Mirror Exposure in Binge-Eating Disorder: Changes in Eating Pathology and Attentional Biases

What is the public health significance of this article? This study suggests that stand-alone mirror exposure contributes to improvement of shape concerns and eating pathology in individuals with binge-eating disorder. Whether these changes occur by changes in selective body-related attention needs further clarification. Objective: Body image disturbance is crucial in the psychopathology of binge-eating disorder (BED). According to cognitive models of eating disorders, biased attentional processes contribute to the development and maintenance of body image concerns. Mirror exposure constitutes an effective method to improve body image. The present study investigated if stand-alone mirror exposure reduces body image concerns, eating pathology as well as body-related attention processes in BED. Method: Women with BED were assigned to an intervention group (IG; four sessions of mirror exposure) or a waiting list control group (WL; 4 weeks waiting period). Prior to and after the intervention/waiting period, body image concerns, and attentional processes toward the self- and a control-body were measured using questionnaires and two eye-tracking tasks. All analyses were conducted on intent-to-treat (ITT) bases. Results: Relative to the WL, the IG reported lower levels of shape concerns and reduced eating pathology after the intervention. Furthermore, while at baseline attention for the most unattractive body part was higher for the self-body (SB) relative to the control body (CB) in both IG and WL, no such difference was found after the intervention in the IG. By contrast, in the WL attention allocation between SB and CB remained unchanged. Conclusions: Evidence was found that stand-alone mirror exposure may contribute to improving shape concerns and the reduction of eating pathology in women with BED. Furthermore, mirror exposure impacts selective body-related attention even though generalizability of results is limited by small sample size, lack of statistical power, and expectancy effects due to comparison with WL.