn Wilson Online First Publication, January 31, 2011. doi: 10.1037/a0022300

CITATION Lambrou, C., Veale, D., & Wilson, G. (2011, January 31). The Role of Aesthetic Sensitivity in Body Dysmorphic Disorder. Journal of Abnormal Psychology. Advance online publication. doi: 10.1037/a0022300

Journal of Abnormal Psychology 2011, Vol. , No. , 000 000

2011 American Psychological Association 0021-843X/11/$12.00 DOI: 10.1037/a0022300

The Role of Aesthetic Sensitivity in Body Dysmorphic Disorder

Christina Lambrou, David Veale, and Glenn Wilson
Kings College London
Individuals with a higher aesthetic sensitivity may be more vulnerable to developing body dysmorphic disorder (BDD). Aesthetic sensitivity has 3 components: (a) perceptual, (b) emotional, and (c) evaluative. Individuals with BDD (n 50) were compared with a control group of individuals with an education or employment in art and design related fields (n 50) and a control group of individuals without aesthetic training (n 50). A facial photograph of each participant was manipulated to create a 9-image symmetry continuum. Presented with the continuum on a computer, participants were required to select and rate the image representing their self-actual, self-ideal, idea of perfect, most physically attractive, most pleasure, and most disgust. Control symmetry continua examined the specificity of the disturbance. As predicted, BDD participants displayed no distortion in their perceptual processing but were disturbed in their negative emotional/evaluative processing of their self-image. A significant discrepancy between their self-actual and self-ideal, resulting from an absent self-serving bias in their self-actual (a bias exhibited by controls) appears to be the source of their disturbance. They also overvalued the importance of appearance and self-objectified. These aesthetic evaluations may predispose individuals to BDD and/or maintain the disorder. Keywords: body dysmorphic disorder, aesthetic sensitivity, body image, self-serving bias, depression

The wish to be attractive is a normal desire. Many express dissatisfaction, to some degree, with at least one facet of their appearance. For those with body dysmorphic disorder (BDD), however, the concern with an imagined or slight defect in their appearance is excessive, causing them significant distress and/or impairment in their social and/or occupational functioning. Etiological understanding of BDD is still in its infancy, and it remains enigmatic. That individuals with BDD appreciate art and beauty to a greater degree than comparative psychiatric groups is suggested by their choice of occupation and/or education (Veale, Ennis, & Lambrou, 2002), which raises an interesting question about the definition of BDD as a preoccupation with an imagined defect or a minor physical anomaly. Perhaps individuals with BDD are more aesthetically sensitive than the mental health professionals who diagnose them and who are therefore unable to appreciate art and beauty to the same degree (Veale & Lambrou, 2002).

Aesthetic Sensitivity Model

Aesthetic sensitivity can be defined as an awareness and appreciation of beauty and harmony. Individuals with BDD may be more aware of subtle differences in facial asymmetry or the size of secondary sexual facial characteristics, or they may be better at

evaluating harmony and balance in appearance. This relates to the concept of aestheticality, a term coined by Harris (1982) to describe an innate sensitivity to aesthetic perception, an attribute that varies among individuals; a high aestheticality would augment an individuals self-consciousness and distress over any defect in their appearance, such that they seek cosmetic surgery. Veale et al. (1996) suggested that being more aesthetically sensitive was a possible risk factor in the development of BDD. To our knowledge, this is the first study to investigate whether aesthetic sensitivity may play such a role. As well as expanding on the original theory by Harris (1982) and applying the ideas to BDD, we used novel techniques to test the hypotheses formulated. Akin to the perceptual and affective/attitudinal components of body image, this study proposed that aesthetic sensitivity has three components: (a) perceptual (the ability to differentiate variations in aesthetic proportions); (b) emotional (the degree of emotion experienced when presented with beauty or ugliness); and (c) evaluative (aesthetic standards, values, and identity). The question is whether those with BDD exhibit a perceptual distortion and/or an emotional/evaluative disturbance.

Components of Aesthetic Sensitivity Aesthetic Perceptual Sensitivity

Christina Lambrou, David Veale, and Glenn Wilson, Department of Psychology, Institute of Psychiatry, Kings College London, London, United Kingdom. This study is part of a doctoral dissertation by Christina Lambrou. Correspondence concerning this article should be addressed to Christina Lambrou, Department of Psychology, Box PO77, Institute of Psychiatry, Kings College London, De Crespigny Park, London SE5 8AF, United Kingdom. E-mail: christina.lambrou@kcl.ac.uk 1

Aesthetic perceptual sensitivity has two components: (a) perceptual understanding/awareness (i.e., idea of perfect) and (b) perceptual accuracy (i.e., self-actual). Although research on perceptual understanding/awareness was not available prior to this study, limited empirical research was available for perceptual accuracy. Mainstream opinion is that self-perception in those with BDD is distorted. However, consistent with the hypothesis of this study, Thomas and Goldberg (1995) found that individuals with

LAMBROU, VEALE, AND WILSON

BDD were more accurate in assessing their facial proportions than controls or individuals seeking cosmetic surgery. Note that they explored accuracy in ability, which is not necessarily the same as accuracy in perception. Mirror gazing heightens self-awareness and accuracy (Jerome, 1992) and reduces attractiveness ratings of ones own face (Mulkens & Jansen, 2009) in those dissatisfied with their appearance. When viewing their appearance in the mirror (a common compulsive behavior in BDD), those with BDD may selectively attend to specific features. This may explain their accuracy and the maintenance of their symptoms (Veale, 2004). Depression is the most common comorbid disorder in BDD (e.g., Gunstad & Phillips, 2003). The relationship with depression may account for their perceptual accuracy. The premise of this study, which is based on the idea of depressive realism (Alloy & Abramson, 1979), was that mildly to moderately depressed individuals (this would likely include most individuals with BDD) would be more accurate in their self-actual perception. In contrast, nondepressed individuals would display a self-serving bias. The accurate self-estimations reported by Thomas and Goldberg (1995) may be tapping the perceptual component of body image, which may not be distorted in BDD. In contrast, the perceptual distortion described by some clinicians may be an account of their response to their internal body image (e.g., Osman, Cooper, Hackmann, & Veale, 2004), which involves the emotional/evaluative component of body image and may be disturbed in BDD.

Aesthetic Emotional Sensitivity

Harris (1982) suggested that a consequence of increased aestheticality is that an individual reacts with a greater emotional response to beauty or ugliness. Individuals with BDD experience more aversion to their face than controls (Feusner et al., 2010). When rating their own face or body, the emotional response for those with BDD may be a mixture of self-disgust (e.g., when viewing themselves in a mirror), depression at the failure to achieve an aesthetic standard, and anxiety about the future consequences of being ugly (Veale & Lambrou, 2002). This study assessed the emotional experience in relation to pleasure and disgust to reveal the precise nature of the hypothesized emotional disturbance in BDD. Because of their heightened aestheticality, individuals with BDD may recognize beauty or ugliness on a perceptual level. However, on an emotional level, they may exhibit a negative bias, experiencing less pleasure and more disgust than those without BDD when viewing attractive and unattractive versions of their own face, respectively.

