Alcohol and inflammation: Examining differences at the intersection of sexual identity and race/ethnicity

Sexual minorities (SMs; e.g., lesbian, gay, bisexual, and other non-heterosexual individuals) are more likely to be current alcohol drinkers than their heterosexual peers while separately experiencing elevated inflammation. Yet, little research has assessed the association between alcohol use and inflammation among subgroups of SMs, let alone potential differences among people with multiple marginal identities (e.g., race/ethnicity and sexual identity). Data came from the National Health and Nutrition Survey 2015-2016. Survey-weighted multivariable linear regression analysis was used to assess the relationship between alcohol use categories, heavy episodic drinking, and log-CRP (C-reactive protein). Models were stratified by sexual identity to determine whether associations between alcohol use and inflammation or between race/ethnicity and inflammation differed by sexual identity. Among 3220 participants, 1000 (36.8%) reported light alcohol use, 870 (32.0%) reported moderate use, and 483 (17.8%) reported heavy use. Mean raw CRP was 4.1 mg/L (SD = 8.1). The association between race/ethnicity and CRP differed in stratified relative to non-stratified models with key differences in CRP among individuals with multiple marginalized identities. We also observed that while the "classic" J-shaped relationship between alcohol use and systemic inflammation persists among heterosexuals in this sample, it does not hold among subgroups of sexual minorities. In particular, bisexuals who report heavy alcohol use, compared to non-users, experience significantly elevated CRP. Finally, we did not observe any association between heavy episodic drinking and CRP among subgroups of sexual minorities. Future studies assessing alcohol and biomarker data need to strive to include subgroups of sexual minorities and people with multiple marginal identities to better target behavioral and biomedical interventions aimed at reducing health disparities.