The link between discrimination and cannabis use varied significantly across sexual orientation and gender identity groups | Research

The takeaway

Analysis of 98,820 participants in the All of Us Research Program found that discrimination had a nonlinear relationship with alcohol and cannabis use, with associations particularly varying among gender minority groups and cisgender sexual minority men.

LGBTQ+ health researchers, substance use counselors working with SGM populations, and public health professionals addressing health disparities.

Nonlinear: discrimination-cannabis link weakens at very high discrimination

What the researchers found

Discrimination (measured by the Everyday Discrimination Scale) was positively associated with cannabis use at low-to-moderate levels but the association was not significant at very high discrimination levels. The relationship between discrimination and substance use varied significantly by sexual orientation and gender identity, with pronounced differences among gender minority groups.

Why it matters

The nonlinear finding is important: at very high levels of discrimination, the association with substance use weakened or reversed, potentially reflecting withdrawal, avoidance coping, or other survival strategies that differ from moderate-stress substance use patterns.

The numbers in context

98,820 participants. Mean EDS scores highest among gender minority people assigned female at birth (14.78) and lowest among cisgender heterosexual men (6.14). EDS positively associated with cannabis use at low-moderate levels. Association not significant at 2+ SD above mean EDS. Interaction by sexual orientation and gender group significant (p<.05).

How the study worked

Cross-sectional analysis of 98,820 participants from the All of Us Research Program (2017-2022). Adjusted logistic regression examined the relationship between Everyday Discrimination Scale scores and past 3-month cannabis use. Interaction terms assessed differences across sexual orientation and gender modality groups.

What this study cannot tell us

Cross-sectional design. All of Us Research Program may not be representative of all SGM populations. Self-reported discrimination and substance use. Imprecise estimates for some subgroups due to small sample sizes.

How to read the evidence

Strong: large sample from a national research program with appropriate statistical methods and subgroup analysis.

When this study was published

Published 2026. Data from 2017-2022.

The bigger picture

Most minority stress research assumes a linear relationship between discrimination and substance use. This study reveals a more nuanced picture where very high discrimination may paradoxically not predict increased use, challenging simple stress-coping models.

Questions still open

Why does the discrimination-cannabis link weaken at very high discrimination levels? Do different SGM subgroups use cannabis for different reasons? Would longitudinal data show different patterns over time?

Common questions

Does discrimination increase cannabis use in LGBTQ+ people?

The relationship is nonlinear. Low-to-moderate discrimination was associated with more cannabis use, but at very high levels of discrimination, the association weakened or disappeared.

Do all LGBTQ+ subgroups show the same patterns?

No. The relationship between discrimination and substance use varied significantly across groups, with particular differences among gender minority people and cisgender sexual minority men.

Read the original research

Alcohol Use, Cannabis Use, and Discrimination by Sexual Orientation and Gender Identity Within the All of Us Research Program.

LGBT health, 13(1), 11-22

View on PubMed ↗Publisher’s paper ↗

Citation

Sunder, Gowri; Tran, Nguyen K; Peña, Juan M; Lunn, Mitchell R; Obedin-Maliver, Juno; Flentje, Annesa. (2026). Alcohol Use, Cannabis Use, and Discrimination by Sexual Orientation and Gender Identity Within the All of Us Research Program.. LGBT health, 13(1), 11-22. https://doi.org/10.1177/23258292251390584