2008), who have higher beauty standards than the rest of the population and demand perfection in themselves as an ideal (Veale & Lambrou, 2002). Relative to controls, those with BDD display a greater discrepancy between their self-actual and their self-ideal (Veale, Kinderman, Riley, & Lambrou, 2003). The question that needs to be addressed is the source of their self-discrepancy. Is it due to a distorted self-actual, an exaggerated self-ideal, or both? Aesthetic standard. The discrepancy between an individuals idea of perfect and their ideal personal standard represents their aesthetic standard. Individuals with BDD may display a minimal discrepancy between their self-ideal and their idea of perfect because they are demanding perfection in their appearance. Attractiveness standard. By the nature of their disorder, individuals with BDD may value attractiveness more than the rest of the population (e.g., Buhlmann, Teachman, Naumann, Fehlinger, & Rief, 2009). This may partly explain why they are more stringent when rating their own attractiveness (e.g., Buhlmann et al., 2008). Aesthetic values. Not everyone who believes they are ugly or that they have a defect in their appearance develops BDD. Individuals with BDD may place a greater value on the importance of appearance in their identity, which would predispose them to and/or perpetuate the disorder. Aesthetic identity. Veale (2004) posited that those with BDD are more likely to judge themselves almost exclusively by their appearance and view the self as an aesthetic object. This process, known as self-objectification, leads to an assiduous engagement in appearance enhancing behaviors, self-consciousness, and body shame (Fredrickson & Roberts, 1997), which characterize BDD.

Present Study
The aim of the present study was to investigate three proposed components of aesthetic sensitivity (perceptual, emotional, and evaluative) to determine whether for those with BDD, their views of their appearance are due to a perceptual distortion and/or an emotional/evaluative disturbance. The central premise was that rather than having a distortion in their perceptual processing, individuals with BDD would have an enhanced understanding of aesthetic proportions and an increased accuracy in their self-actual estimation. The source of their disturbance would be in their emotional/evaluative processing, which is specific to the self. Individuals with an interest in art and beauty, such as those with an education or employment in art and design related fields, might have developed an enhanced understanding or appreciation of aesthetic proportions compared with individuals without aesthetic training. Therefore, art and design controls as well as non-art controls were used. Conceptualized as the reference group, the selection of art and design controls was based on the expectation that they would resemble individuals with BDD in their perceptual processing, expressed as an increased understanding of aesthetic proportions, and that they would resemble non-art controls in their emotional/evaluative processing. Each of the three proposed components of aesthetic sensitivity has a perceptual and emotional/evaluative element to illustrate the dichotomy. For instance, in the assessment of aesthetic emotional sensitivity, participants selected the images that gave them the most pleasure and the most disgust (perceptual) and reported their pleasure and disgust ratings (emotional/evaluative). Two control conditions (other face and building) were included to test whether the hypothesized emotional/evaluative disturbance

Aesthetic Evaluations
Individuals with BDD may hold certain dysfunctional aesthetic evaluations, which interact with aesthetic perceptual and emotional sensitivity to predispose them to and/or maintain the disorder. Aesthetic evaluations divide into three broad categories: (a) aesthetic standards, (b) aesthetic values, and (c) aesthetic identity. Aesthetic standards. There are three types of standards: personal, aesthetic, and attractiveness. Personal standard (self-ideal and self-discrepancy). Perhaps individuals with BDD evaluate their appearance negatively because they are perfectionists (Buhlmann, Etcoff, & Wilhelm,

ROLE OF AESTHETIC SENSITIVITY IN BDD

in those with BDD is specific to their own face. The hypothesized increased perceptual understanding would generalize to other faces (facial control condition) and inanimate objects in the general surroundings, represented by a building (nonfacial control condition). In the control conditions, individuals with BDD would be indistinguishable from art and design controls.

(d) snowball sampling. The ethics committees of the Institute of Psychiatry and the Priory Hospital North London approved the study protocol.

Computer Graphic Techniques

We investigated the three proposed components of aesthetic sensitivity by manipulating a digital photograph of each participants face, the control faces, and the building, using Adobe Photoshop 7.0 and Ulead MorphStudio 1.0 to create nine image symmetry continua. Participants viewed the images in SuperLab on a laptop. Generation of individual test photographs. The first author took a digital colored photograph of each participants face under standardized conditions. Each participant sat upright in front of a white background, with a light source positioned on either side to reduce shadowing. A 3.0 mega-pixel digital camera, placed on a tripod and positioned a constant distance of 0.6 m from the participant, was used to take the frontal view photographs. To eliminate emotional expression as a possible confounding variable, participants received instructions to look directly at the camera and display a neutral facial expression. The male and female individuals recruited to represent the two control facial images were selected because they matched the age criteria of the participants and were judged by 10 volunteers to fall within the average range of attractiveness. The first author took their original digital colored photograph under the same standardized conditions outlined for participants. A building represented the nonfacial inanimate control stimulus to compare with the faces because buildings are comparable with faces in terms of visual complexity; buildings contain internal features, have a global structure, and are, to some degree, symmetrical. The first author used a 3.0 mega-pixel digital camera, placed on a tripod, to generate the original digital colored photograph of the building. The building was Grovelands House, a Grade I listed Regency mansion designed by the celebrated architect John Nash in 1797. Definition of perfection. Empirical evidence implies that attractiveness increases in male and female faces as symmetry is enhanced and that attractiveness decreases as symmetry is reduced (Rhodes, Proffitt, Grady, & Sumich, 1998). For the purposes of this study, perfection was therefore defined by symmetry. To simplify the design, the manipulations were limited to this one dimension of beauty. Symmetry continuum manipulations. The first author created a symmetry continuum of nine images (one real image and eight manipulated images) for each participant, for the male control facial image, for the female control facial image, and for the building image. The real image, that is, the original digital photograph, represented the midpoint on the symmetry continuum. Creating the images representing symmetry and asymmetry. The first author created the two extremes of the continuum, symmetry (most attractive) and asymmetry (least attractive), using Adobe Photoshop 7.0. For all the manipulations, the feather was set at 10 pixels to reduce the visibility of harsh lines in the image from using the lasso or marquee tools to select isolated features or larger sections of the image. Symmetry. To create the facial and building images representing symmetry, the first step was to duplicate a layer of the real

