خودارائه گری کمال، تصور ذهنی از بدن و علائم اختلال تغذیه

Abstract A specific model for eating disorder symptoms involving perfectionistic self-presentation and two different moderators (i.e., body image evaluation and body image investment) was tested. Participants completed measures of perfectionistic self-presentation, body image dysfunction, and eating disorder symptoms. Findings indicated that all three dimensions of perfectionistic self-presentation were associated with eating disorder symptoms. Results also showed that perfectionistic self-presentation predicted eating disorder symptoms in women who were dissatisfied with their bodies, but that it did not predict eating problems in women who liked their bodies and felt there was little or no discrepancy between their actual and ideal appearances. Body image investment did not moderate the relationship between perfectionistic self-presentation and eating disorder symptoms, suggesting that ego-involvement alone is insufficient to promote eating disturbance in the context of perfectionism. The importance of self-presentation components of perfectionism and specific body image difficulties in predicting eating disorder symptoms are discussed.

Introduction Perfectionism has long been recognized as a central feature of eating disorders, and has been hypothesized to have an etiological role in eating problems. Early work by Bruch (1978) described eating disorder patients as perfectionistic, overly submissive, and constantly fearful of not being respected or valued. Others have noted that eating disordered patients may emerge from familial environments that emphasize a public image of perfection (Humphrey, 1992), where mistakes are greeted with dismay and the patient has learned to derive self-worth from the rigid pursuit of physical perfection (Reindl, 2001). Moreover, these clinical observations fit with findings that have tied perfectionism to eating disturbances in cross-sectional samples (e.g., Davis, Claridge, & Fox, 2000; Hewitt, Flett, & Ediger, 1995; Pryor, Wiederman, & McGilley, 1996), and with longitudinal evidence that perfectionism is a potent risk factor for eating disorder development (e.g., Lilenfeld et al., 2000; Tyrka, Waldron, Graber, & Brooks-Gunn, 2002; Vohs, Bardone, Joiner, & Abramson, 1999). While there is a clearly supported link between perfectionism and eating disorders, it is meaningful to ask what type of perfectionism has been examined. As a construct, perfectionism has been variously described, ranging from unidimensional cognitive perspectives (Burns, 1983) to multidimensional trait models (Frost, Marten, Lahart, & Rosenblate, 1990; Hamachek, 1978 and Hewitt and Flett, 1991). From a multidimensional perspective, two broad components have become prevalent in the field of perfectionism: the trait component (Hewitt & Flett) and, more recently, the self-presentation component (Hewitt et al., 2003). Trait perfectionism entails a need to be perfect and it speaks to the source of the perfectionistic demands (i.e., self or others). Although perfectionism may be manifest within the individual, it can also be expressed interpersonally. This suggests an important distinction between an individual's need to be perfect and his or her need to appear perfect in the eyes of others. To account for these entrenched interpersonal styles, a perfectionistic self-presentation component was developed and added to the multidimensional model ( Hewitt et al., 2003). Perfectionistic self-presentation, or the need to appear to be perfect, centers on how perfectionists behave in expressing their supposed perfection to others. Although a desire to actually be perfect (as in trait perfectionism) may involve a desire to appear to be perfect (as in perfectionistic self-presentation), this is not invariably true. Nor is it inevitable that a desire to appear to be perfect necessarily entails a corresponding need to actually be perfect. Analyses involving both clinical and nonclinical samples have demonstrated that trait perfectionism and perfectionistic self-presentation are distinct and separable components of personality (e.g., Hewitt et al., 1995 and Hewitt et al., 2003), and are predictive of different maladaptive outcomes ( Hewitt et al., 2003). For the purposes of this study, we elected to focus on perfectionistic self-presentation. Three perfectionistic self-presentation facets have been described (Hewitt et al., 2003): perfectionistic self-promotion (PSP), nondisclosure of imperfection (NDC), and nondisplay of imperfection (NDP). Perfectionistic self-promotion involves actively proclaiming one's successes, strengths, and achievements to others. Conversely, the latter two facets are protective or defensive orientations geared toward concealing imperfections. Each style has as its goal the maintenance of a flawless image by obscuring perceived mistakes or weaknesses, but each achieves that end in a different