Psychometric properties and factor structure of the Eating Disorder Examination-Questionnaire in omnivores, vegetarians, and vegans

Abstract

The psychometric properties of the Eating Disorder Examination-Questionnaire (EDE-Q) have been widely reported, but there remains considerable uncertainty regarding the factor structure of the tool, with vegetarians and vegans remaining no exception. Due to the utility of the EDE-Q as a screening and outcome measure, we assessed the theoretical assumptions and psychometric properties of the EDE-Q in vegetarians (n = 278), vegans (n = 580), and omnivores (n = 413) separately, owing to the expectation of different structures within each dietary population given their varying degrees of restraint. We undertook confirmatory factor analysis of five models previously described demonstrating inadequate fit. Exploratory factor analysis supported unique three-factor models across dietary groups incorporating Weight and Shape Concern subscales, potentially suggesting that community samples of varying of dietary restraint consistently interpret weight/shape to be related to general body image concerns. These novel findings also suggest a shorter version of the EDE-Q may be more suitable in community samples to reduce the time burden of the tool. The predominately poor test-retest reliability raises doubt around the overall utility and stability of the EDE-Q in all dietary populations, regardless of whether a shortened version is employed. Future research is needed to validate of other eating disorder tools across dietary populations.

Keywords: Eating Disorder Examination-Questionnaire; Eating disorder; Factor analysis; Psychometric properties; Veganism; Vegetarianism.