Method Participants
The study sample comprised 150 participants in three groups: (a) 50 individuals with BDD (test group), (b) 50 art and design controls (nonclinical control group), and (c) 50 non-art controls (nonclinical control group). Male and female adults ages 18 through 40 were recruited. This age group was selected because it is representative of the time when individuals are more commonly and most affected by BDD. The groups were age and gender matched. The inclusion criteria for the BDD group were (a) a primary diagnosis of BDD based on the Diagnostic and Statistical Manual of Mental Disorders (4th ed., DSMIV; American Psychiatric Association, 1994); (b) a total score of at least 20 on the Yale Brown ObsessiveCompulsive Scale modified for BDD (BDDYBOCS; Phillips et al., 1997), including a score of at least 2 on Item 1 (13 hr per day of preoccupation with the perceived defect); and (c) facial concerns as the main preoccupation, because participants were rating their faces and facial concerns are the most common concerns in BDD. Most participants (88%) had multiple concerns, with facial concerns being the main for each BDD participant. Concerns with skin (70%), whole face (44%), eyes (38%), and nose (38%) were the most common. Aside from their primary BDD diagnosis, as confirmed by the Structured Clinical Interview for DSMIV (First, Spitzer, Gibbon, & Williams, 1997), their comorbid disorders were major depression (n 21), delusional disorder (n 19), social phobia (n 5), obsessive compulsive disorder (n 4), alcohol misuse (n 2), adjustment disorder (n 1), and bulimia nervosa (n 1). Participants with BDD were either individuals receiving treatment at the Priory Hospital North London or individuals who had contacted a BDD support group. The two nonclinical control groups excluded (a) those who responded yes to Have you ever been diagnosed with a psychiatric disorder? and/or Are you currently suffering with a psychiatric disorder? and (b) those who had excessive appearance concerns defined by a total score of 20 or above on the BDDYBOCS, including a score of at least 2 on Item 1. The only difference between the control groups was the additional inclusion criterion of an education or occupation in art and design, necessary for the art and design group. This included a current or completed education (at least advanced level) or training in art, fine art, art history, architecture, or design or an occupation as an artist, an art teacher, an architect, or a graphics, fashion, or textile designer. To classify a participant, both current and past occupation, training, and education in art and design were used. For instance, a participant with an art degree but working as a waiter qualified for the art and design group. The two control groups were recruited by (a) advertisements in a local newspaper, (b) email circulars in 11 universities and colleges, (c) leaflets delivered to 1,000 homes, and

LAMBROU, VEALE, AND WILSON

image to allow for the manipulations. Half of the facial or building image was then flipped horizontally (with the edit, transform, and flip horizontal commands). The first author used a randomized process to determine whether to flip the left side or right side. This involved tossing a coin. The randomized order matched across the three groups. To circumvent the technical problems reported in previous research, the first author took care to ensure that, when creating the symmetrical image, the image maintained the original widths of the face and building, as well as the widths and positioning of the internal features. By using the scale option in the edit tool, the transform tool restored any slight deviations from the original. Skin blemishes transferred to the other side of the face because of the flipping, were removed with the cloning stamp and healing brush tools. The final step was to flatten the layers and save the image. The first author saved the images in a bitmap format because it was the only format common to the software used in the manipulation and in the presentation of the images. Asymmetry. To create the facial and building images representing asymmetry, the first step was to duplicate a layer of the real image to allow for the manipulations. For the facial images, one side of the jaw line was pinched (with the filter, distort, and pinch tools) by 30% to reduce the symmetrical proportions of the face. To increase the asymmetry of the internal features, the same side of the mouth was pinched by 30% to reduce the size on one side, and the same side of the nose was pinched by 30% to increase the size on one side. For the building image, one side of the building was pinched by 10% to reduce the symmetrical proportions of the building. On the same side of the building, pinching the outer window on the first floor by 20% and outer window on the second floor by 20% increased the asymmetry of the internal features. The final step was to flatten the layers and save the images in a bitmap format. The asymmetry manipulations occurred on either the left side or the right side of the face and the building depending on the side originally flipped to create the symmetry image. If, to create symmetry, this entailed flipping the left side to the right side, then the asymmetry manipulations were on the right side and vice versa. Creating the images within the continuum. The first author created the six images within the continuum, using Ulead MorphStudio 1.0. In this software, an option is available to compose an image by morphing two images to varying percentages. The real image and symmetry were morphed to create three images of varying symmetry: 25% symmetry, 50% symmetry, and 75% symmetry. The real image and asymmetry were morphed to create three images of varying asymmetry: 25% asymmetry, 50% asymmetry, and 75% asymmetry.

related to depression. Total scores range from 0 to 63 (none or minimal: 10; mild to moderate: 10 18; moderate to severe: 19 29; severe: 30 63). Internal consistency was high in the present sample (Cronbachs .948). Values Scale. A simple scale specifically devised for the study measured aesthetic values. Given a relative scale, respondents are required to allocate 100 points to a choice of 10 values (physical appearance, family, friends, health, academic/ occupational success, religion, art, music, money, scientific truth) and to rate each one on degree of importance. Values that hold no importance to them receive a zero rating. The final points yield a total score of 100 for all respondents. Self-Objectification Questionnaire. The Self-Objectification Questionnaire (Noll & Fredrickson, 1998), a 10-item ranking scale was used to measure aesthetic identity. Total scores range from 25 to 25. A higher positive score indicates a greater emphasis on appearance and, thus, a higher level of self-objectification.

Procedure
Participants met with the first author on two occasions. In the first meeting, participants read the information form and had the opportunity to ask questions. After consenting to the study, participants had their photograph taken in preparation for the second meeting, where they viewed the real image and the manipulated versions of their image and the control images on a computer. Participants were then administered the BDD-YBOCS and the BDI. In the second meeting, before the computer study commenced, participants viewed the questions of the study. They received a thorough explanation of the procedure with the opportunity to ask questions for clarification. Participants were then presented with the nine images of their face (one real and eight manipulated) simultaneously on the computer screen. They were required to select and rate, in the following order, their actual self, their ideal self, their perfect face, the most physically attractive image and give their attractiveness rating for their selected image (scale: 0 extremely unattractive, 10 extremely attractive), the image that gave them the most pleasure and give their pleasure rating for their selected image (scale: 0 extreme disgust, 10 extreme pleasure), and the image that gave them the most disgust and give their disgust rating for their selected image (scale: 0 extreme pleasure, 10 extreme disgust). Viewing and rating time was 1 min for each selection question. A blank screen preceded each new selection question and series presentation. A neutral instructional format was used (i.e., please select which image represents your actual self) to eliminate a possible instructional effect. Participants subsequently completed the Values Scale and the Self-Objectification Questionnaire. Next, participants viewed and rated the images of the facial control condition, followed by the nonfacial control condition. The same procedure and question order used for the self-image condition applied, with two exceptions. First, to make their judgments for actual other and actual building, participants viewed the real image for 10 s. This image was followed by the nine images of the symmetry continuum for 1 min for the image selection. Second, in contrast to the facial image conditions, participants did not select their most attractive building or rate its attractiveness. To reduce

Questionnaires
BDD-YBOCS. The BDD-YBOCS (Phillips et al., 1997), a 12-item semistructured interview was administered to assess the severity of BDD symptoms during the past week. Total scores range from 0 to 48 (mild: 20 26; moderate: 2734; severe: 35 ). Internal consistency was high in the present sample (Cronbachs .951). Beck Depression Inventory (BDI). The BDI (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961), a 21-item self-report inventory, was used to assess the presence and severity of symptoms

ROLE OF AESTHETIC SENSITIVITY IN BDD

the risk of overtesting, male participants viewed only the male control face, and female participants viewed only the female control face. All participants were screened to ensure they were not familiar with the control face. There were two presentation orders for the position of the images on the screen, with the allocation of the order randomized by tossing a coin. Participants viewed the same presentation, in either the first or the second order, across the three conditions. The numbers viewing each presentation were equal across the three groups. At the conclusion of the study, each participant was debriefed.