way. The nondisclosure of imperfection facet entails a reluctance to verbally admit personal shortcomings, whereas the nondisplay of imperfection facet involves an avoidance of behavioral displays of imperfection. Although there is a substantial body of evidence for an association between trait perfectionism and eating disorder symptoms (e.g., Bastiani, Rao, Weltzin, & Kaye, 1995; Hewitt et al., 1995; McLaren, Gauvin, & White, 2001; Pliner & Haddock, 1996), investigators have only recently assessed the role of perfectionistic self-presentation in eating disorders. For example, Cockell et al. (2002) demonstrated that anorexic patients had higher scores on nondisclosure of imperfection than did other psychiatric patients and normal controls, suggesting that anorexic patients are concerned with presenting themselves as perfect by not admitting their imperfections. Additional work revealed that all three perfectionistic self-presentation facets were associated with anorexics’ tendency to suppress negative feelings and to give priority to others’ feelings (Geller, Cockell, Hewitt, Goldner, & Flett, 2000). More recently, investigators found that perfectionistic self-presentation predicted dietary restraint, and that this relationship was mediated by an individual's psychological commitment to exercise (McLaren et al.). Finally, research involving female university students demonstrated that all three self-presentational facets of perfectionism were related to eating disorder symptoms, increased body image avoidance, and decreased appearance self-esteem (Hewitt et al.). Thus, presenting a public image of perfection is associated with eating disorder symptoms and other weight and shape concerns. Current theory conceptualizes the link between perfectionism and eating pathology in a diathesis–stress framework (e.g., Heatherton & Baumeister, 1991; Hewitt & Flett, 2002; Joiner, Heatherton, Rudd, & Schmidt, 1997). In this model, perfectionism acts as a vulnerability factor that promotes psychopathology in the presence of stressful failures. A key aspect of this theory is that ego-involvement alone will not lead to psychological symptoms in the context of perfectionism. Only a failure in an ego-involving domain is predicted to induce symptoms in perfectionists. Hewitt and Flett (2002) have added a level of complexity to this relationship by suggesting that perfectionism is not an inert attribute that merely reacts to stress, but is instead a dynamic trait that also interacts with and creates stress. For example, perfectionism can affect the impact of distressing events ( Hewitt & Flett, 1993). Individuals with stringent evaluative criteria are more likely to be faced with a failure to meet a goal, and such failures, even when minor, will be viewed as serious downfalls. In this way, perfectionism serves to enhance stress. Thus, for a woman who holds rigid appearance standards and fails to achieve her weight goal, the event will be experienced as a calamity rather than as a temporary setback. With respect to eating pathology, a diathesis–stress model could help explain evidence that perfectionism persists following long-term recovery from eating disorders (Bastiani et al., 1995, Kaye et al., 1998, Srinivasagam et al., 1995 and Stein et al., 2002). That is, perfectionism may be a persistent vulnerability factor that is malignant only under certain conditions. Moreover, the particular symptoms that arise from the interaction of perfectionism with stress may depend upon the nature of the stressor. For example, while body dissatisfaction may interact with perfectionism to produce eating disorder symptoms, achievement stress in the context of perfectionism may result in depression. Therefore, the addition of variables that may moderate the effects of perfectionistic self-presentation could help clarify the perfectionism–eating disturbance relationship. We propose body image as one such moderator. Consistent with this, Joiner et al. (1997) demonstrated that perfectionism (as measured by the Eating Disorders Inventory Perfectionism subscale) acts as a vulnerability factor for bulimic symptoms only in those individuals who perceive themselves to be overweight. In their conceptualization, perfectionism is a risk factor for eating problems only when an individual fails to meet weight standards. Moreover, actual weight status did not affect this relationship. To refine and extend this work, Vohs et al. (1999) showed that women high in perfectionism who perceive themselves to be overweight experience an increase in bulimic symptoms over time only if they also have low self-esteem. More recently, investigators used a longitudinal design with different measurement techniques to provide further support for the model's predictive ability (Vohs et al., 2001). Thus, past work has demonstrated that a diathesis–stress model explains variance in eating disorder symptoms. However, the aforementioned