Regimented and lifestyle restraint in binge eating disorder.

OBJECTIVE

This study tested the psychometric properties of two commonly used measures of dietary restraint, the Three Factor Eating Questionnaire and the Eating Disorder Examination Questionnaire.

METHOD: Restraint data from 512 overweight/obese participants with binge eating disorder (BED) were subjected to exploratory and confirmatory factor analyses.

RESULTS

Factor analyses of the restraint variables indicated a two-factor solution, interpreted as "Regimented" and "Lifestyle" restraint. Stepwise regression analyses revealed that Regimented restraint was more predictive of eating pathology, whereas Lifestyle restraint appeared to be protective of eating problems. Neither type of restraint was related to binge eating. Cluster analysis of the restraint dimensions yielded three distinct subgroups of patients who differed significantly on several important eating- and weight-related features.

DISCUSSION: Future research is needed to test the significance of these restraint constructs over time in both the development of obesity and binge eating problems as well as their treatment.