Sexual minority tobacco use disparities across adolescence and the transition to young adulthood

Abstract

Introduction:

Sexual minority (SM; e.g., lesbian, gay, bisexual) youth are disproportionately more likely to use tobacco than non-SM youth, yet there exist several critical gaps in knowledge. This study assessed (a) the timing of SM tobacco use disparities (e.g., during adolescence or early adulthood), (b) whether disparities generalize across different tobacco products, and (c) whether disparities differ by sex.

Methods:

Data were from a 6-year prospective cohort of diverse high school students from Southern California who were followed into early adulthood (9 waves, 2013-2019). SM (vs. non-SM) differences in past 6-month use were assessed for: any tobacco products, cigarettes, e-cigarettes, other products (e.g., hookah), and multiple products. Disparities were modeled longitudinally across adolescence (high school) and the transition to early adulthood (end of high school to post-high school). Differences were tested by sex.

Results:

Among females, SM disparities were evident for all outcomes during both adolescence and early adulthood; no differences were observed among males. For example, SM (vs. non-SM) females had higher odds of cigarette (aOR=4.4 [3.0-6.5]) and e-cigarette (aOR=1.7 [1.2-2.4]) use, averaged across adolescence. The timing of disparities varied by product. For example, cigarette use disparities emerged prior to high school and persisted through adolescence and young adulthood, while e-cigarette use disparities were present in early adolescence and young adulthood only.

Conclusions:

Young SM females are at especially high risk for tobacco use, across various tobacco products, throughout adolescence and young adulthood. Interventions must consider differences in the timing of disparities by product type.

INTRODUCTION

There exist considerable disparities in tobacco product use between sexual minority (SM; e.g., lesbian, gay, bisexual, queer) and non-SM populations, with SM individuals being disproportionately more likely than non-SM individuals to report use of, and dependence on, tobacco products (American Lung Association, 2017; Lee et al., 2009). For instance, in a recent U.S. national sample of adults, approximately 35% of both lesbian females and gay males reported past-year cigarette smoking, compared to 20% and 26% of heterosexual females and males, respectively (McCabe et al., 2018). These disparities are long-standing and they are driven by a number of factors, including a long history of targeting marketing to SM populations by tobacco companies (Dilley et al., 2008; Smith et al., 2008), the historical significance of bars and other places where tobacco use is common as safe places for SM people to socialize (Blosnich et al., 2013), and use of tobacco to cope with social stressors such as discrimination (Krueger et al., 2020). However, several critical gaps in knowledge remain, limiting the degree to which SM tobacco use disparities can effectively be reduced.

The bulk of SM tobacco use research has focused on documenting disparities in cigarette use (Azagba et al., 2014; CDC, 2018b; Corliss et al., 2014; Hoffman et al., 2018; Watson et al., 2018), yet use of cigarettes has declined considerably over the past two decades in the United States (Johnston et al., 2019). While cigarette smoking remains an important public health concern, the landscape of commercially available tobacco products has changed markedly in recent years with the introduction and rapid increase in availability and use of e-cigarettes (e.g., JUUL) (CDC, 2020). Emerging evidence suggests SM populations use e-cigarettes at higher rates than non-SM populations (Dai, 2017; Hoffman et al., 2018; Wheldon et al., 2018). Further, poly-tobacco use refers to the use of multiple tobacco products (e.g., both cigarettes and e-cigarettes). Given higher rates of both cigarette and e-cigarette use among SM populations, it is possible SMs also use multiple products at higher rates than non-SMs. To our knowledge, no prior studies have assessed this question, though prior research has shown differences in poly-tobacco use among SM subgroups, with bisexual female adults reporting higher odds of poly-tobacco use than gay male adults (Delahanty et al., 2019). Thus, while tobacco use disparities are well-documented, they are not well-characterized. In order to adequately address SM tobacco use disparities, data on the specific products used by SM individuals, and patterns of use among SM and non-SM individuals, are needed.

