There is consistent evidence that sexual minority youth (SMY) use more traditional cigarettes than their non-SMY counterparts. However, there is relatively less information on e-cigarettes and, importantly, differences between and within SMY populations by race and ethnicity and sex. This study examines e-cigarette use by sexual orientation status and the intersection of race and ethnicity and sex.
Data come from high school students in the 2020 and 2021 National Youth Tobacco Surveys (N = 16 633). Current e-cigarette use prevalence by sexual identity categories was calculated for racial and ethnic subgroups. Multivariable logistic regression analysis examined the association between sexual identity and e-cigarette use by race and ethnicity groups and sex.
E-cigarette use prevalence was higher for most SMY racial and ethnic groups than their non-SMY counterparts. However, multivariable logistic analysis showed varied results by race and ethnic groups, with higher e-cigarette use odds for SMY populations, although not statistically significant for some race and ethnic groups. Black gay or lesbian (adjusted odds ratio: 3.86, 95% confidence interval, 1.61–9.24) and bisexual (adjusted odds ratio: 3.31, 95% confidence interval, 1.32–8.30) high school students had significantly higher e-cigarette use odds than Black heterosexuals. Non-Hispanic Black females e-cigarettes use odds are 0.45 times that of non-Hispanic white males, and non-Hispanic other gay or lesbian had 3.15 times higher e-cigarette use odds than non-Hispanic white heterosexuals.
E-cigarette use is more prevalent among SMY populations. Disparities in e-cigarette use vary depending on race and ethnicity and sex.
There is consistent evidence that sexual minority youth (SMY) use more traditional cigarettes than their non-SMY counterparts. However, there is less information on SMY e-cigarette use and, importantly, differences within SMY populations by race and ethnicity and sex.
This provides the first evidence on SMY e-cigarette use by race and ethnicity and sex. E-cigarette use prevalence was higher for most SMY groups than for non-SMY counterparts, and differences varied significantly across race and ethnic groups and sex.
Over the past few years, electronic cigarettes (e-cigarettes) have become the most used tobacco product among youth and young adults in the United States,1 with approximately 1 in 5 high school students and 1 in 20 middle school students identified as current users.2 Evidence suggests that e-cigarettes are far from harmless,1,3 and the long-term health risks of e-cigarette use remain undecided. As e-cigarette use has become popular among youth1 and as most adults begin tobacco use during adolescence,4 it is pertinent to examine disparities in e-cigarette use, especially given the history of targeted marketing of tobacco products toward sexual minority populations.5,6 Furthermore, because sexual minorities uniquely face prejudice and stigma because of their identity, they may differentially use substances to manage such social stressors compared to heterosexuals.7,8 Analysis of this type is timely and can further our understanding of subgroup differences in e-cigarette use by race and ethnicity and sex, which is beneficial in promoting intervention targeting.9
Prior research has explored variations of e-cigarette use across numerous categories, including place of residence, race and ethnicity, occupation, sex, education level, socioeconomic status, sexual orientation, and age.6 Studies focused on sexual orientation found e-cigarette use differences among sexual minority youth (SMY) relative to heterosexual youth more broadly,5,10 and previous research suggested bisexual youth,11 sexual minority females,12 and bisexual females13 had notably higher rates of e-cigarette use among SMY. Furthermore, considerable variations in youth e-cigarette use across racial and ethnic groups have also been documented, with studies showing higher rates of e-cigarette use specifically among non-Hispanic white youths.14–19 Nonetheless, there is limited research on sociodemographic differences in e-cigarette use among sexual minority populations.20
The intersection of sexual orientation with race and ethnicity and sex has been examined for cigarette use and other tobacco products.21–24 For example, Blosnich et al (2011) found that sexual minority racial groups of adults had higher cigarette smoking prevalence than their heterosexual counterparts, with white and Hispanic sexual minorities being more likely to use hookahs than their respective heterosexuals.22 Similarly, data from the 2005 to 2007 Youth Risk Behavior Survey by Corliss and her colleagues in 2014 found that SMY used cigarettes more than heterosexual racial groups, and minority lesbians, younger bisexuals, and bisexual girls had a higher risk of cigarette smoking. However, this line of research has not explored the most used tobacco product, e-cigarettes, among SMY, although e-cigarette use has become widespread among youth, and previous research has shown differences in use by sexual orientation13,25 and across various races and ethnicities.14,16,17 Assuming homogeneity within sexual minority populations could mask differences in race and ethnicity and sex, which may be important for intervention targeting.21,22 We used the intersectionality approach, a useful framework for examining tobacco-related disparities across and within many marginalized groups.26 Intersectionality is an approach that considers the interaction of multiple factors (eg, social identities, societal factors) when analyzing inequities, and it provides a more holistic view of tobacco-related disparities than a single-factor approach provides.26 Our focus was nonetheless narrowed because many dimensions constituting intersectionality (eg, systemic exclusion, trauma) were beyond the scope of the data used in this study. The current study addresses the lack of research on e-cigarettes by examining the intersection of race and ethnicity and sex by sexual orientation among youth by examining the intersection of race and ethnicity and sex for e-cigarette use by youth sexual orientation status.
