عزت نفس ضمنی در بولیمیا

Implicit and explicit self-esteem were compared in a group of female participants with bulimia nervosa or binge eating disorder (n = 20) and a healthy control group (n = 20). Lower explicit and a less positive implicit self-esteem bias in the clinical group was predicted. Participants completed a self-esteem implicit association test and two explicit self-esteem measures. The eating disordered group had lower explicit self-esteem, but a more positive implicit self-esteem bias than controls. The results are discussed in relation to the idea that discrepancies between implicit and explicit self-esteem reflect fragile self-esteem and are related to high levels of perfectionism, which is associated with eating disorders.

Cognitive behavioural models of bulimia nervosa (BN) emphasise the central importance of overvalued ideas about body, weight, and shape in the aetiology and maintenance of bulimic symptoms (Cooper and Fairnburn, 1993, Fairburn and Cooper, 1997, Vitousek, 1996 and Vitousek and Hollon, 1990). These models suggest that BN occurs because low self-esteem becomes linked with weight and shape concerns, which in turn become linked with self-worth. The models propose that low self-esteem is the most proximal risk factor for developing overvalued ideas about body, weight, and appearance (Meijboom, Jensen, Kampman, & Schouten, 1999) and is therefore a key maintaining factor in the disorder (Fairburn, Cooper, & Shafran, 2003). If self-esteem acts centrally in the cycle perpetuating eating disorder symptoms, then treatment outcome should improve if self-esteem is addressed directly (Fairburn et al., 2003 and Yellowlees, 1997). Patients who respond poorly to treatment do indeed have lower levels of self-esteem than those who respond well (Baell and Wertheim, 1992, Fairburn et al., 1987 and Fairburn et al., 1993). Likewise, treating self-esteem improves eating disorder symptomatology (Newns, Bell, & Thomas, 2003) and raising self-esteem helps to maintain change following treatment (Beresin et al., 1989, Hsu et al., 1992, Peterson and Rosenvinge, 2002 and Rorty et al., 1993). However, more general evidence for the role of self-esteem in BN is inconsistent. A number of studies have failed to find a relationship between global measures of self-esteem and BN (Bell, 2002, Garner et al., 1993 and Turnbull et al., 1997). As a consequence, self-esteem is not listed as a primary predictor of outcome for the treatment of BN in the current NICE guidelines (National Institute of Clinical Excellence: National Collaborating Centre for Mental Health, 2004). This failure to find a clear and consistent relationship between self-esteem and BN may be related to the conceptualisation and measurement of self-esteem as a single, global domain of self-concept. This unitary conceptualisation has been open to criticism (see Marsh, Craven, & Martin, 2006 for example), and in the eating disorders more specific measurements of weight and shape based self-esteem do a better job of distinguishing individuals with and without eating disorders than global self-esteem self-report questionnaires (Geller et al., 1998). Another significant limitation to the investigation of self-esteem in BN to date is that current studies rely exclusively on explicit measures of self-esteem, albeit global or specific. Explicit self-esteem is viewed as conscious feelings of self-evaluation, whereas implicit self-esteem is thought to involve unconscious, automatic self-evaluations. Dual process models of information processing provide a theoretical framework that points to the need to examine both implicit and explicit self-esteem. For example, Smith and DeCoster (2000) propose that there are two distinct processing modes that draw on the memory system in contrasting ways. One is an associative processing mode that operates pre-consciously and which produces affective responses to events or stimuli. The other is a rule-based processing mode that operates through using symbolic representations and is characterised by conscious effortful retrieval of information. Strack and Deutsch (2004) propose a model that draws on dual processing theory in which implicit evaluations are related to impulsive processes and explicit evaluations are related to reflective processes. Reflective and impulsive processes lead to different behavioural consequences. Thus, behavioural decisions elicited as a consequence of knowledge about facts and values are generated by the reflective system, and behaviour generated through associative links and motivational orientations is elicited by the impulsive system. If current cognitive models of eating disorders are correct in giving prominence to self-esteem as a maintaining factor in eating disorders, then we need to examine explicit and implicit self-esteem separately because they could influence behavioural outputs differently. For example, binge eating is frequently initiated impulsively and one of the diagnostic criteria for bulimia nervosa is a sense of loss of control. Dual process theory suggests that the implicit self-evaluations that characterise self-esteem are important determinants in the generation of impulsive actions and therefore implicit self-esteem may play a more important role in initiating binge episodes than explicit self-esteem. The most widely used measure of implicit self-esteem is the Implicit Associations Test (IAT; Greenwald and Farnham, 2000 and Greenwald et al., 1998). The IAT measures the relative strengths of automatic associations between target concepts and specific attributes so that in the self-esteem IAT (SE-IAT: Greenwald & Farnham, 2000), the concepts of self and other are paired with positive and negative words (a full description of the IAT is given in the method section). The idea behind the IAT is that when the associations presented are congruent with those stored in memory (e.g. self and positive attributes) participants will be faster to respond than when they are incongruent (e.g. self and negative attributes). There is consistent evidence that healthy individuals display a positive implicit self-esteem bias using the IAT (Greenwald et al., 2002). Although to our knowledge, there is no research on implicit self-esteem in eating disorders, in other clinical samples, implicit self-esteem remains positive albeit reduced compared with healthy individuals. For example, in recovered (Gemar, Segal, Sagrati, & Kennedy, 2001), currently depressed (De Raedt, Schacht, Franck, & De Houwer, 2006), and currently depressed individuals with suicidal ideation (Franck, De Raedt, Dereu, & Van den Abbeele, 2007) a positive implicit self-esteem bias remained. There is also some evidence of a reduced positive bias in high socially anxious groups (de Jong, 2002 and Tanner et al., 2006). One suggestion for why individuals retain a positive implicit self-esteem bias in clinical disorders, is that positive implicit self-esteem acts as a defence against threats to the self (Rudman, Dohn, & Fairchild, 2007). This threat defence mechanism may be a normal healthy bias, which remains intact, albeit often somewhat reduced in clinical populations. The aim of our study was to investigate implicit and explicit self-esteem in a group of participants with eating disorders and a healthy control group. We predicted that the eating disordered group would have lower levels of both global and specific (weight and shape based) explicit self-esteem and a less positive implicit self-esteem bias compared to individuals without eating disorders. As depression is often present as a co-morbid diagnosis with eating disorders, we measured the severity of depressive symptoms using the Beck Depression Inventory (Beck, Steer, & Brown, 1996) in order to control for the effects of depression on explicit and implicit self-esteem.