Do inequalities add up? Intersectional inequalities in smoking by sexual orientation and education among U.S. adults

Abstract

Inequalities in smoking by socio-economic status (SES) are well-known. A growing body of literature has demonstrated additional inequalities in smoking by sexual orientation. This study used an intersectional lens to examine smoking at the intersection of sexual orientation and education. Data come from 28,362 adult participants in Wave 2 (2014-2015) of the Population Assessment of Tobacco and Health (PATH) Study. We used educational level (less than high school education (HS); HS or more) and sexual orientation (heterosexual; sexual minority) to form four intersectional positions.We estimated prevalence differences in smoking corresponding to joint, referent, and excess intersectional inequalities using weighted linear binomial regression models. Results were stratified by gender and adjusted for ethnicity and age. The adjusted joint inequality represented 7.6% points (p.p.) (95% CI: 2.5, 12.8) difference in smoking between the doubly advantaged (heterosexual with HS or more) and doubly disadvantaged (sexual minority with less than HS) positions. Joint inequality was decomposed into referent SES inequality (12.5 p.p. (95% CI: 10.5, 14.4)); referent sexual orientation inequality (9.7 p.p. (95% CI: 6.8, 12.6)); and a substantial negative excess intersectional inequality (-14.6 p.p. (95% CI: -20.8, -8.3)), attributed to an unexpectedly low prevalence of smoking among doubly disadvantaged persons. Similar overall patterns were found in the stratified analyses. We found that "doubly-disadvantaged" group of low-educated sexual minority adults did not have the greatest burden of smoking; whereas, low-educated heterosexual adults had the highest smoking prevalence. Our findings support tailoring cessation interventions to disadvantaged groups' different needs.