Further, SM tobacco use disparities often emerge during adolescence (Marshal et al., 2008; Marshal et al., 2009), and earlier tobacco use initiation has been linked to higher risk for nicotine dependence in adulthood (Schuler and Collins, 2019; U.S. Department of Health and Human Services, 2014). It is therefore vital to understand when disparities in SM and non-SM tobacco products first emerge; such information will help practitioners and policymakers to identify effective solutions for reducing tobacco product initiation among a high-risk population. However, the majority of related research – particularly with regard to use non-cigarette tobacco products – has been conducted using cross-sectional samples. As such, critical gaps remain related to when (e.g., early adolescence vs. later adolescence) these disparities emerge, for which products (e.g., cigarettes vs. e-cigarettes), and whether they persist, dissipate, or widen as youth transition to adulthood (Schuler and Collins, 2019).¹ To effectively address SM disparities in tobacco use across the life course, it is important to identify critical periods for the onset of disparities among youth, for a variety of products.

Finally, there exist marked sex differences in tobacco product use among youth, with males generally using tobacco at higher rates than females among youth in the general population (Harrell et al., 1998; World Health Organization, 2007) (though it should be noted that tobacco use among young females is of increasing public health concern (CDC, 2018a)). Emerging research has shown sex differences in tobacco use within the SM population as well; however, this trend is often “flipped,” with SM females commonly using tobacco at higher rates than SM males (Austin et al., 2004; Dai, 2017). This sex difference is not well-understood, but it is important to compare females to males in terms of tobacco product use – both within SM populations and across SM and non-SM populations. Failure to do so may result in disparities being underestimated and ineffective policies being implemented that do not address those most at risk.

Current study

To address these research gaps, the current study uses data from a 6-year prospective cohort of youth and young adults in order to: (1) compare SM (versus non-SM) differences in past 6-month use of: any tobacco products, cigarettes, e-cigarettes, other products (e.g., cigars, hookah), and multiple products; (2) compare differences in the timing of the emergence of disparities across each of the outcomes; and (3) assess whether SM tobacco use disparities vary by sex.

METHODS

Study Design

Data were from a prospective cohort study of substance use among high school students in Southern California. The cohort was recruited during fall semester, 2013 from 10 high schools in Los Angeles County when students were in 9^th grade (first year of high school, mean age: 14.1; N=3,396) and were subsequently followed bi-annually (each semester, approximately every 6 months) through the completion of high school in 2017 (waves 1-8). At each wave, surveys were administered in class via paper and pencil. Students who were not present in class instead completed surveys by telephone or Internet. The cohort had a total retention rate of 92.8% from wave 1 to wave 8. Respondents were surveyed again in January-September 2019 after graduation from high school via Internet (wave 9, mean age: 19.8; N=2,548; 81.3% retention from wave 8 to wave 9). Respondents who completed a wave 9 survey, and who had complete demographic information were eligible for the current study. The final analytic sample was 2,323 (see consort diagram, Figure S1).

Ethics Statement

The study was approved by the institutional review board at the University of Southern California. Parents provided written or verbal informed consent, and students provided assent prior to data collection.

Study Variables

Tobacco product use. Five outcomes were assessed at each wave. Respondents were asked if they had used any of the following in the past 6 months (yes/no): any tobacco products, cigarettes, e-cigarettes, and all other tobacco products (inclusive of hookah, cigars, and cigarillos). Additionally, a three-level outcome assessed poly-tobacco product use at each wave (use of 0, 1, or 2+ different products).

Sexual minority status was assessed at the final wave (wave 9) using three survey items, each capturing a different component of sexual orientation: identity, attraction, and behavior. For identity, respondents were asked “do you consider yourself to be:” (single choice: asexual, bisexual, gay, straight, lesbian, pansexual, queer, questioning or unsure, another identity not listed here, prefer not to disclose). Those selecting “asexual” or “prefer not to disclose” were marked as missing for sexual identity. For sexual attraction, respondents were asked, “people are different in their sexual attraction to other people. Which best describes your feelings?” (Likert scale from only attracted to members of the same sex – only attracted to members of the opposite sex, attracted to all genders, not attracted to anyone, not sure, prefer not to disclose). Those selecting “not attracted to anyone,” “not sure,” or “prefer not to disclose” were marked as missing for attraction. For sexual behavior, respondents were asked “(“in the past year, who did you have sex with?” (single choice: males, females, transgender males, transgender females, gender variant/non-binary, additional gender category, have not had sex, prefer not to disclose). Those selecting “have not had sex” or “prefer not to disclose” were marked as missing for sexual behavior. For the present analyses, respondents were coded as SM if they reported a non-straight sexual identity (i.e., bisexual, gay, lesbian, pansexual, queer, questioning or unsure), reported being attracted to members of the same sex or all genders, and/or reported having engaged sexual activity with members or the same sex, transgender, or non-binary partners. Respondents were coded as non-SM if they identified as straight, reported exclusively opposite-sex attractions, and reported no same-sex behavior.