studies used the Eating Disorders Inventory Perfectionism subscale (EDI-P; Garner, Olmstead, & Polivy, 1983). Although generally treated as a unidimensional subscale, the EDI-P is actually a composite of two dimensions of trait perfectionism, self-oriented perfectionism and socially prescribed perfectionism (Bardone, Vohs, Abramson, Heatherton, & Joiner, 2000; Joiner et al., 1997 and Joiner and Schmidt, 1995; Sherry, Hewitt, Besser, McGee, & Flett, 2004). The use of the EDI-P may obscure the differential relations of its self-oriented and socially prescribed components to eating pathology and it ignores the possible role of perfectionistic self-presentation in a diathesis–stress model of eating disturbance. Further, while we know that a perceived failure in an appearance domain interacts with perfectionism to predict eating problems, no one has yet tested whether it is only a failure that yields these results. Perhaps simply being preoccupied by appearance, or putting great importance on appearance is enough to provoke eating disorder symptoms in perfectionists. For example, Ruggiero, Levi, Ciuna, and Sassaroli (2003) examined the relationship between perfectionism and eating disorder symptoms in high school students on an average school day, an examination day, and the day on which the students were to receive the evaluation of their performance on the exam. The results indicated that perfectionism was associated with drive for thinness only on the day that the grades were to be returned, suggesting that simply anticipating academic failure may have been enough to prompt eating disturbance amongst perfectionists. It raises the question of whether the same could be true if the moderator was body image investment rather than academic investment. Therefore, in the interest of building on past work, we attempted to examine two key features. First, we elected to examine a moderational model using perfectionistic self-presentation to explore whether the need to appear to be perfect interacts with body image dysfunction to predict eating disorder symptoms. Second, we chose to measure body image using multiple measures and to examine two aspects of body image as potential moderators: body image evaluation, or the degree to which a person likes the way she looks and how close she feels she is to her ideal appearance, and body image importance or investment, the degree to which a person spends time on her appearance and how important her ideal appearance is to her (Brown, Cash, & Mikulka, 1990; Cash, 2000; Cash & Szymanski, 1995). We wanted to establish whether it is specifically a discrepancy between an individual's actual and ideal appearance that moderates the perfectionism–eating pathology relationship, or whether believing that appearance is important will also affect the relationship between perfectionism and eating disorder symptoms. The former constitutes a perceived failure experience, while the latter does not. To summarize, the purpose of this study was to explore the relationship between self-presentational facets of perfectionism and eating disorder symptoms and to determine whether body image evaluation or body image investment influences that relationship. Consistent with research on multidimensional perfectionism (e.g., Cockell et al., 2002 and Hewitt et al., 1995), we predicted that all three of the perfectionistic self-presentation facets (PSP, NDP, and NDC) would be positively related to eating disorder symptoms. Second, consistent with theory (e.g., Heatherton & Baumeister, 1991; Hewitt & Flett, 2002; Joiner et al., 1997) we predicted that the perfectionism × body image evaluation interactions would predict significant variance in eating disorder symptoms. Specifically, we anticipated that the severity of eating disturbance amongst perfectionistic self-presenters would be worse when negative body image evaluation was high. Third, we predicted the perfectionism × body image investment interactions would not predict significant variance in eating disorder symptoms. That is, level of body image investment would not affect the relationship between perfectionism and eating disturbance. This was based on the belief that, unlike body image evaluation, body image investment does not in and of itself constitute a perceived failure to meet expectations.

