Evidence for a specific distortion in perceptual body image in eating disorders: a replication and extension

A core feature of eating disorders, such as anorexia nervosa, is an overestimation of body size. A key question is whether this overestimation arises solely from body image concerns typical in eating disorders, or if there is an additional perceptual disturbance. To address this question, we applied a two-component model of body size estimation that has been thoroughly replicated in the body image literature concerning healthy individuals. This model shows statistically independent, additive effects on body size estimates of: a) body image concerns, and b) a perceptual component known as contraction bias. Here body image concerns were defined by a principal components analysis (PCA) of psychometric tasks including the: Eating Disorder Examination Questionnaire, Beck Depression Inventory, Body Shape Questionnaire, Rosenberg Self-Esteem Scale, and Sociocultural Attitudes Towards Appearance Questionnaire-4. The PCA identified three components referred to as PSYCH, FAMPEER, and ATHIN. We investigated the influence of age, personal body mass index (BMI), and these three components on body size estimation in 33 women with a current or past history of eating disorders and 100 healthy controls. Low-BMI control participants overestimated their size, while high-BMI controls underestimated their size, exhibiting the expected normal perceptual contraction bias. However, the women with a history of eating disorders showed no evidence of contraction bias, suggesting a different processing of perceptual aspects of body size estimation compared to controls. We discuss two putative mechanisms that can explain these differences in accuracy of personal body size estimation.