BDI ( p .0001, r .85). Control groups did not differ. The groups were equivalent in age and gender ratio (see Table 1).

Aesthetic Perceptual Sensitivity

Table 2 presents the group means, standard deviations, test statistics, significance values, and effect sizes for the variables of aesthetic perceptual sensitivity. The findings were consistent with expectations. Perceptual understanding/awareness (idea of perfect). BDD participants and art and design controls (who did not differ) selected a significantly more symmetrical self-image for their idea of perfect than did non-art controls ( p .0001, r .62). Their perceptual superiority relative to non-art control participants extended to the control conditions (other face: p .0001, r .31; building: p .001, r .29). Perceptual accuracy. BDD participants were significantly more accurate in perceiving their actual self than were control participants ( p .0001, r .59). The groups did not differ in their accuracy for the control conditions. Effects of mirror checking. BDD participants checked their appearance in a mirror significantly more frequently on a weekly basis than did control participants, t(147) 5.992, p .0001, r .44 (BDD: M 3.9, SD 1.0; art and design: M 2.9, SD 1.0; non-art: M 2.8, SD 1.0). However, mirror checking did not influence the self-actual estimation results. When it was included as a covariate, the pattern of results did not change. Depressive realism. There was a significant negative correlation between self-actual estimation and depression (BDI total score; r .52, p .0001). Participants with mild-to-moderate depression were the most accurate. Participants with minimal or no depression showed the greatest positive distortion. Of the BDD group, 90% were at least mildly depressed. When we controlled for BDD severity (BDD-YBOCS total score), the significant association ceased (r .16, p .058). There was a significant negative correlation between self-actual estimation and BDD severity (r .54, p .0001). BDD participants with moderate BDD were the most accurate. BDD participants with severe BDD showed the greatest negative distortion. A significant positive corre-

Statistical Analysis
Multivariate analyses of variance (MANOVAs) were conducted to confirm the body image components. A series of one-way analyses of variance (ANOVAs) were conducted on each variable. When variances were considered heterogeneous (variance ratio 3:1), a BrownForsythe F* test was selected. Significant main effects were followed up with planned comparisons designed accordingly to test each specific prediction. A post hoc Tukey honestly significant difference test followed up an unexpected outcome. Cohens f (for ANOVAs) or Pearsons correlation coefficient r (for planned comparisons) were calculated to measure effect sizes (small: f 0.10, r .10; medium: f 0.25, r .30; large: f 0.40, r .50). Correlational analyses were performed to assess the depressive realism explanation. The following coding system was used in the analyses for the images selected from the symmetry continuum: symmetry 4; 75% symmetry 3; 50% symmetry 2; 25% symmetry 1; real image 0; 25% asymmetry 1; 50% asymmetry 2; 75% asymmetry 3; asymmetry 4. In light of the multiple comparisons, a Bonferroni corrected alpha level was set at .001 for all analyses.

Results Demographic and Clinical Characteristics

Relative to control participants, BDD participants scored significantly higher on the BDD-YBOCS ( p .0001, r .86) and the

Table 1 Demographic and Clinical Characteristics of BDD and Control Participants

BDD participants (n 50) Variable Demographic characteristic Age Gender Male Female Clinical characteristic BDD-YBOCS BDI M 27.7 18 32 29.6a 24.4a 6.3 10.6 8.2b 4.9b n SD 6.9 Art and design controls (n 50) M 26.2 16 34 5.7 4.7 8.4b 4.9b n SD 6.5 Non-art controls (n 50) M 26.3 18 32 6.1 4.5 n SD 5.1 Test statistic F(2, 147)
2

p .375 .930 .0001 .0001

Effect size Cohens f 0.11

0.988

(2)

0.235

F(2, 147) 209.218 F (2, 86.546) 121.431

1.67 1.27

Note. Means in the same row that do not share subscripts differ at p .001 in planned comparisons. BDD body dysmorphic disorder; BDD-YBOCS YaleBrown ObsessiveCompulsive Scale modified for body dysmorphic disorder; BDI Beck Depression Inventory. p .0001.

LAMBROU, VEALE, AND WILSON

Table 2 Group Differences in Aesthetic Perceptual Sensitivity

BDD participants (n 50) Variable and condition Perceptual understanding Self Other Building Perceptual accuracy Self Other Building M 3.4a 2.9a 3.3a 0.4a 0.3 0.4 SD 0.7 0.9 0.7 1.0 1.1 1.1 Art and design controls (n 50) M 3.2a 2.9a 3.2a 0.9b 0.3 0.6 SD 0.7 1.0 0.8 0.9 1.5 1.8 Non-art controls (n 50) M 1.9b 2.1b 2.7b 1.3b 0.4 0.6 SD 1.1 1.5 1.2 1.0 1.7 2.1 Test statistic: F(2, 147) 46.452 7.983 7.351 41.154 0.119 0.183a p .0001 .001 .001 .0001 .888 .833 body dysmorphic disorder; self Effect size Cohens f 0.79 0.33 0.31 0.74 0.04 0.05 self-image

Note. Means in the same row that do not share subscripts differ at p .001 in planned comparisons. BDD condition; other other face condition; building building condition. a F (2, 121.279). p .0001.

lation was observed between depression and BDD severity (r p .0001).

.82,

Aesthetic Emotional Sensitivity

Table 3 presents the group means, standard deviations, test statistics, significance values, and effect sizes for the variables of aesthetic emotional sensitivity. The outcomes were as predicted.

Pleasure. BDD participants and art and design controls (who did not differ) selected a significantly more symmetrical selfimage than did non-art controls to represent the image that gave them most pleasure ( p .0001, r .48). When rating their selected self-image, BDD participants experienced significantly less pleasure than did control participants ( p .0001, r .63). Control groups did not differ. The perceptual superiority in BDD participants and art and design controls extended to the control

Table 3 Group Differences in Aesthetic Emotional Sensitivity

BDD participants (n 50) Variable and condition Pleasure Perceptual selection Self Other Building Rating Self Other Building Disgust Perceptual selection Self Other Building Rating Self Other Building M SD Art and design controls (n 50) M SD Non-art controls (n 50) M SD Test statistic: F(2, 147) p Effect size Cohens f

2.9a 2.6a 2.7a 2.8a 5.6 6.1 2.9a 3.0a 3.1a 8.0a 5.0 5.8

0.9 1.1 1.4 1.8 1.4 1.8 1.0 1.0 0.9 1.7 1.7 1.9

2.6a 2.4a 2.8a 5.2b 5.8 6.5 2.8a 2.9a 2.8a 5.9b 4.6 5.3

1.0 1.1 1.2 1.3 1.2 1.6 0.9 1.1 1.1 1.5 1.3 1.7

1.5b 0.9b 1.9b 5.3b 6.0 6.1 1.9b 1.6b 1.6b 5.4b 4.4 4.9

1.2 1.9 1.4 1.3 1.3 1.5 1.2 1.7 1.6 1.5 1.5 1.7

23.146 20.330 6.951 47.308 0.942 1.252 12.562 16.748 21.111a 36.035 2.177 3.384

.0001 .0001 .001 .0001 .392 .289 .0001 .0001 .0001 .0001 .117 .037 body dysmorphic disorder; self

0.55 0.52 0.30 0.79 0.11 0.13 0.40 0.47 0.53 0.69 0.17 0.21 self-image

Note. Means in the same row that do not share subscripts differ at p .001 in planned comparisons. BDD condition; other other face condition; building building condition. a F (2, 113.275). p .0001.

