Eating Disorder-15 (ED-15): Factor structure, psychometric properties, and clinical validation

Abstract

Objective The purpose of the current study was to explore the factor structure and psychometric properties of the Portuguese version of the Eating Disorder-15 (ED-15), as well as to establish cutoff scores and normative data for the Portuguese version. Methods Participants from a nonclinical sample (N = 860) and an eating disorders clinical sample (N = 260) were invited to complete a set of questionnaires, including the Portuguese version of the ED-15. Results The first-order two-factor structure originally proposed by the ED-15 authors was endorsed through a confirmatory factor analysis (chi(2)/df = 2.610; standardized root-mean-square residual = 0.0325; root-mean-square error of approximation = 0.079; Tucker-Lewis index/goodness-of-fit index/incremental fit index > 0.95). Items revealed adequate construct validity (lambda = .54-.90; R-2 = .29-.81). The ED-15 revealed excellent internal consistency (alpha = .91) and temporal stability (intraclass correlation coefficient = .92; 95% CI [.84-.95]). Normative data for the ED-15 were provided. The ED-15 demonstrated acceptable concurrent and convergent validity. Receiver operating characteristic analysis revealed that the ED-15 total score accurately discriminates between participants with and without an eating disorder (area under de curve = .80; SE = .017; p <= .001; 95% CI [.766-.834]). A cutoff score for clinical significance and a reliable change index were computed. Conclusions The Portuguese version of the ED-15 is a reliable and valid measure of eating psychopathology and symptoms, whenever a brief measure is needed, as in session-by-session assessment of therapy progress and outcome.