Methods
Data
The current study data come from the 2020 and 2021 National Youth Tobacco Survey (NYTS). The NYTS is a comprehensive national data set on tobacco measures for students in middle school (grades 6–8) and high school (grades 9–12) to support surveillance and evaluation of tobacco prevention programs at the state and national levels. A rich set of tobacco measures collected by the NYTS are tobacco products (e-cigarettes, cigarettes, cigars, smokeless tobacco, hookahs, roll-your-own cigarettes, pipes, snus, dissolvable tobacco, bidis, and heated tobacco products); exposure to secondhand smoke and e-cigarette aerosol; smoking cessation; minor ability to purchase or obtain tobacco products; and knowledge and perception about tobacco products. The NYTS uses a stratified, 3-stage cluster probability-based sampling frame to capture a nationally representative sample of middle and high school students.27 The NYTS design involves the selection of (1) primary sampling units (PSUs) (defined as a county, or a group of small counties, or part of a large county) within each stratum, (2) secondary sampling units (SSUs) (defined as schools or linked schools) within each selected PSU, and (3) students within each selected school. Students in randomly selected states and the District of Columbia are invited to participate each year, with participation confidential, voluntary, and approved by parents. Participants are given a tablet to complete the survey using a programmed survey application; those absent or unavailable on the survey day could participate in make-up surveys using a web-based questionnaire. The 2020 NYTS had a school participation rate of 49.9%, and 87.4% of students completed questionnaires, which yielded an overall participation rate of 43.6% (ie, the product of the school-level and student-level participation rates). The 2021 NYTS was administered as an online survey, with virtual assistance provided by trained technical support personnel. The school participation rate for the 2021 survey was 54.9%, the student participation rate was 81.2%, and the overall participation rate was 44.6%. This study was restricted to high school students.
Measures
The outcome variable, current e-cigarette use, was derived to capture any use in the past 30 days before the survey. Participants were provided a brief description of e-cigarettes, “The next several questions are about electronic cigarettes or e-cigarettes, such as JUUL, SMOK, Suorin, Vuse, blu, Puff Bar, or STIG. You also may know them as vapes, mods, e-cigs, e-hookahs, or vape-pens.” The primary independent variable of interest was sexual identity, which we captured from the question,” which of the following best describes you?” with responses such as “heterosexual (straight),” “lesbian or gay,” “bisexual,” and “not sure.” We used these four response categories to represent the sexual identity status of high school students. Other independent variables captured were grade levels (grades 9, 10, 11, and 12), sex (male and female), race and ethnicity (non-Hispanic white, non-Hispanic Black, Hispanic, and non-Hispanic other), tobacco use by household members, and other tobacco use. Non-Hispanic other includes those who self-reported as non-Hispanic Asian, non-Hispanic American Indian or Alaska Native, non-Hispanic Native Hawaiian or Other Pacific Islander, and those who self-reported as multiple non-Hispanic races. Other nicotine products used included any use of cigarettes, cigars, smokeless tobacco, hookahs, roll-your-own cigarettes, pipes, snus, dissolvable tobacco, or bidis in the past 30 days before the survey.