Covariates. Several sociodemographic characteristics, assessed at baseline (wave 1), were included as covariates. Sex was assessed with the question, “what is your sex?” (female, male). Age was assessed in years (continuous). Race/ethnicity was assessed using the question “please choose one term that best describes you,” (American Indian or Alaska Native, Asian, Black or African American, Hispanic or Latino, Native Hawaiian or Pacific Islander, White, Multiracial, Other), which was collapsed into a 6-level variable (Hispanic, Asian, White, Black, Other, Multiracial). Parental education was also assessed for each parent (“what is the highest grade completed by your [mother/father]?”), and a 4-level variable (less than high school, high school, college, advanced degree) was created for the parent with the highest level of education.

Data Analysis

Sociodemographic characteristics were first calculated, separately by sex. Then, we estimated the prevalence of each tobacco use outcome, averaged across adolescence (waves 1-8) and the transition to young adulthood (waves 8-9), separately by time period and SM status. Finally, we used a series of mixed effect logistic regression models to estimate the longitudinal differences between SM and non-SM respondents for each outcome, adjusting for age, race/ethnicity, parental education and random effects by wave (controlling for multiple observations per respondent) and school (controlling for nesting of respondents by school). All models were sex-stratified and were estimated separately across the two study periods.

To assess whether relative differences in tobacco product use varied on the basis of sex, each adjusted model was repeated in a pooled sample including both males and females and included a multiplicative interaction term (SM status*sex); significance of the interaction was used to determine whether estimates differed by sex. Similar models were used to evaluate whether estimates differed by wave, by including a SM*wave interaction term. All analyses were conducted using Stata SE version 15. For all analyses, Benjamini-Hochberg-adjusted p values were calculated to adjust for multiple comparisons.

RESULTS

Sociodemographic characteristics

For detailed sociodemographic comparisons, refer to Supplemental Table S1. Compared to males (11.2%), a larger proportion of female respondents were sexual minorities (21.1%; p<0.001). There were also small, but significant age differences, with males (mean: 14.1, SD: 0.01) being older than females (mean: 14.0, SD: 0.1; p=0.005). There were no differences by race/ethnicity (p=0.108) or parental education (p=0.526).

SM Disparities in Tobacco Product Use Across Adolescence

Table 1 presents differences in tobacco product use by SM status across adolescence, separately by sex. Among females, SM respondents had higher prevalence rates of each tobacco use outcome across waves 1-8 (difference range: 3.4-8.6%). Unadjusted and adjusted mixed effects logistic regression models also estimated the association between SM status and each tobacco use outcome, across adolescence. In adjusted models, SM females had higher odds of using each tobacco product, compared to non-SM females averaged across the entire period (any: aOR=2.8 [1.9, 3.9]; cigarettes: aOR=4.4 [3.0, 6.5]; e-cigarettes: aOR=1.7 [1.2, 2.4]; other: aOR=1.8 [1.3, 2.6]; poly [1 vs. 0 products]: aOR=1.9 [1.3, 2.7]; poly [2+ vs. 0 products]: aOR=2.8 [2.0, 4.1]).

Table 1.

