The study, led by David A. Stupplebeen of the University of Hawaiʻi at Mānoa, with Candice D. Donaldson and Tam D. Vuong of the California Department of Public Health and UC Davis Comprehensive Cancer Center, drew on pooled data from the 2022 and 2023 cycles of the Online California Adult Tobacco Survey. That survey is designed to be representative of California adults, and the researchers restricted their analysis to people between the ages of 18 and 64, yielding a weighted analytic sample of 10,332 respondents. Because the data were weighted, the findings are intended to reflect the broader adult population of the state rather than just the people who happened to answer the questionnaire.

What sets this investigation apart is its measurement strategy. Rather than simply recording whether respondents belonged to a racial or ethnic minority group or identified as sexual or gender minorities, the researchers measured discrimination directly using the Everyday Discrimination Scale, a widely used instrument that asks about routine experiences of being treated with less courtesy or respect, being insulted, or being perceived as less capable. Respondents were then classified into categories based on the type of discrimination they attributed their experiences to: none at all, discrimination related to something other than race or sexual orientation, racial or ethnic discrimination only, sexual orientation discrimination only, or the joint experience of both racial-ethnic and sexual orientation discrimination.

The researchers also captured a dimension that is often overlooked in health disparities research: frequency. Experiencing discrimination once is not the same as living with it as a chronic, repeating feature of daily life. By standardizing discrimination scores and modeling both the type and the reported frequency of discriminatory experiences, the team could ask whether the chronicity of mistreatment mattered independently of its kind. The answer, according to their statistical models, was yes, and the patterns were strikingly different across categories.

The core results come from logistic regression models that adjusted for a range of demographic and socioeconomic factors, allowing the researchers to isolate the association between discrimination and tobacco use from confounders such as age, income, education, and other characteristics. Compared with adults who reported no discrimination, those who experienced discrimination tied to other reasons had 32 percent higher odds of current cigarette use, with an adjusted odds ratio of 1.32 and a 95 percent confidence interval of 1.11 to 1.57. Those who experienced sexual orientation discrimination alone faced roughly double the odds of smoking, with an adjusted odds ratio of 2.05 and a confidence interval of 1.47 to 2.85, the single largest effect observed in the study.

Perhaps the most counterintuitive finding involved people who reported experiencing both racial-ethnic and sexual orientation discrimination. Their adjusted odds of smoking were elevated, at 1.26 with a confidence interval of 1.10 to 1.45, but the effect was smaller than for sexual orientation discrimination alone. The same held true for vaping: joint discrimination was associated with 31 percent higher odds of current vaping, while racial-ethnic discrimination alone carried 53 percent higher odds and other discrimination 37 percent higher odds. The researchers interpret this not as evidence that intersectional discrimination is somehow protective, but as a reminder that the relationship between layered prejudice and health behavior is not a simple additive one, where each additional form of discrimination mechanically stacks risk on top of risk.

Across both outcomes, the direction of association was consistent: more discrimination, and more frequent discrimination, went together with higher odds of tobacco product use. For vaping, the adjusted odds ratios were 1.37 for other discrimination, 1.53 for racial-ethnic discrimination only, and 1.31 for joint discrimination, all with confidence intervals that excluded the null value of 1.0. The consistency of the positive association across two distinct tobacco products, one a combustible product with a century of epidemiological baggage and the other a relatively recent phenomenon, suggests that the underlying mechanism is not specific to any single delivery device but may relate to how people cope with chronic social stress.

The biological plausibility for such a link comes from a substantial body of research on minority stress and allostatic load. Chronic exposure to discriminatory treatment is thought to activate the body’s stress response systems repeatedly, elevating cortisol and sympathetic nervous system activity over time. Tobacco use, in this framework, can function as both a pharmacological coping mechanism, delivering nicotine that temporarily blunts negative affect, and a behavior embedded in social environments where stress and marketing pressure converge. The California findings cannot prove causation, since the data are cross-sectional and discrimination and tobacco use were measured at the same time, but they add to a growing pattern of evidence that perceived discrimination behaves like a modifiable risk factor for one of the world’s leading causes of preventable death.

The methodological implications may be as important as the substantive ones. Public health surveillance has long relied on demographic categories, race, ethnicity, sexual orientation, gender identity, to identify populations at elevated risk. Those categories are useful but blunt, and they describe characteristics that are immutable. By measuring the actual experience of discrimination, including its type and its frequency, the study demonstrates a way to identify risk that tracks something potentially changeable. If discrimination contributes to tobacco use, then reducing discrimination, through policy, enforcement, or community intervention, becomes a legitimate tobacco control strategy, not merely a matter of social justice pursued for its own sake.

The authors are careful about the limits of their work. The study is observational, the discrimination measures rely on self-report, and the attribution of discriminatory treatment to a specific cause, race versus sexual orientation versus something else, is inherently subjective. The analysis was also restricted to California adults of working age, and patterns could differ elsewhere. Still, the study, which was funded in part by the California Department of Public Health and reviewed as exempt by the University of Hawaiʻi at Mānoa Institutional Review Board, offers a template for future research: model discrimination as an exposure with both a type and a chronicity, rather than as a proxy embedded in demographic labels. If replicated, the finding that sexual orientation discrimination alone roughly doubles the odds of cigarette smoking would rank among the strongest social-environmental correlates of tobacco use ever quantified at the population level, and it would give tobacco control programs a concrete, measurable target in their efforts to close the gaps in health that demographics alone cannot explain.

Subject of Research: The association between racial/ethnic and sexual orientation discrimination and tobacco product use among California adults

Article Title: Tobacco product use at the intersection of racial/ethnic and sexual orientation discrimination among California adults, 2022–2023: an observational study

Article References: Stupplebeen, D. A., Donaldson, C. D., & Vuong, T. D. (2026). Tobacco product use at the intersection of racial/ethnic and sexual orientation discrimination among California adults, 2022–2023: an observational study. International Journal for Equity in Health. https://doi.org/10.1186/s12939-026-03054-z

Image Credits: AI Generated

DOI: 10.1186/s12939-026-03054-z

Keywords: tobacco use, discrimination, health disparities, sexual orientation, race and ethnicity, cigarette smoking, vaping, minority stress, California, epidemiology, public health, equity