Statistical Analyses
Descriptive statistics were estimated for the full sample by using the combined 2020 and 2021 NYTS data among high school students in the United States. The prevalence of current e-cigarette use by sexual identity categories was calculated. The analysis also included current e-cigarette use by sexual identity categories among racial and ethnic subgroups. For all analyses, the reference groups reflect majority or privileged social identities (eg, non-Hispanic white, heterosexual [straight], male). Multivariable logistic regression analysis examined the association between sexual identity and current e-cigarette use by race and ethnicity groups (within-group differences), adjusting for grade levels, sex, tobacco use by household members, other nicotine products use, and survey year. Multivariable analysis was used to examine the intersection of race and ethnicity and sexual identity (between-group differences) through the interaction of race and ethnicity and sexual identity. A similar analysis was performed to examine the intersection of sex and race and ethnicity. A stratified multivariable analysis by sex was also conducted to examine the association between sexual identity categories and current e-cigarette use for each race and ethnicity group. Additional logistic regression analysis examined the association between sexual identity and current e-cigarette use by sex, adjusting for grade levels, race and ethnicity groups, sex, tobacco use by household members, other nicotine products use, and survey year.28 All analyses used sampling weight to account for the complex survey design.
Results
Table 1 reports the characteristics of the weighted study sample using the 2020 and 2021 NYTS. The proportion of students was almost balanced by grade level, with grades 9 (26.7%) and 10 (25.6%) slightly more than grades 11 and 12. Of 16 633 students, 81.4% were heterosexual (straight), 3.3% were lesbian or gay, 9.7% were bisexual, and 5.6% were unsure. Much of the sample was non-Hispanic white (51.6%), followed by Hispanic (24.9%), non-Hispanic other (11.9%), and non-Hispanic Black (11.6%). Males (51.1%) were slightly more than females (48.9%), and ∼66% of students came from households with no tobacco use. About 14.7% of high school students were current e-cigarette users.
Descriptive Statistics, 2020–2021
. | Full Sample . |
---|---|
Na | 16 633 |
Grade | |
9th | 26.69 (25.13–28.25) |
10th | 25.61 (24.36–26.86) |
11th | 24.36 (23.24–25.47) |
12th | 23.35 (22.13–24.56) |
Sex | |
Male | 51.11 (49.05–53.17) |
Female | 48.89 (46.83–50.95) |
Race and ethnicity | |
Non-Hispanic white | 51.60 (47.93–55.27) |
Non-Hispanic Black | 11.57 (9.37–13.76) |
Hispanic | 24.92 (22.04–27.80) |
Non-Hispanic other | 11.91 (10.20–13.61) |
Current e-cigarette use | |
Yes | 14.77 (13.02–16.53) |
No | 85.23 (83.47–86.98) |
Other nicotine products useb | |
Yes | 7.73 (6.55–8.91) |
No | 92.27 (91.09–93.45) |
Tobacco use by household members | |
Yes | 34.09 (32.07–36.11) |
No | 65.91 (63.89–67.93) |
Sexual identity | |
Heterosexual (straight) | 81.40 (80.25–82.55) |
Gay or lesbian | 3.30 (2.89–3.72) |
Bisexual | 9.70 (8.88–10.53) |
Not Sure | 5.60 (5.10–6.09) |
. | Full Sample . |
---|---|
Na | 16 633 |
Grade | |
9th | 26.69 (25.13–28.25) |
10th | 25.61 (24.36–26.86) |
11th | 24.36 (23.24–25.47) |
12th | 23.35 (22.13–24.56) |
Sex | |
Male | 51.11 (49.05–53.17) |
Female | 48.89 (46.83–50.95) |
Race and ethnicity | |
Non-Hispanic white | 51.60 (47.93–55.27) |
Non-Hispanic Black | 11.57 (9.37–13.76) |
Hispanic | 24.92 (22.04–27.80) |
Non-Hispanic other | 11.91 (10.20–13.61) |
Current e-cigarette use | |
Yes | 14.77 (13.02–16.53) |
No | 85.23 (83.47–86.98) |
Other nicotine products useb | |
Yes | 7.73 (6.55–8.91) |
No | 92.27 (91.09–93.45) |
Tobacco use by household members | |
Yes | 34.09 (32.07–36.11) |
No | 65.91 (63.89–67.93) |
Sexual identity | |
Heterosexual (straight) | 81.40 (80.25–82.55) |
Gay or lesbian | 3.30 (2.89–3.72) |
Bisexual | 9.70 (8.88–10.53) |
Not Sure | 5.60 (5.10–6.09) |
The weighted column percentage and its 95% confidence interval are presented.
The final sample was restricted to 16 633 high school students with valid responses to the sexual identity question.
Other nicotine products included any use of cigarettes, cigars, smokeless tobacco, hookahs, roll-your-own cigarettes, pipes, snus, dissolvable tobacco, or bidis in the past 30 d before the survey.
