Gender Identity Moderates the Association Between Alcohol Use and Eating Disorder Risk in U.S. College and University Students.

Abstract

Objective: Studies have demonstrated associations of alcohol use and alcohol use disorder (AUD) with eating disorders; however, it is unknown whether the pattern of association differs by gender identity, particularly for transgender or gender-expansive (TGE) people.

Methods: Data from a US national sample of college and university students (n = 21,623) participating in the Healthy Minds Study were used to examine associations of Alcohol Use Disorder Identification Test (AUDIT) scores, categorized into no drinking, low-risk drinking, and mild, moderate, and severe AUD, with a positive screen on the five-item SCOFF eating disorder screening tool. Predicted prevalence estimates were calculated from logistic regression models that included interaction terms for gender identity by level of alcohol involvement using marginal standardization.

Results: The range of predicted prevalence of a positive eating disorder screen between those who did not drink alcohol and those with severe AUD was greatest among TGE people (26.0%-69.3%, prevalence difference [PD] 43.3, p<.001), followed by cisgender women (27.0%-52.3%, PD = 25.3, p<.001), and cisgender men (18.2%-35.0%, PD = 16.8, p<.001). Among cisgender women the association was monotonic, while among cisgender men and TGE people, there was a threshold effect, with increased predicted prevalence only among those with scores in the severe or moderate and severe AUD categories.

Conclusions: These findings suggest patterns of association between level of alcohol involvement and eating disorders among college and university students differ by gender identity, underscoring the need for targeted, gender identity-sensitive measures to aid the prevention and treatment of this comorbidity.