For the purposes of this chapter, studies that included those younger than age 18 years (but could include young adults) or students at high school level or lower are cited. The disparities in smoking prevalence observed in the adult LGBT population likely originate during adolescence and emerging adulthood. Early on, Garofalo et al. [27] observed higher rates of smoking among Massachusetts youth in grades 9–12, who identified as gay, lesbian, or bisexual in the state’s Youth Risk Behavior Surveillance Survey (YRBSS) [13], a CDC-designed, state-based survey of health risk behaviors of youth in the U.S. They found that 59.3 % of LGB youth were current smokers compared to 35.2 % of the non-LGB youth. This tobacco use disparity in Massachusetts was confirmed subsequently by YRBSS findings in Colorado [10].

A solid and growing body of published research using national, population-based sampling indicates that the prevalence of smoking among lesbians, gay men, and bisexual men and women is significantly elevated when compared to smoking prevalence for their heterosexual counterparts. Smoking prevalence data for transgender persons have not been collected [42] in most population-based surveys, and the existing studies that do use varying methods to sample the transgender subpopulation. A systematic review of the tobacco research literature conducted by Lee et al. [42], and spanning nearly two decades up to mid-2007, found that sexual minority men and women have 1.5 to 2.5 times the odds of being current smokers when compared to their heterosexual counterparts. More recent tobacco use data from 10 states participating in the CDC Behavioral Risk Factor Surveillance Survey (BRFSS) [4] and assessing sexual orientation and gender identity in 2010, are consistent with that review. Specifically, lesbians had nearly twice the odds of being a current smoker compared to heterosexual women (OR = 1.91, 95 % CI = 1.26, 2.91). Bisexual women had over twice the odds of being a smoker (OR = 2.13, 95 % CI = 1.33, 3.42). In terms of smoking prevalence, 19.1 % of lesbians were current smokers, whereas 29.7 % of bisexual women and 11.7 % of heterosexual women smoked. Both gay men and bisexual men had nearly twice the odds of being a current smoker compared to heterosexual men (respectively, OR = 1.93, 95 % CI = 1.27, 2.93; OR = 1.92, 95 % CI = 1.04, 3.53) [4]. The prevalence of current smoking was 22.9 % for gay men, 33.3 % for bisexual men, and 15.8 % for heterosexual men. As the authors noted, the limitation of these data include the absence of any southern state among the 10 states reporting BRFSS data; therefore, they are not representative of the U.S. population, and assessment of sexual orientation varied across states.

LGB adolescents compared to heterosexual youth are more likely to start smoking at an earlier age and report current smoking [19, 20]. In general, adolescents who smoke compared to those who don’t are at greater risk to become regular smokers in adulthood [17]. There is variability in risks for smoking within sexual minority subgroups. A meta-analysis of 18 studies examining sexual orientation and adolescent substance use found that being female (vs. male) or bisexual (vs. gay or lesbian) were associated with higher risk for smoking [49]. Bisexual identity, compared to lesbian/gay or heterosexual identity, as a risk factor for higher risk for tobacco use and other risk behaviors [49] was confirmed in later YRBSS data [11]. Victimization, such as verbal or physical harassment due to being LGB, is longitudinally associated with LGB smoking status [52], and LGBQ (queer-identified youths were included) youths experiencing high victimization levels also had higher smoking levels and other health risk behavior compared with heterosexual youths reporting low levels of victimization [5]. In the latter study, LGBQ youth who reported low levels of victimization had health risk behavior profiles similar to heterosexual youth, and LGBQ youth reporting the highest victimization were at higher risk than heterosexual youth also reporting high victimization [5]. Approaching the influence of social environment on smoking risk from an alternative perspective, Hatzenbuehler et al. [35] found that among 11th graders in Oregon, a greater supportive social environment for LGB youth (e.g., presence of gay-straight alliances, school nondiscrimination and anti-bullying policies that specifically protected LGB students) was associated with reduced tobacco use. The potential mitigating effect of supportive school environments on LGBT youths’ risk for tobacco use, along with the link between victimization and risk for smoking, strongly suggest that social stigma, harassment, and absence of clear support for LGBT identity are potent risk factors for smoking in these youth. Parallel to these data are consistent findings across studies that use of other substances and levels of other high risk health behaviors are elevated among LGB youth [11, 44]. Interviews with LGBT youths about smoking have highlighted the perceived utility of smoking in managing stressors and stress reactions [60], although other data suggest that smoking can amplify the association between stress burden and subsequent psychological distress [62].

A 2011 systematic review of publications reporting sociodemographic, behavioral, and mental health associations with smoking among sexual minorities noted a fragmented research literature lacking coherence in sampling methods, measures, and definitions of sexual orientation and smoking. The reviewers stated “the current evidence base constructs an incomplete and challenging glimpse into the etiology of smoking disparities among sexual minorities,” ([3], p. 4). Although some progress has been made since, their statement largely holds. That review identified alcohol use, depression or depressive symptoms, younger age, and lower education level as associated with LGB tobacco use, all of which are also associated with current smoking in the general population [3, 12, 30, 56]. Similarly, the Legacy Young Adults Cohort study [59] conducted multivariate logistic regressions to determine covariates of tobacco use by sexual identity. Current cigarette use by LGB young adults was significantly associated with having a high school education (vs. some college or greater), OR = 4.27 (95 % CI: 1.51–12.12); current alcohol use, OR = 12.65 (95 % CI: 2.99–53.54); and current other drug use, OR = 9.67 (95 % CI: 2.22–42.09). The same three covariates were significant in the statistical model using any current tobacco use as the outcome variable. Notably, when the same statistical models were developed for heterosexuals, two of the three variables significant for LGB youth, i.e., high school education and current alcohol use, were significant for them as well. As the authors noted, current alcohol use was most strongly associated with current cigarette and any tobacco use by LGB young adults.