Methods: Data were drawn from a national convenience sample of LGBTQ college students who completed a web-survey (n=952; age M=22.8 years; cisgender men 34.8%; cisgender women 44.6% transgender 20.6%; person of color 24.7%; gay/lesbian 52.2%). Problem drinking was assessed using the AUDIT scale (theoretical range 0-40; ≥ 8 indicates problem drinking).

We used logistic regression to examine relationships between problem drinking and gender, race, demographics, social support, LGBTQ-identity variables (outness, saliency, pride, shame), perceived LGBTQ campus climate, and LGBTQ discrimination (microaggressions, victimization).

Analysis was conducted with the full sample, then stratified by transgender/cisgender status and further stratified by female/male among cisgender students. Each model tested interactions between race/ethnicity and LGBTQ-identity variables. Monte-Carlo simulation was used in the logistic regression intersectionality estimates to adjust for multiple comparisons.

Results: The average AUDIT score was 5.6, with approximately 19% indicating problem drinking. Problem drinking did not vary significantly between cisgender and transgender students. Among cisgender students, men (24.2%) had a significantly higher rate of problem drinking than women (16.6%, p=.037).

Regression Results: For cisgender and transgender students combined, microaggressions significantly increased the odds of problem drinking (AOR=1.41, p=.004). Regarding the intersection of race and LGBTQ identity, LGBTQ shame among students of color was associated with higher odds of problem drinking (AOR=2.28, p=.005).

Comparing transgender and cisgender students, positive perceptions of campus climate was significant for only transgender students (AOR=0.31, p=.035) and microaggressions was only significant for cisgender students (AOR=1.48, p=.004). A significant interaction between being a White transgender student and outness was found (AOR=2.66, p=.048), indicating that greater outness was associated with higher odds of problem drinking.

The cisgender sample had the same trend as the overall sample. Microaggressions increased the odds of problem drinking (AOR=1.48, p=0.004). Greater shame was associated with problem drinking among cisgender students of color (AOR=2.39, p=0.007). For women, only microaggressions was a risk factor (AOR=1.57, p=.042) and there were no interactions between LGBTQ identity and race/ethnicity. For cisgender men, only microaggressions was a risk factor (AOR= 1.54, p=.030); disturbance was significant risk factor among men of color (AOR=2.29, p=.042).

Implications: Research that actively attends to race, gender, and other intersecting identities is needed to advance knowledge and inform empirically grounded interventions that can support diverse LGBTQ students. The findings suggest that interventions to reduce problem drinking among LGBTQ students must consider the role of intersecting social locations. Implications for social work research and practice will be discussed.