Women with anorexia nervosa and bulimia nervosa : Individual and family characteristics, with particular emphasis on perfectionism
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
This study investigated socio-cultural, family and individual factors associated with anorexia and bulimia nervosa, with particular emphasis on dysfunctional perfectionism, and adopting a general social learning perspective. Theories of the development of eating disorders were interwoven with theories of the development of perfectionism. A model was proposed for the development of anorexia and bulimia nervosa via a dysfunctional perfectionism pathway. The 135 participants, aged 18 to 40 years, were women with anorexia nervosa (N=25), bulimia nervosa (N=32), Type 1 diabetes (N= 53, a North Canterbury population-based sample), and healthy women students (N=25). The women with eating disorders were recruited from various treatment centres throughout New Zealand. Participants completed a battery of seven self-report psychometric tests, namely, the Eating Disorder Inventory-2 (EDI-2), Beck Depression Inventory (BDI), Multidimensional Perfectionism Scale (MPS), Setting Conditions for Anorexia Nervosa Scale (SCANS), Tridimensional Personality Questionnaire (TPQ), Parental Bonding Instrument (PBI), and Family Environment Scale (FES). Analysis of Covariance, using the BDI as a covariate, revealed that, in addition to measures concerned with weight, shape and dieting, both anorexia and bulimia nervosa group means were significantly higher than both healthy and diabetes group means for EDI-2 Interpersonal Distrust and Social Insecurity; MPS Concern over Mistakes, Personal Standards, and Parental Criticism; and TPQ Harm A voidance, and significantly different from the healthy group mean for MPS Parental Expectations; SCANS Perfectionism; and PBI Maternal Protection, Maternal Care, and Paternal Care. Correlational analyses confirmed hypothesized moderate or strong associations between some perfectionism measures and other characteristics of women with eating disorders, such as a harm-avoidant temperament, and perceptions of maternal overprotection. Discriminant function analysis revealed seven variables, in combination, that maximally discriminated between eating disordered and non-eating disordered groups: three EDI-2 variables of Drive for Thinness, Ineffectiveness, and Social Insecurity, three MPS subscales of Concern over Mistakes, Personal Standards, and Doubts about Actions, and the BDI. Of the three instruments measuring perfectionism, in this study, only the MPS effectively discriminated between eating disordered and non-eating disordered groups. Findings indicated the importance of controlling for depression when comparing eating disordered groups with other groups, and that dysfunctional perfectionism is largely independent of the mood of the respondent. Findings suggest that the PBI may be limited by cultural sensitivity. Findings led to questioning of the applicability of the EDI-SC to diabetes groups and of the validity of the Novelty Seeking and Reward Dependence Dimensions of the TPQ. In concluding that dysfunctional perfectionism is a key personality characteristic of women with anorexia and bulimia nervosa, it is argued that multidimensional measures of perfectionism provide more insight than unidimensional measures into the dysfunctional facets of perfectionism, and that perfectionism per se is not necessarily problematic. Dysfunctional perfectionism may distinguish psychopathology associated with anorexia and bulimia nervosa from numerous other forms of psychopathology, including depression. Although aetiological factors were not assessed in this study, the MPS and PBI, considered in conjunction with the theoretical literature, may provide insight into the development of dysfunctional perfectionism. This has implications for the treatment and prevention of eating disorders.