ROLE OF AESTHETIC SENSITIVITY IN BDD

conditions (other face: p .0001, r .46; building: p .0001, r .29). The groups did not differ in their pleasure rating for the control conditions. Disgust. BDD participants and art and design controls (who did not differ) selected a significantly less symmetrical self-image than did non-art controls to represent the image that gave them most disgust ( p .0001, r .45). When rating their selected self-image, BDD participants experienced significantly more disgust than did control participants ( p .0001, r .57). Control groups did not differ. The perceptual superiority in BDD participants and art and design controls extended to the control conditions (other face: p .0001, r .43; building: p .0001, r .55). The groups did not differ in their disgust rating for the control conditions.

Aesthetic Evaluative Sensitivity

Table 4 presents the group means, standard deviations, test statistics, significance values, and effect sizes for the variables of aesthetic evaluative sensitivity. With the exception of personal and aesthetic standard, the results were as predicted. Personal standard. There were no significant differences between the groups in their self-ideal selections. Contrary to ex-

pectations, BDD participants were not demanding a higher selfideal standard than were control participants. However, there were significant group differences in the degree of self-discrepancy (see Figure 1). Consistent with our prediction, BDD participants expressed a greater discrepancy between their self-actual and selfideal than did control participants ( p .0001, r .62). Control groups did not differ. The groups did not diverge in their ideal standard or their actual/ideal discrepancy for the control conditions. Aesthetic standard. There was a significant group effect in the discrepancy between their self-ideal and perfect self. Contrary to expectations, planned comparisons revealed that BDD participants and control participants did not differ ( p .011, r .21), and control groups did diverge ( p .0001, r .45). A post hoc Tukey honestly significant difference test identified the source of the significant group effect. BDD participants and art and design controls (who did not differ) expressed a greater discrepancy between their self-ideal and perfect self than did non-art controls ( ps .0001). The groups did not differ in their aesthetic standard for the control conditions. Attractiveness standard. Relative to non-art controls, BDD participants and art and design controls (who did not differ)

Table 4 Group Differences in Aesthetic Evaluative Sensitivity

BDD participants (n 50) Variable and condition Aesthetic standards Personal standard Ideal Self Other Building Actual/ideal discrepancy Self Other Building Aesthetic standard (perfect vs. ideal) Self Other Building Attractiveness Standard Perceptual selection Self Other Rating Self Other Aesthetic values Values Scalephysical appearance Aesthetic identity Self-Objectification Questionnaire M SD Art and design controls (n 50) M SD Non-art controls (n 50) M SD Test statistic: F(2, 147) p Effect size Cohens f

2.2 2.5 2.7 2.6a 2.2 2.3 1.2a 0.4 0.6 2.8a 2.6a 3.3a 5.2 28.8a 8.7a

0.9 1.2 1.1 1.3 1.3 1.2 1.2 1.0 0.9 0.8 1.0 2.1 1.6 16.7 11.2

1.9 2.4 2.5 1.0b 2.1 1.9 1.3a 0.5 0.7 2.5a 2.4a 5.1b 5.6 8.6b 2.5b

1.2 1.2 1.5 1.0 1.8 1.9 1.2 1.1 1.1 1.3 1.2 1.2 1.3 4.6 9.9

1.9 2.2 2.3 0.6b 1.8 1.7 0.0b 0.1 0.3 1.6b 1.3b 4.8b 5.8 9.1b 2.1b

1.2 1.4 1.5 1.0 1.7 2.2 0.9 1.4 1.1 1.2 1.9 1.5 1.5 5.6 11.3

1.701 1.036 1.076 48.028 1.170 1.426a 22.104 2.959 1.978 15.852c 12.316d 16.636 1.769 60.073e 17.032

.186 .357 .343 .0001 .313 .243 .0001b .055 .142 .0001 .0001 .0001 .174 .0001 .0001

0.15 0.12 0.12 0.80 0.12 0.14 0.48 0.20 0.16 0.46 0.41 0.47 0.15 0.89 0.48

Note. Means in the same row that do not share subscripts differ at p .001 in planned comparisons. BDD body dysmorphic disorder; self self-image condition; other other face condition; building building condition. F (2, 126.940). b A post hoc Tukey honestly significant difference test identified the source of the significant group effect. c F (2, 133.473). d F (2, 108.377). e F (2, 67.935). p .0001.

8
-4 -3

LAMBROU, VEALE, AND WILSON

-2

-1

|_____|_____|_____|_____|_____|_____|______|_____|
100% AS 75% AS 50% AS 25% AS RI 25% S 50% S 75% S 100% S

BDD participants Art and design controls Non-art controls

Figure 1. Group differences in self-actual versus self-ideal discrepancy. AS S symmetry; BDD body dysmorphic disorder. asymmetry; RI real image;

selected a significantly more symmetrical self-image for their most physically attractive image ( p .0001, r .41). BDD participants rated their selected self-image significantly lower in attractiveness than did control participants ( p .0001, r .53). Control groups did not differ. The perceptual superiority in BDD participants and art and design controls extended to the other face condition ( p .0001, r .46). The groups did not differ in their attractiveness ratings for the control other face. Aesthetic values. BDD participants were significantly more likely to value the importance of physical appearance than were control participants ( p .0001, r .75). Control groups did not differ. Aesthetic identity. BDD participants reported a significantly higher positive self-objectification score than did control participants ( p .0001, r .43). Control groups did not differ.

not influence the results. There were no significant differences between the two BDD subgroups for any variable in this study ( ps .05).