Table 2 presents the prevalence of current e-cigarette use by sexual identity categories for the full sample and race and ethnicity. Gay or lesbian (21.5%) and bisexuals (18.1%) had a higher prevalence of e-cigarette use than heterosexuals (14.4%) and those unsure about their sexual identity (10.5%). Similarly, analysis by race and ethnicity showed a higher prevalence of e-cigarette use by gay or lesbian, and bisexual students. For non-Hispanic white students, e-cigarette use by gay or lesbian (23.8%) and bisexual (22.5%) groups was >20%, whereas lower for heterosexual groups (17.7%). Also, among non-Hispanic Black students, results showed e-cigarette use was higher for gay or lesbian (17.2%) and bisexual groups (13%) compared to heterosexual groups. For Hispanic and non-Hispanic groups, e-cigarette use was consistently higher for gay or lesbian and bisexual students.
Percentage of e-Cigarette Use Among Sexual Identity Groups by Race and Ethnicity, 2020 to 2021 (Na = 16 633)
Sexual Identity . | E-Cigarettes . |
---|---|
Full sample | |
Heterosexual (straight) | 14.40 (12.66–16.15) |
Gay or lesbian | 21.46 (16.32–26.60) |
Bisexual | 18.09 (14.67–21.51) |
Not sure | 10.51 (7.21–13.80) |
Male | |
Heterosexual (straight) | 14.28 (12.32–16.24) |
Gay or lesbian | 30.15 (21.28–39.01) |
Bisexual | 17.70 (11.87–23.52) |
Not sure | 12.28 (7.60–16.96) |
Female | |
Heterosexual (straight) | 14.52 (12.60–16.44) |
Gay or lesbian | 15.06 (9.58–20.53) |
Bisexual | 18.27 (14.67–21.87) |
Not Sure | 9.24 (5.60–12.89) |
Non-Hispanic white | |
Heterosexual (straight) | 17.75 (15.77–19.73) |
Gay or lesbian | 23.79 (16.51–31.07) |
Bisexual | 22.53 (18.10–26.97) |
Not sure | 12.91 (7.99–17.84) |
Non-Hispanic Black | |
Heterosexual (straight) | 4.83 (3.27–6.40) |
Gay or lesbian | 17.19 (6.32–28.06) |
Bisexual | 13.02 (4.89–21.15) |
Not sure | 6.86 (1.75–11.96) |
Hispanic | |
Heterosexual (straight) | 12.48 (9.62–15.34) |
Gay or lesbian | 18.46 (8.70–28.21) |
Bisexual | 15.15 (9.85–20.45) |
Not sure | 11.51 (5.06–17.97) |
Non-Hispanic Other | |
Heterosexual (straight) | 13.25 (10.49–16.02) |
Gay or lesbian | 25.32 (11.47–39.17) |
Bisexual | 12.02 (6.48–17.57) |
Not sure | 4.74 (1.13–8.35) |
Sexual Identity . | E-Cigarettes . |
---|---|
Full sample | |
Heterosexual (straight) | 14.40 (12.66–16.15) |
Gay or lesbian | 21.46 (16.32–26.60) |
Bisexual | 18.09 (14.67–21.51) |
Not sure | 10.51 (7.21–13.80) |
Male | |
Heterosexual (straight) | 14.28 (12.32–16.24) |
Gay or lesbian | 30.15 (21.28–39.01) |
Bisexual | 17.70 (11.87–23.52) |
Not sure | 12.28 (7.60–16.96) |
Female | |
Heterosexual (straight) | 14.52 (12.60–16.44) |
Gay or lesbian | 15.06 (9.58–20.53) |
Bisexual | 18.27 (14.67–21.87) |
Not Sure | 9.24 (5.60–12.89) |
Non-Hispanic white | |
Heterosexual (straight) | 17.75 (15.77–19.73) |
Gay or lesbian | 23.79 (16.51–31.07) |
Bisexual | 22.53 (18.10–26.97) |
Not sure | 12.91 (7.99–17.84) |
Non-Hispanic Black | |
Heterosexual (straight) | 4.83 (3.27–6.40) |
Gay or lesbian | 17.19 (6.32–28.06) |
Bisexual | 13.02 (4.89–21.15) |
Not sure | 6.86 (1.75–11.96) |
Hispanic | |
Heterosexual (straight) | 12.48 (9.62–15.34) |
Gay or lesbian | 18.46 (8.70–28.21) |
Bisexual | 15.15 (9.85–20.45) |
Not sure | 11.51 (5.06–17.97) |
Non-Hispanic Other | |
Heterosexual (straight) | 13.25 (10.49–16.02) |
Gay or lesbian | 25.32 (11.47–39.17) |
Bisexual | 12.02 (6.48–17.57) |
Not sure | 4.74 (1.13–8.35) |
The final sample was restricted to 16 633 high school students with valid responses to the sexual identity question.