Associations between Sexual Minority Status and Tobacco Product Use During Adolescence (W1-W8)

	Females				Males				Females vs. Males
	Absolute Prevalence	Difference in Prevalence	Unadjusted Model	Adjusted Model	Absolute Prevalence	Difference in Prevalence	Unadjusted Model	Adjusted Model
	% (95% CI)	% (95% CI)	OR (95% CI)	OR (95% CI)	% (95% CI)	% (95% CI)	OR (95% CI)	OR (95% CI)	P-Value
Any Tobacco Use									0.014
Non-Sexual Minority	11.9 (11.2, 12.6)		Ref	Ref	14.1 (13.3, 14.9)		Ref	Ref
Sexual Minority	20.5 (18.8, 22.2)	8.6 (6.7, 10.4)	2.6 (1.8, 3.7)	2.8 (1.9, 3.9)	12.9 (10.7, 15.6)	−1.2 (−3.7, 1.4)	0.9 (0.5, 1.6)	0.9 (0.5, 1.6)
Cigarettes									0.011
Non-Sexual Minority	4.8 (4.4, 5.3)		Ref	Ref	7.5 (6.9, 8.2)		Ref	Ref
Sexual Minority	12.7 (11.3, 14.2)	7.9 (6.4, 9.4)	4.4 (3.0, 6.4)	4.4 (3.0, 6.5)	9.1 (7.2, 11.4)	1.6 (−0.6, 0.4)	1.3 (0.7, 2.5)	1.3 (0.7, 2.6)
E-Cigarettes									0.026
Non-Sexual Minority	12.6 (11.9, 13.3)		Ref	Ref	15.9 (15.1, 16.8)		Ref	Ref
Sexual Minority	17.6 (16.0, 19.3)	5.0 (3.2, 6.8)	1.7 (1.2, 2.3)	1.7 (1.2, 2.4)	14.0 (11.6, 16.7)	−2.0 (−4.6, 0.7)	0.7 (0.4, 1.3)	0.7 (0.4, 1.2)
Other Tobacco Products									0.082
Non-Sexual Minority	9.4 (8.8, 10.0)		Ref	Ref	10.4 (9.7, 11.2)		Ref	Ref
Sexual Minority	13.3 (11.9, 14.8)	3.9 (2.3, 5.4)	1.7 (1.2, 2.5)	1.8 (1.3, 2.6)	8.7 (6.8, 10.9)	−1.8 (−3.9, 0.4)	0.7 (0.4, 1.3)	0.7 (0.4, 1.4)
Poly Tobacco Use									0.020
Number of Products (ref = 0)
1 Product
Non-Sexual Minority	11.9 (11.3, 12.6)		Ref	Ref	12.6 (11.8, 13.4)		Ref	Ref
Sexual Minority	15.3 (13.8, 16.9)	3.4 (1.7, 5.1)	1.8 (1.3, 2.6)	1.9 (1.4, 2.7)	11.4 (9.3, 13.9)	−1.2 (−3.6, 1.2)	0.8 (0.4, 1.5)	0.8 (0.4, 1.4)
2+ Products
Non-Sexual Minority	6.7 (6.2, 7.2)		Ref	Ref	9.2 (8.6, 9.9)		Ref	Ref
Sexual Minority	12.4 (11.1, 13.9)	5.7 (4.2, 7.2)	2.7 (1.9, 3.8)	2.8 (2.0, 4.1)	8.4 (6.6, 10.6)	−0.8 (−3.0, 1.3)	0.8 (0.4, 1.5)	0.8 (0.4, 1.5)

Note. Adjusted models controlled for wave, school, race/ethnicity, age, parental education. Difference between females and males were tested between adjusted models using likelihood ratio tests. Though there were significant direct effects of wave on all outcomes, there were no significant sexual minority*wave interactions, for either females or males. Boldface indicates statistical significance. All significance tests were corrected for multiple comparisons using the Benjamini-Hochberg correction method

Among males, there were no significant differences in the prevalence of tobacco product use for any of the outcomes. Similarly, no significant differences were found between SM and non-SM males, in either unadjusted or adjusted models, for any of the outcomes (any: aOR=0.9 [0.5, 1.6]; cigarettes: aOR=1.3 [0.7, 2.6]; e-cigarettes: aOR=0.7 [0.4, 1.2]; other: aOR=0.7 [0.4, 1.2]; poly [1 vs. 0 products]: aOR=0.8 [0.4, 1.4]; poly [2+ vs. 0 products]: aOR=0.8 [0.4, 1.5]). For all outcomes except other tobacco products, the adjusted models estimating the associations between SM status and tobacco use differed significantly between females and males (p<0.008 for all likelihood ratio tests). However, for all models the sexual minority status*wave interaction was not significant, signifying that the magnitudes of the disparities did not differ significantly over time, relative to baseline (p>0.05 for all outcomes, data not shown).

