Development and validation of the BRief Eating Disorder Screener (BREDS) for US veterans in healthcare and community settings.

Abstract

Objective: To develop a Diagnostic and Statistical Manual of Mental Disorders (DSM-5) eating disorder screener.

Method: Veterans enrolled in VA healthcare (N = 344) completed a survey of screening items and established measures. A validation subset (n = 166) participated in diagnostic interviews to confirm an eating disorder diagnosis. Items most predictive of diagnosis were included or adapted for a confirmatory national survey (N = 405).

Results: Item selection models using machine learning techniques resulted in a five-item screen with good sensitivity, specificity and area under the curve (0.75, 0.87, and 0.83). Items included: (1) Do you do things to rid your body or stomach of calories or food (vomiting, excessive exercise, laxative or diuretic use, or something else)? (2) Do you ever eat an extremely large amount of food without thinking? (3) Are you or other people in your life concerned about a recent change in your weight? (4) Is your weight or shape one of the main ways you judge yourself? (5) Do you need to eat in order to get back to sleep when you wake up at night?

Conclusion: The 5-item BRief Eating Disorder Screener (BREDS) is the first screen developed and validated for a broad range of DSM-5 eating disorder diagnoses.