Testing visual self-misperception in anorexia nervosa using a symmetrical body size estimation paradigm

Abstract

Objective: Individuals diagnosed with anorexia nervosa (AN) often report seeing themselves as overweight. While body size estimation tasks suggest that such individuals overestimate their body size, these tasks have failed to establish whether this misestimation stems from visual misperception. Misestimation might, instead, be due to response bias. We designed a paradigm to distinguish between visual and response bias contributions to body size misestimation: the symmetrical body size estimation (s-BSE) paradigm.

Method: The s-BSE paradigm involves two tasks. In the conventional task, participants estimate the width of their photographed body by adjusting the size of a rectangle to match. In the transposed task, participants adjust the size of a photograph of their body to match the rectangle. If overestimation stems exclusively from visual misperception, then errors in each task would be equal and opposite. Using this paradigm, we compared the performance of women diagnosed with AN (n = 14) against women without any eating disorder (n = 40).

Results: In the conventional task, we replicated previous findings indicating that both women with AN and women without any eating disorder overestimate their body size. In the transposed task, neither group adjusted the bodies to be narrower than the rectangle. Participants with AN set their photographs to be significantly wider.

Discussion: While we replicated previous findings of body size overestimation amongst women with AN and those without any eating disorder, our results are inconsistent with the hypothesis that such overestimation stems exclusively from visual misperception and instead suggest a substantial response bias effect.

Public significance: Women with anorexia nervosa overestimate their own body size. Research has not yet determined whether this overestimation stems from them seeing themselves as larger or other, non-visual factors. We employ a new method for distinguishing these possibilities and find that non-visual factors influence size estimates for women with and without anorexia nervosa. This method can help future research control for non-perceptual influences on participant responses.