Discussion
It was hypothesized that a higher aesthetic sensitivity contributes to the development and maintenance of BDD. This study examined the central premise that rather than being distorted in their perceptual processing, those with BDD have an enhanced understanding of aesthetic proportions and an increased accuracy in their self-actual estimation. We expected the source of their disturbance to be in their emotional/evaluative processing, which is specific to the self. This would manifest in art and design controls resembling BDD participants in their superior perceptual understanding of aesthetic proportions and non-art controls in their emotional/ evaluative processing. Overall, the findings corroborate these hypotheses. Effect sizes for these key findings were medium to large. Individuals with BDD and art and design controls seemed to have a clearer idea of the criteria of attractiveness levels defined by symmetry. They displayed a greater awareness of their aesthetic facial proportions relative to non-art controls, and this extended to another persons face and a building. This reinforces the idea that individuals with BDD possess a more critical eye and appreciation of aesthetics, which they then apply to their own appearance (Veale & Lambrou, 2002). BDD participants were also more accurate in perceiving their actual self compared with controls. According to Thomas (1990), the body image of each healthy member of the population tends to mirror their position within the attractiveness distribution. Dissatisfaction arises when there is a mismatch between perceived appearance and actual appearance. However, the results of this study suggest that individuals with BDD, who are particularly dissatisfied, present a closer match between perceived appearance and actual appearance. On average, they displayed only a slight negative bias in their estimation of themselves. Hence, they did not display a distortion in their perceptual processing. Rather this study substantiates preliminary evidence (Thomas & Goldberg, 1995) that individuals with BDD are superior in their discriminatory abilities. In contrast, control participants expressed a pro-

Confirmation of Body Image Components: Perceptual and Emotional/Evaluative

We postulated that the following variables assessed the perceptual component of body image: (a) perfect, (b) pleasure perceptual selection, (c) disgust perceptual selection, and (d) physical attractiveness perceptual selection. MANOVAs conducted with these variables revealed, as expected, a significant group effect for the self-image condition, F(8, 290) 11.896, p .0001, for the other face condition, F(8, 290) 6.788, p .0001, and for the building condition, F(6, 292) 7.281, p .0001. In addition, we postulated that the following variables would assess the emotional/evaluative component of body image: (a) actual, (b) pleasure rating, (c) disgust rating, (d) ideal, (e) actual/ ideal discrepancy, (f) physical attractiveness rating, (g) aesthetic values, and (h) aesthetic identity. MANOVAs conducted with these variables revealed, as expected, a significant group effect for the self-image condition, F(14, 284) 11.650, p .0001, but not for the other face condition, F(10, 288) 0.922, p .513, or for the building condition, F(8, 290) 1.700, p .089.

Influence of the BDD Concerns

Of the BDD group, 52% were preoccupied with at least one of the facial features altered in the manipulations. However, this did

ROLE OF AESTHETIC SENSITIVITY IN BDD

nounced positive bias toward symmetry, overestimating the attractiveness of their actual self. In other words, control participants were looking at themselves through rose-tinted spectacles. Control participants overestimation of attractiveness was specific to their own face and is consistent with previous findings. Jansen, Smeets, Martijn, and Nederkoorn (2006) observed that, contrary to eating-symptomatic participants who expressed a realistic view of their attractiveness, control participants overestimated their own independently rated attractiveness. Similarly, Noles, Cash, and Winstead (1985) reported that, although depressed students underestimated their attractiveness (akin to the slight negative distortion in the BDD group), they were nevertheless more accurate in their self-appraisals than nondepressed students, who positively exaggerated their level of attractiveness. Consistent with the depressive realism explanation, those with mild-to-moderate depression in the present study were the most accurate in their self-perception. Participants with minimal or no depression showed the greatest positive distortion. The findings suggest that overestimating personal attractiveness may be important for psychological well-being, providing protection from developing body image conditions and depression. Selfperceived facial appearance correlates positively with global selfworth (Pope & Ward, 1997). The present results imply that individuals with BDD are not using a self-serving bias when evaluating their physical appearance. Thus, their minor imperfections disappoint them, and they experience negative emotions, such as depression. Future studies could examine whether individuals with BDD also fail to use self-serving attributional biases to assess self-relevant information unrelated to physical appearance. There was no perceptual accuracy in the severely depressed, all of whom were BDD participants. The situation is probably made more complex by the fact that clinicians have described individuals with BDD who clearly display distortion in their self-portraits. Severity of the BDD condition may play an important role. Perhaps at some point on the BDD severity scale, individuals with BDD become exclusively preoccupied with their internal body image. Presumably, at this stage they have a grossly distorted internal body image because of the limited validation with objective reality. The present study provides preliminary support for this idea, because those with severe BDD showed the greatest negative distortion in their self-actual estimation. Participants with moderate BDD were the most accurate. Although BDD severity ranged from mild to severe in the present BDD sample, most BDD participants had moderate BDD, which may account for their lack of distortion. It is important to disentangle the effects of the BDD symptomatology and of depression on participants self-actual estimations. Replication with a sample of depressed and nondepressed BDD participants will determine whether depression influences perceptual accuracy over and above the effects accounted for by BDD. Future research could also compare BDD participants with depressed participants. Aesthetic values and identity could be the key factors separating them. Only those with BDD will overvalue the importance of appearance and self-objectify, which may promote BDD symptomatology. Although BDD participants checked their appearance in the mirror more often than controls, this did not account for their perceptual accuracy. The quality of the mirror-checking experience may be more important than the quantity of checks (e.g., Mulkens & Jansen, 2009). To enhance ecological validity, an eye

tracker could register participants eye movements when they are looking at themselves in the mirror to examine how they scan their face. Individuals with BDD display visual processing and frontostriatal abnormalities when viewing their own face, which imply a bias for detail encoding and analysis rather than holistic processing (Feusner et al., 2010). It is therefore likely that individuals with BDD approach the mirror with a more analytical eye, deliberately assessing features and focusing on their perceived defects. This approach heightens their self-perception and reinforces their dissatisfaction. The increased perceptual accuracy in the BDD group was specific to their own faces. The groups did not differ in the control conditions. Thus, BDD participants did not display impairments in their face- or object-recognition abilities, further substantiating the idea that there is no perceptual deficit. Indeed, Stangier, AdamSchwebe, Muller, and Wolter (2008) found that individuals with BDD were more accurate than control participants in recognizing aesthetic alterations to other faces. The accuracy may not extend to other body image conditions. Individuals with eating disorders were less accurate than controls at detecting the facial flaws of others (Legenbauer, Kleinstauber, Muller, & Stangier, 2008). The outcome might have been different had these authors assessed the body, because weight and shape are the concerns in eating disorders. The role of facial versus body concerns on aesthetic sensitivity needs to be investigated. The emotional sensitivity results confirm the main hypothesis that individuals with BDD are superior in their perceptual processing and disturbed in their emotional processing. BDD participants and art and design controls were perceptually superior to non-art controls. They selected a more symmetrical face to represent the self-image providing the most pleasure and a less symmetrical face to represent the self-image providing the most disgust. Crucially, when rating their selected self-image for most pleasure and most disgust, BDD participants experienced less pleasure and more disgust, respectively, than either control group. Surprisingly, although the insula, one of the main regions implicated in self-recognition, plays a role in the experience of emotions such as disgust, insula hyperactivity is not evident in those with BDD (Feusner et al., 2010). The perceptual superiority observed in BDD participants and art and design controls extended to another persons face and a building. The emotional ratings were the pivotal difference between the self-image and the control conditions. In contrast to their self-image, where they exhibited a marked negative emotional bias, BDD participants resembled controls in their pleasure and disgust ratings for the other face and the building. This corroborates the premise that the emotional disturbance in those with BDD is specific to their own face. The aesthetic evaluations component of aesthetic sensitivity is subdivided into standards, values, and identity. Contrary to the commonly held view and our prediction, individuals with BDD did not demand a higher standard of beauty for themselves compared with individuals from the healthy population. BDD participants and controls did not differ, with each selecting a modest self-ideal. Of interest, Silver and Reavey (2010) reported that for some with BDD, their ideal self was an idealized version of their perceived childhood self. The groups were also analogous in their modest ideal standard for another persons face and for a building. In line with Veale et al. (2003), the key difference between the groups was the self-discrepancy. As a result of overestimating their actual attractiveness, controls expressed only a marginal difference