Figure 1 shows the predicted probability of each tobacco use outcome across adolescence (waves 1-8), separately by SM status and sex, controlling for covariates. For any tobacco product use except wave 6, SM females reported higher use than non-SM females at each wave of the study (p<0.05 for all pairwise comparisons). Similarly, SM females reported using cigarettes at higher rates than non-SM females across all study waves (all p<0.05). SM females and non-SM females used e-cigarettes at similar rates at wave 1, but a disparity emerged at waves 2 (1st year of high school, 2^nd semester), and persisted through wave 4 (2^nd year of high school, second semester; all p<0.05). However, this disparity lost significance across waves 5-8, (all p>0.05). There were also disparities for other tobacco product use between SM and non-SM females early in high school (waves 1 and 3), and for use of multiple products both in early and later high school (waves 1-3, 7-8; all p<0.05).

Figure 1. SM Disparities in Tobacco Product Use Across Adolescence, Separately by Sex.

Figure displays the predicted probability of each tobacco use outcome for SM and non-SM respondents across waves 1-8, separately by sex, and controlling for covariates. For all models except the poly-tobacco use model, asterisks signify a significant difference (p<0.05) between SM and non-SM respondents at each wave. For poly-product use, asterisks signify a significant difference (p<0.05) in use of 1 (versus 0) products between SM and non-SM respondents; crosses (†) signify a significant difference (p<0.05) in use of 2+ (versus 0) products between SM and non-SM respondents. All significance tests were corrected for multiple comparisons using the Benjamini-Hochberg correction method.

For all models except poly-tobacco use:

For poly-tobacco use model:

SM Disparities in Tobacco Product Use Across the Transition to Young Adulthood

Table 2 presents differences in tobacco product use by SM status across the transition to young adulthood, separately by sex. Among females, SM females had higher prevalence rates of any tobacco use (difference: 9.1%), cigarette use (difference: 10.0%), e-cigarette use (difference: 9.2%), and use of two or more tobacco products (difference: 9.9%). Unadjusted and adjusted mixed effects logistic regression models also estimated the association between SM status and use of each tobacco outcome, across the transition to young adulthood. In adjusted models, SM females had higher odds of most outcomes, compared to non-SM females (any: aOR=2.4 [1.6, 3.5]; cigarettes: aOR=4.7 [2.7, 8.0]; e-cigarettes: aOR=2.4 [1.6, 3.5]; other: aOR=1.7 [1.1, 2.6]; poly [2+ vs. 0 products]: aOR=3.9 [2.5, 6.0]).

Table 2.

Associations between Sexual Minority Status and Tobacco Product Use During Transition to Adulthood (W8-W9)

