Differential item functioning analysis for the Eating Disorder Examination Questionnaire among heterosexual and sexual minority men.

Abstract

Sexual minority (SM) men are at a higher risk for eating disorders and related issues, relative to heterosexual men. However, it is currently unknown whether commonly used measures of eating pathology are appropriate to use among diverse groups of men. Determining the unique functioning of existing assessments may help better and more accurately understand eating disorder pathology within this population. The present study examined differences in item endorsement between sexual orientation in the Eating Disorder Examination Questionnaire (EDE-Q) through differential item functioning (DIF). Heterosexual and SM men (N = 703) completed the EDE-Q and a demographic questionnaire. EDE-Q scores were examined for clinically significant DIF based on participants' self-reported sexual orientation (e.g., heterosexual men vs SM men). SM men reported higher EDE-Q symptom composite scores than heterosexual men. DIF was observed for all EDE-Q items relative to the global score; however, only one item met clinical significance (EDE-Q #19; ∆R2 ≥ 0.13). Results suggest that SM men experience greater levels of ED pathology than heterosexual men. While the EDE-Q is a commonly used measure of eating pathology, findings suggest that sexual orientation bias may impact many items on the EDE-Q. However, results from this study indicated that only one item introduces bias that has clinical implications. Additional research is needed to further explore and replicate this finding among more diverse samples of SM and heterosexual men. Level III, case-control analytic study.