Table 3 presents results from logistic regression models with the interaction between sexual identity and race and ethnicity (Model 1) and the interaction between sex and race and ethnicity (Model 2). In Model 1, controlling for interaction between sexual identity and race and ethnicity, non-Hispanic Black heterosexuals had significantly lower odds of e-cigarette use than non-Hispanic white heterosexuals (adjusted odds ratio [aOR]: 0.54, 95% confidence interval [CI], 0.38–0.77). Among non-Hispanic white, those who were unsure about their sexual identity were less likely to use e-cigarettes than their heterosexual counterparts. In addition, non-Hispanic other gay or lesbian had 3.15 times higher odds of e-cigarette use than non-Hispanic white heterosexuals. In Model 2, adjusting for interaction between sex and race and ethnicity, gay or lesbian males had significantly higher odds of e-cigarette use (aOR: 1.38, 95% CI, 1.12–1.72) compared to heterosexual males, whereas those unsure of their sexual identity had lower odds of e-cigarette use. The odds for non-Hispanic Black females to use e-cigarettes are 0.45 times that of non-Hispanic white males.
Association Between Sexual Identity and Current e-Cigarette Use Intersection With Race and Ethnicity and Sex
. | Model 1a,b . | Model 2b,c . |
---|---|---|
Sex | ||
Female | 1.07 (0.94–1.23) | 0.89 (0.75–1.05) |
Male | Ref | Ref |
Race and ethnicity | ||
Non-Hispanic white | Ref | — |
Non-Hispanic Black | 0.54 (0.38–0.77) | 0.85 (0.43–1.69) |
Hispanic | 1.13 (0.87–1.46) | 1.55 (1.00–2.39) |
Non-Hispanic other | 1.00 (0.76–1.30) | 0.70 (0.43–1.17) |
Sexual identity | ||
Heterosexual (straight) | Ref | Ref |
Gay or lesbian | 1.64 (1.26–2.14) | 1.38 (1.12–1.72) |
Bisexual | 1.10 (0.83–1.47) | 1.13 (0.90–1.42) |
Not sure | 0.59 (0.43–0.81) | 0.62 (0.47–0.82) |
Interaction between sexual identity and race and ethnicity | ||
Non-Hispanic Black gay or lesbian | 1.06 (0.55–2.01) | — |
Hispanic gay or lesbian | 0.68 (0.41–1.14) | — |
Non-Hispanic other gay or lesbian | 1.92 (1.17–3.16) | — |
Non-Hispanic Black bisexual | 1.37 (0.70–2.69) | — |
Hispanic bisexual | 0.93 (0.62–1.40) | — |
Non-Hispanic other bisexual | 0.70 (0.39–1.25) | — |
Non-Hispanic Black not sure | 1.11 (0.56–2.19) | — |
Hispanic not sure | 1.35 (0.81–2.26) | — |
Non-Hispanic other not sure | 0.62 (0.33–1.17) | — |
Interaction between sex and race and ethnicity | ||
Non-Hispanic Black female | — | 0.60 (0.37–0.98) |
Hispanic female | — | 0.86 (0.66–1.13) |
Non-Hispanic other female | — | 1.35 (0.95–1.93) |
. | Model 1a,b . | Model 2b,c . |
---|---|---|
Sex | ||
Female | 1.07 (0.94–1.23) | 0.89 (0.75–1.05) |
Male | Ref | Ref |
Race and ethnicity | ||
Non-Hispanic white | Ref | — |
Non-Hispanic Black | 0.54 (0.38–0.77) | 0.85 (0.43–1.69) |
Hispanic | 1.13 (0.87–1.46) | 1.55 (1.00–2.39) |
Non-Hispanic other | 1.00 (0.76–1.30) | 0.70 (0.43–1.17) |
Sexual identity | ||
Heterosexual (straight) | Ref | Ref |
Gay or lesbian | 1.64 (1.26–2.14) | 1.38 (1.12–1.72) |
Bisexual | 1.10 (0.83–1.47) | 1.13 (0.90–1.42) |
Not sure | 0.59 (0.43–0.81) | 0.62 (0.47–0.82) |
Interaction between sexual identity and race and ethnicity | ||
Non-Hispanic Black gay or lesbian | 1.06 (0.55–2.01) | — |
Hispanic gay or lesbian | 0.68 (0.41–1.14) | — |
Non-Hispanic other gay or lesbian | 1.92 (1.17–3.16) | — |
Non-Hispanic Black bisexual | 1.37 (0.70–2.69) | — |
Hispanic bisexual | 0.93 (0.62–1.40) | — |
Non-Hispanic other bisexual | 0.70 (0.39–1.25) | — |
Non-Hispanic Black not sure | 1.11 (0.56–2.19) | — |
Hispanic not sure | 1.35 (0.81–2.26) | — |
Non-Hispanic other not sure | 0.62 (0.33–1.17) | — |
Interaction between sex and race and ethnicity | ||
Non-Hispanic Black female | — | 0.60 (0.37–0.98) |
Hispanic female | — | 0.86 (0.66–1.13) |
Non-Hispanic other female | — | 1.35 (0.95–1.93) |
The final sample was restricted to 16 633 high school students with valid responses to the sexual identity question.
