Perfectionism, low self-esteem, and family factors as predictors of bulimic behavior

Abstract

A previous study found that self-reported body dissatisfaction, depression, and peer pressure to maintain a thin body shape were significant predictors of bulimic behavior in college women, but that family functioning was not a significant predictor [Eat. Behav. 2 (2001) 323]. The current study examined whether perfectionism, low self-esteem, and a more specific family variable—perceived pressure from the family to be thin—predicted any additional variance in eating-disordered behavior after significant variables from the previous study had been taken into account. As in the previous study, self-reported body dissatisfaction, depression, and peer pressure to maintain a thin body shape were significant predictors of bulimic behavior. Perceived weight-related pressure from the family was also a significant predictor. In contrast, high parental expectations were found to predict lower levels of bulimic behavior and to moderate the effects of peer influence on bulimic behavior. The variables found in this study to be related to bulimic behavior may be useful targets for clinical intervention for women with disturbed eating patterns.

Introduction

To be diagnosed with bulimia nervosa, an individual must engage in binge eating and compensatory behaviors, such as self-induced vomiting or excessive exercise, at least twice a week for a period of 3 months or more (DSM-IV, American Psychiatric Association, 1994). Although diagnosable cases of bulimia nervosa are fairly rare, the prevalence of bulimic behavior that falls short of meeting DSM criteria is much higher (Fairburn & Beglin, 1990). Research with clinical samples of individuals with bulimia nervosa and with college students with less severe bulimic behavior has identified a variety of variables associated with bulimic behavior. These variables include social pressure from family, peers, and the media and individual variables, such as self-esteem, perfectionism, body dissatisfaction, and depression.

Women with bulimia or with bulimia-like behavior often report their families to be low in expressiveness, cohesion, and caring and high in anger and conflict Johnson & Flach, 1985, Kent & Clopton, 1992. Other studies have suggested that family members may model maladaptive eating attitudes and behaviors for adolescent girls and young women. Mothers of daughters with bulimia or with symptoms of eating disorders are often critical of their daughters' weight and physical attractiveness, and they often have a history of disordered eating or dieting themselves (e.g., Pike & Rodin, 1991). These studies support the notion that daughters may often adopt their parents' concerns about body shape and weight.

Several studies have examined the relation between peer influence and binge eating. Crandall (1988) found a positive relationship between binge eating and popularity in a study of university women in two sororities. Criticism from peers about body appearance has also been shown to correlate with increases in bulimic behavior (Cattarin & Thompson, 1994).

Low self-esteem (i.e., a large discrepancy between the ideal image of oneself and the self that one perceives as real) can result in a variety of negative consequences, including eating disorders (Bardone, Vohs, Abramson, Heatherton, & Joiner, 2000). Low self-esteem was found to correlate significantly with scores on an eating disorder measure in a sample of high-school students (Fisher, Pastore, Schneider, Pegler, & Napolitano, 1994).

According to one theoretical model, perfectionism consists of the following features: (a) an excessive level of worry over making mistakes; (b) constant doubts about the quality of one's performance, which may delay task completion; (c) overvaluation of parental expectations; and (d) excessive emphasis on order and precision (Frost, Marten, Lahart, & Rosenblate, 1990). Perfectionism has been identified as a risk factor for bulimic behavior Hewitt et al., 1995, Joiner et al., 1997.

Bulimic behavior has been found to be associated with body dissatisfaction (i.e., negative feelings about one's body) in both clinical and nonclinical samples (Williamson, Cubic, & Gleaves, 1993). Women face greater cultural pressure than men to strive for physical attractiveness, and they show greater dissatisfaction with their bodies than men Feingold & Mazzella, 1998, Striegel-Moore et al., 1986. In one study, body dissatisfaction was the primary determinant of disordered eating among college women, and sociocultural pressures to be thin contributed to disordered eating by increasing students' body dissatisfaction (Stice, Schupak-Neuberg, Shaw, & Stein, 1994).

Individuals with eating disorders often have other serious problems, and depression is the most common comorbid diagnosis (Braun, Sunday, & Halmi, 1994). In one sample of women with bulimia, most exhibited symptoms of mild to moderate depression, and many also reported having a close relative with an affective disorder (Lee, Rush, & Mitchell, 1985). In another study, women with bulimic behavior reported significantly higher levels of depression than either obese women or women in a normal comparison group (Williamson, Kelley, Davis, Ruggiero, & Blouin, 1985).

A study by Young, McFatter, and Clopton (2001), using a sample of college women, examined the relationship between bulimic behavior and the following eight variables: family functioning, body dissatisfaction, depression, peer influence, media influence, extraversion, neuroticism, and weight ratio (WTR; each participant's reported weight divided by the standard weight for women of that height). These variables were assessed using self-report questionnaires, with responses on a measure of bulimic behavior as the criterion variable. The results of the Young et al. study indicated that the following four variables were significant predictors: body dissatisfaction, depression, peer influence, and extraversion. Although media influence was not in itself a significant predictor of bulimic behavior, it interacted with body dissatisfaction and with WTR to increase the likelihood of bulimic behavior. In addition, an increase in WTR was found to be a better predictor of bulimic behavior in depressed women than in women who were not depressed.

The current study expanded on the Young et al. (2001) study in three important ways. First, it examined whether perfectionism and low self-esteem predicted any additional variance in bulimic behavior in the presence of variables that had already been found to be significant: body dissatisfaction, depression, and peer influence. The purpose of including perfectionism and low self-esteem with the other predictors was to attempt to identify correlates of eating-disordered behavior that would make useful targets for therapeutic intervention. Second, given that self-reports of global family functioning were not significantly related to bulimic behavior in the previous study, the current study instead used a more specific measure of weight-related family attitudes. Specifically, this study examined whether self-reports of perceived pressure from family members to maintain a thin body shape accounted for any additional variance in bulimic behavior in comparison to the other variables. Third, the current study examined whether any of the subscales of the Frost et al. (1990) measure of perfectionism accounted for more variance in bulimic behavior than the global perfectionism score for that measure.