	Females				Males				Females vs. Males
	Absolute Prevalence	Difference in Prevalence	Unadjusted Model	Adjusted Model	Absolute Prevalence	Difference in Prevalence	Unadjusted Model	Adjusted Model
	% (95% CI)	% (95% CI)	OR (95% CI)	OR (95% CI)	% (95% CI)	% (95% CI)	OR (95% CI)	OR (95% CI)	P-Value
Any Tobacco Use									0.159
Non-Sexual Minority	22.7 (21.0, 24.5)		Ref	Ref	28.7 (26.6, 30.9)		Ref	Ref
Sexual Minority	31.8 (28.0, 35.9)	9.1 (4.7, 13.4)	1.7 (1.3, 2.2)	2.4 (1.6, 3.5)	28.7 (22.6, 35.6)	−0.0 (−6.9, 6.8)	1.0 (0.6, 1.7)	1.2 (0.6, 2.2)
Cigarettes									0.041
Non-Sexual Minority	6.8 (5.8, 7.9)		Ref	Ref	13.4 (11.8, 15.0)		Ref	Ref
Sexual Minority	16.7 (13.8, 20.2)	10.0 (6.6, 13.3)	4.4 (2.7, 7.4)	4.7 (2.7, 8.0)	15.8 (11.2, 21.8)	2.4 (−3.1, 7.9)	1.3 (0.6, 3.0)	1.4 (0.6, 3.1)
E-Cigarettes									0.040
Non-Sexual Minority	22.0 (20.3, 23.8)		Ref	Ref	26.8 (24.7, 28.9)		Ref	Ref
Sexual Minority	31.2 (27.4, 35.3)	9.2 (4.9, 13.5)	1.9 (1.4, 2.5)	2.4 (1.6, 3.5)	25.0 (19.3, 31.8)	−1.8 (−8.4, 4.8)	0.9 (0.5, 1.5)	1.0 (0.5, 1.8)
Other Tobacco Products									0.503
Non-Sexual Minority	8.9 (7.8, 10.1)		Ref	Ref	12.3 (10.8, 14.0)		Ref	Ref
Sexual Minority	11.7 (9.2, 14.7)	2.8 (−0.2, 5.8)	1.5 (1.0, 2.2)	1.7 (1.1, 2.6)	11.9 (8.0, 17.4)	−0.4 (−5.3, 4.5)	1.0 (0.5, 2.0)	1.3 (0.6, 2.6)
Poly Tobacco Use									0.032
Number of Products (ref = 0)
1 Product
Non-Sexual Minority	18.4 (16.8, 20.0)		Ref	Ref	17.3 (15.6, 19.2)		Ref	Ref
Sexual Minority	17.7 (14.7, 21.2)	−0.6 (−4.3, 2.9)	1.3 (0.9, 1.9)	1.6 (1.0, 2.4)	17.8 (13.0, 24.0)	0.5 (−5.3, 6.3)	1.1 (0.6, 2.0)	1.1 (0.6, 2.3)
2+ Products
Non-Sexual Minority	8.5 (7.4, 9.7)		Ref	Ref	14.9 (13.3, 16.6)		Ref	Ref
Sexual Minority	18.4 (15.4, 22.0)	9.9 (6.4, 13.4)	3.0 (2.1, 4.4)	3.9 (2.5, 6.0)	14.6 (10.2, 20.5)	−0.3 (−5.6, 5.1)	1.0 (0.5, 1.9)	1.2 (0.6, 2.5)

Note. Adjusted models controlled for wave, school, race/ethnicity, age, parental education. Difference between females and males were tested between adjusted models using likelihood ratio tests. Though there were significant direct effects of wave on all outcomes, there were no significant sexual minority*wave interactions, for either females or males. Boldface indicates statistical significance. All significance tests were corrected for multiple comparisons using the Benjamini-Hochberg correction method.

Among males, there were no significant differences in the prevalence of tobacco product use for any of the outcomes. Similarly, no significant differences were found between SM and non-SM males, in either unadjusted or adjusted models, for any of the outcomes (any: aOR=1.2 [0.6, 2.2]; cigarettes: aOR=1.4 [0.6, 3.1]; e-cigarettes: aOR=1.0 [0.5, 1.8]; other: aOR=1.3 [0.6, 2.6]; poly [1 vs. 0 products]: aOR=1.1 [0.6, 2.3]; poly [2+ vs. 0 products]: aOR=1.2 [0.6, 2.5]). For cigarettes, e-cigarettes, and poly-tobacco use, the adjusted models estimating the associations between SM status and tobacco use differed significantly between females and males (all p<0.015). The magnitudes of the disparities at Wave 9 did not differ significantly over time, relative to Wave 8 (p>0.05 for all outcomes, data not shown).

Figure 2 shows the predicted probability of each tobacco use outcome across the transition to young adulthood (waves 8-9), separately by SM status and sex, controlling for covariates. For any tobacco product use and cigarettes, SM females had higher rates of use than non-SM females across both waves (p<0.05 for all pairwise comparisons). SM and non-SM females did not differ in use of other tobacco products for either waves 8 or 9 (p>0.05 for all pairwise comparisons). SM females (vs. non-SM females) had a greater probability of e-cigarette use, and of using both 1 and 2+ tobacco products at wave 9.