10

LAMBROU, VEALE, AND WILSON

between their actual self and their ideal self, suggesting that they were satisfied with their perceived appearance. Korabik and Pitt (1980) observed that individuals with a high self-concept tended to see themselves as closer to the ideal than they actually were. Relative to controls, BDD participants expressed a greater discrepancy between their perceived actual self and their desired ideal self, because of an absent self-serving bias in their self-actual estimation. Thus, the crucial determining factor for those with BDD, and probably the source of their disturbance, is that a discrepancy exists between their actual and ideal, rather than that they possess an unrealistic ideal or a distorted actual self. According to Higgins (1987), individuals displaying a discrepancy between their actual self and their ideal self are vulnerable to dejection-related emotions, such as depression. This may explain the increased depressive symptomatology in the present BDD sample. As they did not differ from controls in their self-ideal standards, treatment strategies that aim to develop self-serving biases, such as those evident in controls, may prove effective in reducing the self-discrepancy in those with BDD and, consequently, may alleviate their symptomatology. BDD participants resembled control participants in their discrepancy between actual and ideal for the control conditions. Their significant actual versus ideal discrepancy was therefore unique to their self-image. This further attests to the idea that for those with BDD, the source of their disturbance is the specific discrepancy between their perceived actual self and their desired ideal self. Contrary to expectations, BDD participants expressed a discrepancy between their perfect and ideal self. They selected a more symmetrical self-image for their perfect self than for their ideal self, verifying the earlier observation that most do not desire perfection in their appearance. Indeed, they chose images analogous to those chosen by art and design controls. On a perceptual level, BDD participants and art and design controls displayed a higher attractiveness standard, probably the result of their increased understanding of aesthetic proportions. They were superior to non-art controls in selecting a more symmetrical self-image to represent their most physically attractive image. This perceptual superiority extended to the control other face. Akin to the trend observed for emotional sensitivity, what differentiated the BDD group was their self-rating. They rated the attractiveness of only their chosen self-image markedly lower than both control groups. There were no group differences when rating the attractiveness of the control other face. This did not concur in Buhlmann et al.s (2008) study, in which they found that participants with BDD rated attractive faces as more attractive compared with participants with obsessive compulsive disorder and control participants. However, participants in the present study rated the image they perceived as the most physically attractive as opposed to rating images classified by others as attractive. This study did not include the objective rating of the participants actual attractiveness. Should research unequivocally show individuals with BDD to be less attractive than the general population and than others with a psychiatric disorder, this would raise some questions regarding the validity of the BDD diagnosis. Thomas and Goldberg (1995) provided modest support for the possibility of those with BDD being less attractive, but they did not assess symmetry or include a control psychiatric group. Buhlmann et al. (2008), however, reported that BDD, obsessive compulsive disorder, and control groups did not differ in their independently rated

attractiveness. Perhaps for BDD participants in the present study, their appearance dissatisfaction and negative self-appraisals were, to some extent, justified, as a panel in Jansen et al. (2006) found in an eating-symptomatic group. Alternatively, perhaps BDD participants and controls did not differ in their actual attractiveness; the discrepancy in their ratings may have been merely a reflection of control participants overestimating their attractiveness to maintain their positive self-concept. These issues need to be addressed. Physical appearance was the most important value only for BDD participants. They valued its importance three times more than control participants did, implying that they may be valuing appearance to a dysfunctional degree. Consistent with the physical attractiveness stereotype, Buhlmann et al. (2009) reported that BDD participants attached more meaning and consequences to thoughts about the importance of appearance (e.g., attractive people are more competent). Valuing the importance of appearance is likely to reinforce processing of the self as an aesthetic object (Veale, 2004). BDD participants were more likely to self-objectify than controls. It is reasonable to hypothesize that these aesthetic evaluations predispose an individual to BDD, as well as maintaining the condition. One must interpret the present outcomes in light of several limitations. The main limitation was the unavoidable selection bias in the BDD group. Only those who were willing to view and rate their image agreed to participate. BDD participants were also preselected for their predominant facial concerns and for falling within a restricted age range of 18 to 40 years. Further studies would verify whether the results generalize to those who refused to participate, those primarily concerned with nonfacial features, or those ages above 40 or below 18. Given that BDD usually begins in adolescence, prospective longitudinal studies based on young vulnerable samples may elucidate whether a high aesthetic sensitivity predisposes an individual to BDD and/or is a consequence of the disorder. Comorbidity was present in the BDD group. Although this is typical in BDD, the question does arise as to whether the comorbid disorders partially influenced the outcomes. Comparing a BDD group without comorbidities with related clinical control groups, such as groups with eating disorders, depression, obsessive compulsive disorder, and social phobia, would address issues regarding a possible degree of overlap in the pathogenesis of BDD and these comorbid disorders. Future investigations could use electroculography to measure how BDD and comparative groups differentiate and/or resemble one another in their strategies when viewing, selecting, and rating the images. It would have been preferable for participants to view both the male and female control images to allow comparisons of cross-gender differences, but there was a risk of overtesting. In addition, the order in which the three conditions were presented should have been randomized. The present study explored the role of aesthetic sensitivity in BDD. Results support the distinction between perceptual and emotional/ evaluative modalities of body image. Results also substantiate the underlying premise of the study that, rather than suffering from a perceptual deficit, individuals with BDD possess an increased understanding of aesthetic proportions, which extends to other faces and a building, and a superior accuracy in their self-actual estimation. The source of the disturbance is in their emotional/evaluative processing when viewing their self-image. Such findings hold clear implications for treatment. Rather than engaging in a debate with them about the existence of their perceived defects, it may be more effective for clinicians to focus on their emotional/evaluative processing.

ROLE OF AESTHETIC SENSITIVITY IN BDD

11

A higher aesthetic sensitivity may contribute to an explanation of why a small defect in their appearance severely disturbs those with BDD. As predicted, participants with BDD responded the same way as art and design controls in the control conditions. It is therefore important to understand what differentiates these groups. This will not only aid the development of effective treatment strategies but also contribute to the prediction of which individuals are vulnerable to BDD. Differences in personality and psychological factors, such as the experience of shame, may be pivotal to appreciating the idiosyncratic pathways of BDD. The promising outcomes of this study highlight that aesthetic sensitivity plays a role in BDD and indicate potential therapeutic strategies. With the development of research methods and the questions generated from this study, future research on self-perception in BDD may unlock the mysteries of this enigmatic disorder.

