This study sought to confirm a previously identified race by sexual orientation interaction and to clarify men’s alcohol-related risk by using an expanded classification of sexual orientation.

We collapsed three waves of National Alcohol Survey data, restricting the analytic sample to White (n = 5,689), Black (n= 1,237), and Latino (n = 1,549) men with complete information on sexual orientation and alcohol use. Using self-reported sexual identity and behavior, respondents were categorized as exclusively heterosexual (referent), behaviorally discordant heterosexuals (i.e., heterosexual identity and same-sex partners), or gay/bisexually identified men. We used multivariable logistic regression to model lifetime alcohol dependence symptoms, lifetime drinking-related consequences, and past-year hazardous drinking, controlling for age, education, employment, and relationship status and accounting for the complex survey design.

There was no difference in risk of past-year hazardous drinking and lifetime drinking-related consequences between heterosexual, behaviorally discordant heterosexual, and gay/bisexual men, independent of race/ethnicity. Among Black men, behaviorally discordant heterosexuals had three-fold higher odds of lifetime alcohol dependence symptoms than exclusively heterosexual peers (aOR = 3.30, 95% CI [1.19, 9.18], p = .02). Gay/bisexual Latino men had marginally significantly lower odds of lifetime alcohol dependence symptoms (aOR = 0.36, 95% CI [0.12, 1.03], p = .06).

There is little support for broad statements of greater alcohol risk among gay/bisexual men; however, for some subgroups and outcomes the direction and degree of risk depend on race/ethnicity. Thus, this study underscores the importance of considering the potential interaction of sexual orientation and race/ethnicity, which may exacerbate or attenuate alcohol-related risk.