Figure 2. SM Disparities in Tobacco Product Use Across the Transition to Young Adulthood, Separately by Sex.

Figure displays the predicted probability of each tobacco use outcome for SM (dashed lines) and non-SM (solid lines) respondents across waves 8-9, separately by sex, and controlling for covariates. For all models except the poly-tobacco use model, asterisks signify a significant difference (p<0.05) between SM and non-SM respondents at each wave. For poly-product use, asterisks signify a significant difference (p<0.05) in use of 1 (versus 0) products between SM and non-SM respondents; crosses (†) signify a significant difference (p<0.05) in use of 2+ (versus 0) products between SM and non-SM respondents. All significance tests were corrected for multiple comparisons using the Benjamini-Hochberg correction method.

For all models except poly-tobacco use:

For poly-tobacco use model:

DISCUSSION

This study found stark disparities in tobacco product use between SM and non-SM adolescents and young adults, across several tobacco use outcomes, inclusive of cigarettes, e-cigarettes, other tobacco products (e.g., hookah), and notably in use of multiple tobacco products. Interestingly, however, the largest differences in prevalence of use were observed for cigarettes. So, while cigarette use has been declining among youth overall (Johnston et al., 2019), SM youth (particularly females) make up a disproportionate fraction of the small population of cigarette smokers. Similar trends have been found for other disparate populations (e.g., those experiencing mental health problems, those of low socioeconomic status) (Drope et al., 2018), which may reflect a greater willingness to engage in “risky” health behaviors to cope with stress (Lehavot and Simoni, 2011; Revell et al., 1985). On the other hand, e-cigarette use was more widespread overall, possibly because e-cigarettes are perceived to be less harmful than cigarettes (Ambrose et al., 2014; Pearson et al., 2012). This might explain, in part, the smaller disparities in use of e-cigarettes observed between SM and non-SM youth.

Importantly, we found SM tobacco use disparities to be concentrated among SM females, consistent with a growing body of research (Austin et al., 2004; Dai, 2017). However, this research provides new evidence that these disparities persisted across a wide range of tobacco use outcomes, and across a 6-year period of follow-up during adolescence and young adulthood. By contrast, the disparity between SM and non-SM males was largely absent. In the general population of youth and young adults, males use tobacco products at higher rates than females, yet among SM youth, this trend is reversed, with females using more frequently than males (Austin et al., 2004; Dai, 2017) – a finding which is not well-understood. Prior research has shown that sensory factors (e.g., environmental cues) more strongly predict tobacco use among women, while pharmacologic factors (e.g., effects of nicotine) more strongly predict use among men (Perkins et al., 2006). Given the interpersonal stressors many SM youth face (e.g., discrimination, bullying) (Meyer, 2003; Russell et al., 2011; Toomey et al., 2013), it is possible SM females are especially drawn to tobacco products for sensory relief/stress management. More research is needed to understand the likely complex mechanisms contributing to sex differences in tobacco product use among SM youth. Regardless, our findings highlight the importance of assessing SM trends in tobacco use separately for females and males.

This study also adds novel information by examining the timing and onset of tobacco use disparities in a contemporary cohort of diverse adolescents and young adults in Los Angeles, CA. Given the rapidly evolving tobacco product marketplace, we were able to assess whether the timing and prevalence of disparities differed across distinct tobacco use products and outcomes. Importantly, we found that disparities for different tobacco use outcomes indeed varied by tobacco product. For some products (e.g., cigarettes and other tobacco products, such as cigars), a disparity between SM and non-SM youth was present across all waves of data collection, suggesting that SM youths’ higher rates of combustible tobacco use began prior to data collection (i.e., before high school). This may reflect concerted industry efforts to market e-cigarettes and other tobacco products to SM populations (Emory et al., 2018; Smith and Malone, 2003) and suggests that tailored prevention activities and policies should be enacted earlier than high school – perhaps in middle schools – particularly because SM youth are increasingly realizing and embracing their sexuality at earlier ages (Grov et al., 2006; Marios et al., 2015).