Ref, reference; —, not applicable.
The weighted multivariable logistic regression models were used to test the intersection between sexual identity and race and ethnicity, adjusting for grade, sex, tobacco use by household members, other nicotine products use, and survey year.
The weighted multivariable logistic regression models were used to examine the intersection between sex and race and ethnicity, adjusting for grade, sexual identity, tobacco use by household members, other nicotine products use, and survey year.
The c statistics, which measures the goodness of fit for binary outcomes in a logistic regression model, is 0.78 for both models 1 and 2, indicating good models.
Multivariable logistic regression analyses examining the association between sexual identity categories and current e-cigarette use by race and ethnicity are shown in Table 4. For non-Hispanic Black students, gay or lesbian (aOR: 3.86, 95% CI, 1.61–9.24) and bisexual (aOR: 3.31, 95% CI, 1.32–8.30) groups had statistically significant higher odds of e-cigarette use compared to Black heterosexuals. E-cigarette use was significantly higher among gay or lesbian students (aOR: 2.58, 95% CI, 1.19–5.61) in the non-Hispanic other group than heterosexual (straight) students.
Association Between Sexual Identity and Current e-Cigarette Use by Race and Ethnicity
. | Non-Hispanic White . | Non-Hispanic Black . | Hispanic . | Non-Hispanic Other . |
---|---|---|---|---|
Sexual identity | ||||
Heterosexual (straight) | Ref | Ref | Ref | Ref |
Gay or lesbian | 1.06 (0.70–1.60) | 3.86 (1.61–9.24) | 1.04 (0.47–2.32) | 2.58 (1.19, 5.61) |
Bisexual | 1.06 (0.82–1.37) | 3.31 (1.32–8.30) | 1.12 (0.59–2.13) | 0.68 (0.31, 1.49) |
Not Sure | 0.56 (0.33–0.97) | 1.25 (0.52–3.03) | 0.75 (0.40–1.40) | 0.32 (0.13, 0.79) |
Grade | ||||
9th | 0.34 (0.26–0.45) | 0.53 (0.24–1.18) | 0.60 (0.38–0.96) | 0.47 (0.23, 0.95) |
10th | 0.58 (0.45–0.75) | 0.61 (0.28–1.31) | 0.93 (0.69–1.27) | 0.73 (0.40–1.33) |
11th | 0.71 (0.59–0.87) | 1.23 (0.54–2.82) | 0.80 (0.57–1.11) | 0.81 (0.43–1.52) |
12th | Ref | Ref | Ref | Ref |
Sex | ||||
Female | 1.29 (1.09–1.52) | 0.48 (0.29–0.79) | 0.77 (0.58–1.02) | 1.23 (0.80–1.87) |
Male | Ref | Ref | Ref | Ref |
Tobacco use by household members | ||||
Yes | 2.39 (2.06–2.77) | 1.51 (0.83–2.76) | 1.90 (1.50–2.41) | 1.93 (1.11–3.35) |
No | Ref | Ref | Ref | Ref |
Other nicotine products use | ||||
Yes | 16.03 (11.61–22.12) | 8.85 (4.91–15.96) | 12.97 (9.10–18.49) | 9.35 (5.51–15.87) |
No | Ref | Ref | Ref | Ref |
Survey year | ||||
2020 | Ref | Ref | Ref | Ref |
2021 | 0.59 (0.48–0.72) | 0.64 (0.32–1.26) | 0.44 (0.29–0.67) | 0.65 (0.45–0.93) |
. | Non-Hispanic White . | Non-Hispanic Black . | Hispanic . | Non-Hispanic Other . |
---|---|---|---|---|
Sexual identity | ||||
Heterosexual (straight) | Ref | Ref | Ref | Ref |
Gay or lesbian | 1.06 (0.70–1.60) | 3.86 (1.61–9.24) | 1.04 (0.47–2.32) | 2.58 (1.19, 5.61) |
