Sexual Orientation Disparities in Adolescent Cigarette Smoking: Intersections With Race/Ethnicity,Gender, and Age

Objectives. We examined sexual orientation differences in adolescent smoking and intersections with race/ethnicity, gender, and age.Methods. We pooled Youth Risk Behavior Survey data collected in 2005 and 2007 from 14 jurisdictions; the analytic sample comprised observations from 13 of those jurisdictions (n = 64 397). We compared smoking behaviors of sexual minorities and heterosexuals on 2 dimensions of sexual orientation: identity (heterosexual, gay–lesbian, bisexual, unsure) and gender of lifetime sexual partners (only opposite sex, only same sex, or both sexes). Multivariable regressions examined whether race/ethnicity, gender, and age modified sexual orientation differences in smoking.Results. Sexual minorities smoked more than heterosexuals. Disparities varied by sexual orientation dimension: they were larger when we compared adolescents by identity rather than gender of sexual partners. In some instances race/ethnicity, gender, and age modified smoking disparities: Black lesbians–gays, Asian American and Pacific Islander lesbians–gays and bisexuals, younger bisexuals, and bisexual girls had greater risk.Conclusions. Sexual orientation, race/ethnicity, gender, and age should be considered in research and practice to better understand and reduce disparities in adolescent smoking.Cigarette smoking continues to be the leading cause of preventable morbidity and premature mortality in the United States.1,2 Preventing adolescent smoking is essential to reducing the burden of cigarettes because smoking typically begins during adolescence.3,4 Approximately 88% of adult daily smokers began smoking before their 18th birthday.5 Research has shown that adolescents with a minority sexual orientation (i.e., lesbian, gay, and bisexual LGB] youths and other adolescents who report same-sex attractions or behavior) are more likely than heterosexual adolescents to smoke cigarettes.6–12 In addition to variation in adolescent smoking by sexual orientation, research has documented variation by race/ethnicity, gender, and age–developmental period.13–17 For instance, national data from the United States collected in 2009 found that White (19.4%) and Hispanic (19.1%) high school students reported higher prevalence of current smoking than Asian (9.7%) and Black (9.1%) students.18 Risk for smoking is typically higher in male than female adolescents and in older than younger adolescents.16,19Although research has shown how sexual orientation, race/ethnicity, gender, and age separately influence variations in adolescent smoking, limited data exist on how sexual orientation differences in adolescent smoking vary across sociodemographic factors such as race/ethnicity, gender, and age. A report published in 2011 by the Institute of Medicine, The Health of Lesbian, Gay, Bisexual, and Transgender People: Building a Foundation for Better Understanding, argued for the importance of examining the health of sexual minorities in the context of sociodemographic diversity to provide a more complete understanding of health disparities.20 Empirical evidence of this nature can improve understanding of the burden of smoking in specific population subgroups and identify high-risk subgroups to target for research, prevention, and cessation efforts.Existing research to understand how smoking patterns of sexual minority youths vary across gender, age, and race/ethnicity is inconclusive and sometimes contradictory. In addition, few studies have used large, representative samples, which limits the ability to draw inferences about the entire population of sexual minority youths.21 Studies examining how sexual orientation differences in adolescent smoking vary by gender have been the most conclusive and have typically found larger disparities between sexual minority and heterosexual adolescent girls than between sexual minority and heterosexual adolescent boys.6,9,22,23 However, studies examining how sexual orientation differences in adolescent smoking vary by age have been inconclusive. One study of mostly White youths followed between ages 12 and 24 years found that smoking disparities were larger between sexual minorities and heterosexuals during younger than older ages.6 However, a study of Asian Americans and Pacific Islanders (APIs) found that smoking disparities were not present in adolescence but emerged in young adulthood.9In addition, scant data exist on how sexual orientation and race/ethnicity jointly influence risk for adolescent smoking. This is an especially difficult area to investigate because studies with a sample size large enough to examine this question are rare. Some evidence suggests that sexual minority youths who belong to racial/ethnic minority groups are more likely to smoke cigarettes than their heterosexual peers of their same race/ethnicity. For instance, a study of college students found that Black, Asian, Hispanic, and multiracial LGB persons were more likely to smoke than their heterosexual racial/ethnic peers.24 This study also found that Black and Asian LGB persons were less likely to smoke than their White LGB peers, but the same was not true for Hispanics and multiracial LGB persons. However, the study did not describe statistical testing to examine whether race/ethnicity modified sexual orientation disparities in smoking.Another important consideration is the multidimensional nature of sexual orientation (e.g., identity, attractions, behaviors), which in research with adolescents has most often been assessed as how respondents identify or the gender of their sexual attractions or partners. How sexual orientation is operationalized in studies may influence findings and conclusions, but studies infrequently include more than 1 dimension. Studies with adults12,25–27 and adolescents10,23,28 have shown differences in the magnitude of the sexual orientation disparities observed depending on which dimension is considered. For example, a study of Mexican youths aged 18 to 29 years found that self-identified LGB participants had approximately twice the odds of reporting current smoking than did heterosexuals, but differences between participants reporting only same-sex partners and those reporting only opposite-sex partners were negligible.23 Such disparate findings are likely to occur because the dimensions capture somewhat different populations with differing risk and protective factors.29It is especially important to assess multiple dimensions of sexual orientation in adolescence because a same-sex orientation commonly develops during this period, and many adolescents with a same-sex orientation may not identify as LGB.30 In addition, when gender of sexual partners is used as an indicator of sexual orientation, only adolescents who have initiated sexual intercourse (approximately 48% of high school students in 200719) can be identified. Because adolescent smoking is a robust correlate of sexual activity,31 the degree to which the selection of a sexually active subgroup may influence sexual orientation findings warrants consideration. Finally, the extent to which the different dimensions may affect conclusions drawn about smoking disparities arising from sexual orientation when also considering intersections with race/ethnicity, gender, and age remain uncertain. To address these questions, we compared sexual orientation differences in smoking during adolescence with 2 dimensions of sexual orientation (identity and gender of lifetime sexual partners) and investigated how these differences were modified by race/ethnicity, gender, and age in Youth Risk Behavior Survey (YRBS) data pooled from 13 jurisdictions and 2 years.