References
Alloy, L. B., & Abramson, L. Y. (1979). Judgment of contingency in depressed and nondepressed students: Sadder but wiser? Journal of Experimental Psychology: General, 108, 441 485. doi:10.1037/00963445.108.4.441 American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression. Archives of General Psychiatry, 4, 561571. Buhlmann, U., Etcoff, N. L., & Wilhelm, S. (2008). Facial attractiveness ratings and perfectionism in body dysmorphic disorder and obsessive compulsive disorder. Journal of Anxiety Disorders, 22, 540 547. doi: 10.1016/j.janxdis.2007.05.004 Buhlmann, U., Teachman, B. A., Naumann, E., Fehlinger, T., & Rief, W. (2009). The meaning of beauty: Implicit and explicit self-esteem and attractiveness beliefs in body dysmorphic disorder. Journal of Anxiety Disorders, 23, 694 702. doi:10.1016/j.janxdis.2009.02.008 Feusner, J. D., Moody, T., Hembacher, E., Townsend, J., McKinley, M., Moller, H., & Bookheimer, S. (2010). Abnormalities of visual processing and frontostriatal systems in body dysmorphic disorder. Archives of General Psychiatry, 67, 197205. doi:10.1001/archgenpsychiatry.2009.190 First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1997). Structured Clinical Interview for DSMIV Axis I Disorders: Clinician Version. Washington, DC: American Psychiatric Association. Fredrickson, B. L., & Roberts, T. (1997). Objectification theory: Toward understanding womens lived experiences and mental health risks. Psychology of Women Quarterly, 21, 173206. doi:10.1111/j.14716402.1997.tb00108.x Gunstad, J., & Phillips, K. A. (2003). Axis I comorbidity in body dysmorphic disorder. Comprehensive Psychiatry, 44, 270 276. doi:10.1016/ S0010-440X(03)00088-9 Harris, D. L. (1982). Cosmetic surgerywhere does it begin? British Journal of Plastic Surgery, 35, 281286. Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect. Psychological Review, 94, 319 340. doi:10.1037/0033-295X.94.3.319 Jansen, A., Smeets, T., Martijn, C., & Nederkoorn, C. (2006). I see what you see: The lack of a self-serving body-image bias in eating disorders. British Journal of Clinical Psychology, 45, 123135. doi:10.1348/ 014466505X50167 Jerome, L. (1992). Body dysmorphic disorder: A controlled study of patients requesting cosmetic rhinoplasty. American Journal of Psychiatry, 149, 577578. Korabik, K., & Pitt, E. J. (1980). Self concept, objective appearance, and

profile self perception. Journal of Applied Social Psychology, 10, 482 489. doi:10.1111/j.1559-1816.1980.tb00727.x Legenbauer, T., Kleinstauber, M., Muller, T., & Stangier, U. (2008). Are individuals with an eating disorder less sensitive to aesthetic flaws than healthy controls? Journal of Psychosomatic Research, 65, 8795. doi: 10.1016/j.jpsychores.2008.02.014 Mulkens, S., & Jansen, A. (2009). Mirror gazing increases attractiveness in satisfied, but not in dissatisfied women: A model for body dysmorphic disorder? Journal of Behavior Therapy and Experimental Psychiatry, 40, 211218. doi:10.1016/j.jbtep.2008.10.001 Noles, S. W., Cash, T. F., & Winstead, B. A. (1985). Body image, physical attractiveness, and depression. Journal of Consulting and Clinical Psychology, 53, 88 94. doi:10.1037/0022-006X.53.1.88 Noll, S. M., & Fredrickson, B. L. (1998). A mediational model linking self-objectification, body shame, and disordered eating. Psychology of Women Quarterly, 22, 623 636. doi:10.1111/j.1471-6402.1998.tb00181.x 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. doi:10.1080/09658210444000043 Phillips, K. A., Hollander, E., Rasmussen, S. A., Aronowitz, B. R., DeCaria, C., & Goodman, W. K. (1997). A severity rating scale for body dysmorphic disorder: Development, reliability, and validity of a modified version of the YaleBrown ObsessiveCompulsive Scale. Psychopharmacology Bulletin, 33, 1722. Pope, A. W., & Ward, J. (1997). Self-perceived facial appearance and psychosocial adjustment in preadolescents with craniofacial anomalies. Cleft Palate Craniofacial Journal, 34, 396 401. doi:10.1597/15451569(1997)0342.3.CO;2 Rhodes, G., Proffitt, F., Grady, J. M., & Sumich, A. (1998). Facial symmetry and the perception of beauty. Psychonomic Bulletin & Review, 5, 659 669. Silver, J., & Reavey, P. (2010). Hes a good-looking chap aint he?: Narrative and visualisations of self in body dysmorphic disorder. Social Science & Medicine, 70, 16411647. doi:10.1016/j.socscimed.2009.11.042 Stangier, U., Adam-Schwebe, S., Muller, T., & Wolter, M. (2008). Dis crimination of facial appearance stimuli in body dysmorphic disorder. Journal of Abnormal Psychology, 117, 435 443. doi:10.1037/0021843X.117.2.435 Thomas, C. S. (1990). Stress and facial appearance. Stress Medicine, 6, 299 304. doi:10.1002/smi.2460060408 Thomas, C. S., & Goldberg, D. P. (1995). Appearance, body image and distress in facial dysmorphophobia. Acta Psychiatrica Scandinavica, 92, 231236. doi:10.1111/j.1600-0447.1995.tb09574.x Veale, D. (2004). Advances in a cognitive behavioural model of body dysmorphic disorder. Body Image, 1, 113125. doi:10.1016/S17401445(03)00009-3 Veale, D., Ennis, M., & Lambrou, C. (2002). Possible association of body dysmorphic disorder with an occupation or education in art and design. American Journal of Psychiatry, 159, 1788 1790. doi:10.1176/ appi.ajp.159.10.1788 Veale, D., Gournay, K., Dryden, W., Boocock, A., Shah, F., Willson, R., & Walburn, J. (1996). Body dysmorphic disorder: A cognitive behavioural model and pilot randomised controlled trial. Behaviour Research and Therapy, 34, 717729. doi:10.1016/0005-7967(96)00025-3 Veale, D., Kinderman, P., Riley, S., & Lambrou, C. (2003). Selfdiscrepancy in body dysmorphic disorder. British Journal of Clinical Psychology, 42, 157169. doi:10.1348/014466503321903571 Veale, D. M., & Lambrou, C. (2002). The importance of aesthetics in body dysmorphic disorder. CNS Spectrums, 7, 429 431.

Received November 8, 2009 Revision received August 3, 2010 Accepted August 4, 2010