The Sum Is Greater than its Parts: Intersectionality and Measurement Validity of the Eating Disorder Examination Questionnaire (EDE-Q) in Latinx Undergraduates in the United States

Previous research has demonstrated disparities in the diagnosis and treatment of eating disorders for men and ethnic minorities. The current study sought to investigate differences in eating pathology and measurement validity of the Eating Disorder Examination Questionnaire (EDE-Q) at the intersection of gender and ethnicity in a nonclinical sample of undergraduates. A total sample of 1173 undergraduates (177 Latino men, 133 non-Latino White men, 554 Latina women, 309 non-Latina White women) completed the EDE-Q as part of a larger study on eating habits. A series of ANOVAs indicated a similar pattern of effects for gender and ethnicity for the original four-factor, 22-item EDE-Q and an alternative three-factor, seven-item variant. Women scored significantly higher than men on all subscales, and Latinx participants scored significantly higher than non-Latinx White participants on the Shape Concern and Weight Concern subscales. A significant interaction of gender by ethnicity was observed on the Shape Concern subscale in which larger effects of ethnicity were observed for men compared to women. Tests of confirmatory factor analysis suggested only the three-factor, seven-item EDE-Q provided acceptable fit and demonstrated scalar invariance across gender and ethnicity. Our study highlights the importance of examining eating pathology and measurement validity at the intersection of demographic characteristics. Results suggest that Latino men may be at higher risk of eating pathology relative to non-Latino White men and that a three-factor, seven-item EDE-Q appears to be a valid measure of eating pathology across gender and ethnicity among U.S. undergraduates.