Conversely, use of e-cigarettes and of multiple products (poly-tobacco use) did not differ significantly between SM and non-SM respondents at baseline. Disparities in these outcomes emerged between SM and non-SM females at wave 2 (1^st year of high school, spring semester). For e-cigarettes, the SM disparity dissipated after wave 4 (2^nd year of high school, spring semester). This initial disparity may reflect greater susceptibility and willingness to try tobacco products among SM youth, which subsided over time. Conversely, poly-tobacco use disparities emerged early (waves 1-3), dissipated during mid-high school (waves 4-6), and re-emerged towards the end of high (waves 7-8). Together these findings highlight that disparities for various tobacco products may emerge at different time points in adolescence development. Importantly, preventing teen use of tobacco products is a priority for all teens (U.S. Department of Health and Human Services, 2012), and as such, prevention campaigns are commonplace in high schools across the country. However, these findings highlight that targeted prevention campaigns – enacted early during the first year of high school – may curb the added burden of tobacco use among SM youth.

The transition from high school to early adulthood may also represent a critical period for reducing SM tobacco use disparities. Indeed, SM females reported higher use of cigarettes, e-cigarettes, and poly-product use than non-SM females at wave 8 (end of high school). These disparities were maintained as respondents transitioned to early adulthood (wave 9), highlighting that the transition to adulthood is a critical period for tobacco use prevention interventions tailored to SM youth – particularly for SM females, for whom nearly half the sample had used a tobacco product in the past 6 months. However, while these disparities between SM and non-SM young adults widened somewhat over 1.5 years of follow-up, this change did not reach significance (i.e., the magnitude of the disparity did not change significantly). This finding may reflect a troubling trend of increased prevalence of tobacco product use increased among SM and non-SM young adults alike.

Limitations

This study has key limitations. First, low prevalence rates of “other” tobacco product use limited our ability to disaggregate this category into to more specific product types. Future research should assess differences in the timing and onset of disparities across specific tobacco products, such as cigars, cigarillos, and hookah. Additionally, a sensitive measure of SM status was used, which was inclusive of SM youth on the basis of their identities, attractions, and/or behaviors. As such, this study quantified the magnitude of tobacco use disparities experienced by SM youth, which is important for crafting effective policies and interventions. However, this measurement approach yields a heterogenous SM population of study. Future research should disentangle SM subgroups (e.g., analyze lesbian/gay and bisexual respondents separately) in analysis when possible. Third, information on SM status was not collected from respondents until early adulthood (wave 9). Sexual orientation development is a dynamic process that unfolds over adolescence (Floyd and Stein, 2002), and so respondents who reported different SM statuses at wave 9, compared to what they would have reported during high school, were miscategorized. Future research should assess SM status contemporaneously to tobacco product use. Finally, while there exist stark tobacco use disparities between transgender/non-binary and cisgender populations, we did not have enough transgender/non-binary respondents in our sample to make meaningful comparisons on the basis of gender identity. As such, we elected to stratify our analyses by sex (female vs. male), and to refer to respondents’ sex rather than their gender identity throughout the manuscript.

Conclusions

Using a diverse longitudinal sample of youth and young adults from Southern California, this study demonstrated marked disparities in tobacco product use between SM and non-SM youth. These disparities were concentrated among SM females, with use of tobacco products among SM males largely mirroring use by non-SM males. These findings highlight potential areas for interventions; tobacco cessation efforts must prioritize those most at risk for tobacco use – including SM females. Further, interventions must be appropriately timed, and depend on the tobacco product; cigarette disparities may be targeted prior to high school, while disparities in e-cigarette use may be more appropriately implemented during high school, and as students transition into young adulthood.

Highlights.

• Sexual minority (SM) tobacco use disparities are concentrated among SM females.

• SM tobacco use disparities extend across several tobacco products.

• SM tobacco use disparities extend through both adolescence and young adulthood.

• The timing of SM tobacco use disparities varies by product type.