Results Although we had numerous subscales available to serve as measures of our proposed moderators, body image evaluation and body image investment, we were concerned that the large number of analyses would seriously reduce power and, thus, hamper our ability to detect effects. Therefore, we created composite measures of body image dysfunction to serve as moderators. Using the five subscale scores from the MBSRQ and the BIQ, we conducted a principal components analysis with varimax rotation. Based on a scree test,2 a two-factor solution best fit the data and accounted for 76.1% of the variance. Selecting variables with loadings above .40, the first factor consisted of the MBSRQ Appearance Evaluation subscale, the MBSRQ Body Areas Satisfaction subscale, and the BIQ Discrepancy subscale. We labelled it “body image evaluation.” Higher scores on this factor indicate a more negative evaluation of one's appearance. The second factor comprised the MBSRQ Appearance Orientation and BIQ Importance subscales and we called it “body image investment.” Higher scores for the body image investment factor connote greater investment in one's appearance. Sums of the respective scales for each of the two factors, body image evaluation and body image investment, were used as variables in the interaction terms of the moderated regression analyses. The means, standard deviations and coefficients alpha of the measures are presented in Table 1 and were consistent with previous reports using nonclinical samples (e.g., Cash, Ancis, & Strachan, 1997; Cash & Henry, 1995; Cash & Szymanski, 1995; Hewitt et al., 1995). Further, all scales showed adequate internal consistency.3 Table 1. Means, standard deviations, alpha reliability, and zero-order correlations of the uncentered perfectionism, body image, eating disorder symptoms, and body mass index variables Variable M SD α EAT-26 Perfectionistic self-promotion 42.02 11.04 .89 .45* Nondisplay of imperfection 44.67 10.25 .87 .37* Nondisclosure of imperfection 23.10 7.76 .84 .43* MBSRQ Appearance Evaluationa 2.83 .77 .89 .54* MBSRQ Appearance Orientation 3.59 .60 .85 .39* MBSRQ Body Areas Satisfactiona 2.82 .62 .77 .58* BIQ Discrepancy 1.08 .61 .80 .34* BIQ Importance 1.60 .55 .84 .46* EAT-26 65.19 18.83 .90 – BMI 20.64 2.69 – .23 Body image evaluation factor 6.73 1.76 .91 .56* Body image investment factor 5.18 .98 .88 .50* Note: The following labels were used—MBSRQ (Multidimensional Body-Self Relations Questionnaire), BIQ (Body Image Ideals Questionnaire), EAT (Eating Attitudes Test-26 total scores), and BMI (body mass index). a The values reported for the MBSRQ Appearance Evaluation and MBSRQ Body Areas Satisfaction subscales were based on the reverse-scored subscales. The means of these subscales in their original format were 3.17 and 3.18, respectively. * p < .001 (two-tailed). Table options Associations among the perfectionism, body image, and eating disorder symptom variables Zero-order bivariate correlations between the eating disorder symptoms and both perfectionism and body image variables are presented in Table 1. Family-wise Type I error rate was controlled in all analyses. A Bonferroni correction (p = .05/11) was applied to bivariate correlations, resulting in a significance level of .005. Consistent with the notion that perfectionistic self-presentation is associated with eating disorder symptoms, perfectionistic self-promotion, nondisplay of imperfection and nondisclosure of imperfection all exhibited significant positive correlations with EAT-26 total scores. In addition, all three body image evaluation variables (MBSRQ Appearance Evaluation, MBSRQ Body Areas Satisfaction, and BIQ Discrepancy) as well as those reflecting body image investment (MBSRQ Appearance Orientation and BIQ Importance) displayed significant positive correlations with eating disorder symptoms. Testing the diathesis–stress model In order to ascertain whether the relationship between perfectionistic self-presentation and eating disorder symptoms is moderated by body image dysfunction, we conducted a series of hierarchical multiple regression analyses. The following variables were entered into the regression analysis: Step 1—body mass index, Step 2—perfectionism dimension (PSP, NDP, or NDC), Step 3—body image factor (body image evaluation or body image investment), and Step 4—the perfectionism by body image factor product vector. The criterion variable was eating disorder symptoms, as measured by EAT-26 total score. Owing to the statistical difficulty of detecting moderator effects (e.g., McClelland & Judd, 1993), the family-wise Type I error rate was controlled at the .10 level for moderation analyses, resulting in a corrected significance level of .016. Eating disorder symptoms and body image evaluation The results presented in Table 2 indicated that each of the perfectionistic self-presentation facets interacted with body image evaluation to predict unique variance in eating disorder symptoms. That is, the PSP × body image evaluation, NDP × body image evaluation, and NDC × body image evaluation interaction terms served as significant predictors of eating disorder symptoms. It is important to note that when a study successfully detects an interaction, the reduction in variation attributable to adding the interaction term to an additive model is likely to be small (McClelland & Judd, 1993). Therefore, even interactions explaining as little as 1% of the total variance should be considered important (Chaplin, 1991 and Evans, 1985). In the present study, the significant