Validation of the French version of the Binge Eating Scale: Examination of its factor structure, internal consistency and construct validity in a non-clinical and a clinical population

Objective. The Binge Eating Scale is a widely used scale to assess binge eating disorder in obese patients. Until now, this scale has not been validated on a French population, and no psychometrically sound tool assesses binge eating disorder in the French. This study aimed to test the psychometric properties of a French version of the Binge Eating Scale by establishing its factor structure, internal consistency, and construct validity in both a non-clinical population and a clinical population (obese patients who are candidates for bariatric surgery). Methods. - A total of 553 non-clinical subjects and 63 morbidly obese patients who were candidates for bariatric surgery were assessed with the BES and the Bulimic Investigatory Test, Edinburgh or BITE (which assesses both binge eating behaviours and use of inappropriate compensatory behaviours). We tested the factor structure of the instrument, its internal consistency, its construct validity with measures of binge eating, and its construct validity with measures of inappropriate compensatory behaviours to avoid weight gain. In 47 out of the 63 obese patients, we assessed binge eating disorder (SCID). Results.- In the non-clinical population, the BES had a one-factor structure (which accounted for 61% of the variance), excellent internal consistency (alpha = 0.93), and high construct validity with measures of binge eating. In this population, construct validity with measures of inappropriate compensatory behaviours was confirmed in overweight and obese subjects (P = 0.42), but not in underweight and optimal weight subjects (P < 0.001). In obese patients candidates for bariatric surgery, we demonstrated that the BES had a one-factor structure (which accounted for 46% of the variance), had high internal consistency (alpha= 0.88) and high construct validity with measures of binge eating and good construct validity with measures of inappropriate compensatory behaviours to avoid weight gain. In the subpopulation of 47 obese patients, sensitivity, specificity, positive predictive value and negative predictive value were respectively 75%, 88.4%, 37.5% and 97.4% (BES threshold = 18). Discussion.- In this study, we validated a psychometrically sound French version of the Binge Eating Scale, both in a non-clinical and a clinical sample. The psychometric properties of the French version of the BES are comparable to its original version with a one-factor structure. The BES is a useful tool to assess binge eating disorder in obese patients (e.g., bariatric surgery candidates), but might not differentiate between binge eating disorder and bulimia nervosa in underweight and optimal weight subjects. (C) 2016 L'Encephale, Paris.