interactions accounted for between 3 and 5% of the total variance in eating disorder symptoms, underscoring the importance of the interactions. Table 2. Summary of Hierarchical Regression Analyses for perfectionism, body image evaluation, and the perfectionism × body image evaluation variables predicting EAT-26 total scores Predictor Total R ΔR2 ΔF df β df Analysis 1: PSP BMI .23 .06* 8.28* 1, 143 .23* 143 PSP .51 .20** 39.05** 2, 142 .45** 142 Evaluation .64 .15** 35.19** 3, 141 .42** 141 PSP × evaluation .68 .05** 13.03** 4, 140 1.33** 140 Analysis 2: NDP BMI .23 .06* 8.28* 1, 143 .23* 143 NDP .43 .13** 23.23** 2, 142 .37** 142 Evaluation .59 .16** 34.03** 3, 141 .46** 141 NDP × evaluation .62 .03* 7.46* 4, 140 1.16* 140 Analysis 3: NDC BMI .23 .06* 8.28* 1, 143 .23* 143 NDC .49 .18** 34.23** 2, 142 .43** 142 Evaluation .62 .15** 34.89** 3, 141 .43** 141 NDC × evaluation .65 .04* 8.55* 4, 140 .95* 140 Note: The following labels were used—BMI (body mass index), SOP (self-oriented perfectionism), SPP (socially prescribed perfectionism), PSP (perfectionistic self-promotion), NDP (nondisplay of imperfection), NDC (nondisclosure of imperfection), evaluation (body image evaluation factor). * p < .016 (two-tailed). ** p < .001 (two-tailed). Table options The significant interactions indicate that the relationship between perfectionistic self-presentation and eating disorder symptoms changes depending on the level of body image evaluation. To clarify the nature of these significant interactions, we calculated the slopes of the regression of eating disorder symptoms on a given facet of perfectionistic self-presentation (e.g., PSP) at two levels of body image evaluation (1SD above and below the mean). This procedure was adopted in each simple slope regression analysis of a significant interaction. The results of the simple slope regression analyses revealed the same pattern for each of the three significant interactions ( Table 3). For example, for the PSP × body image evaluation interaction, the simple slope analysis indicated that the slope for the high value of body image evaluation was significant, but that the slope for the low value was not significantly different from zero. Thus, at high levels of body image evaluation, when women were dissatisfied with how they look, higher levels of perfectionistic self-promotion predicted higher levels of eating disorder symptoms. At low levels of body image evaluation, when women felt their actual appearance was close to their ideal, higher levels of perfectionistic self-promotion did not predict higher levels of eating disorder symptoms. This same pattern of results was found for both the NDP × body image evaluation and NDC × body image evaluation interactions. Table 3. Simple Slope Regression Analyses of significant perfectionism × body image evaluation interactions predicting EAT total scores Predictor t For within-set predictors df β PSP × evaluation PSP at low (positive) evaluation 1.11 140 .10 PSP at high (negative) evaluation 5.82** 140 .51** NDP × evaluation NDP at low (positive) evaluation .10 140 .01 NDP at high (negative) evaluation 3.38* 140 .34* NDC × evaluation NDC at low (positive) evaluation .91 140 .09 NDC at high (negative) evaluation 4.90** 140 .41** Note: The following labels were used—PSP (perfectionistic self-promotion), NDP (nondisplay of imperfection), NDC (nondisclosure of imperfection), evaluation (body image evaluation factor). * p < .016. ** p < .001. Table options Eating disorder symptoms and body image investment As predicted, there were no significant interactions involving body image investment (see Table 4). Thus, the relationship between perfectionistic self-presentation and eating disorder symptoms was not affected by the amount of time spent grooming or by the importance associated with personal appearance. Table 4. Summary of Hierarchical Regression Analyses for perfectionism, body image investment, and the perfectionism × body image investment variables predicting EAT-26 total scores Predictor Total R ΔR2 ΔF df β df Analysis 1: PSP BMI .23 .06* 8.28* 1, 143 .23* 143 PSP .51 .20** 39.05** 2, 142 .45** 142 Investment .58 .08** 16.50** 3, 141 .33** 141 PSP × investment .59 .02 3.30 4, 140 .88 140 Analysis 2: NDP BMI .23 .06* 8.28* 1, 143 .23* 143 NDP .43 .13** 23.23** 2, 142 .37** 142 Investment .56 .13** 25.90** 3, 141 .39** 141 NDP × investment .58 .02 3.95 4, 140 .92 140 Analysis 3: NDC BMI .23 .06* 8.28* 1, 143 .23* 143 NDC .49 .18** 34.23** 2, 142 .43** 142 Investment .61 .14** 30.96** 3, 141 .39** 141 NDC × investment .63 .02 3.82 4, 140 .90 140 Note: The following labels were used—BMI (body mass index), SOP (self-oriented perfectionism), SPP (socially prescribed perfectionism), PSP (perfectionistic self-promotion), NDP (nondisplay of imperfection), NDC (nondisclosure of imperfection), and investment (body image investment factor). * p < .016 (two-tailed). ** p < .001 (two-tailed).