Bisexual | 1.06 (0.82–1.37) | 3.31 (1.32–8.30) | 1.12 (0.59–2.13) | 0.68 (0.31, 1.49) |
Not Sure | 0.56 (0.33–0.97) | 1.25 (0.52–3.03) | 0.75 (0.40–1.40) | 0.32 (0.13, 0.79) |
Grade | ||||
9th | 0.34 (0.26–0.45) | 0.53 (0.24–1.18) | 0.60 (0.38–0.96) | 0.47 (0.23, 0.95) |
10th | 0.58 (0.45–0.75) | 0.61 (0.28–1.31) | 0.93 (0.69–1.27) | 0.73 (0.40–1.33) |
11th | 0.71 (0.59–0.87) | 1.23 (0.54–2.82) | 0.80 (0.57–1.11) | 0.81 (0.43–1.52) |
12th | Ref | Ref | Ref | Ref |
Sex | ||||
Female | 1.29 (1.09–1.52) | 0.48 (0.29–0.79) | 0.77 (0.58–1.02) | 1.23 (0.80–1.87) |
Male | Ref | Ref | Ref | Ref |
Tobacco use by household members | ||||
Yes | 2.39 (2.06–2.77) | 1.51 (0.83–2.76) | 1.90 (1.50–2.41) | 1.93 (1.11–3.35) |
No | Ref | Ref | Ref | Ref |
Other nicotine products use | ||||
Yes | 16.03 (11.61–22.12) | 8.85 (4.91–15.96) | 12.97 (9.10–18.49) | 9.35 (5.51–15.87) |
No | Ref | Ref | Ref | Ref |
Survey year | ||||
2020 | Ref | Ref | Ref | Ref |
2021 | 0.59 (0.48–0.72) | 0.64 (0.32–1.26) | 0.44 (0.29–0.67) | 0.65 (0.45–0.93) |
The final sample was restricted to 16 633 high school students with valid responses to the sexual identity question. Weighted multivariable logistic regression models were used to test the association between sexual identity and current e-cigarette use, adjusting for grade, sex, tobacco use by household members, other nicotine products use, and survey year. The c statistics, which measure the goodness of fit for binary outcomes in a logistic regression model, ranged from 0.68 to 0.81, indicating good models. Ref, reference.
Results from the stratified analysis by sex are reported in Table 5. For non-Hispanic Black, the odds of e-cigarette use were higher for gay males (aOR: 4.98, 95% CI, 1.03–24.21) and gay or lesbian females (aOR: 3.16, 95% CI, 1.19–8.42). Likewise, bisexual females were more likely to be current e-cigarette users (aOR: 4.33, 95% CI, 1.56–12.07). For Hispanic youth, the stratified analysis results were consistent with the baseline model, showing no statistically significant difference in e-cigarette use by sexual identity categories and sex. Gay males were more likely to be e-cigarette users (aOR: 2.89, 95% CI, 1.20–6.97) among the non-Hispanic other group, which confirms that the baseline results could be because of males. Additional analyses focusing on within comparisons, including bisexual youth as the reference group, are shown in Supplementary Tables 6 and 7. The results were largely consistent between the main and stratified analyses.
Association Between Sexual Identity and Current e-Cigarette Use by Race and Ethnicity Groups and Sex
Sexual identity . | Full Male and Female . | Male . | Female . |
---|---|---|---|
Non-Hispanic white | |||
Heterosexual (straight) | Ref | Ref | Ref |
Gay or lesbian | 1.06 (0.70–1.60) | 1.56 (0.84–2.88) | 0.77 (0.38–1.55) |
Bisexual | 1.06 (0.82–1.37) | 0.95 (0.54–1.64) | 1.07 (0.80–1.43) |
Not sure | 0.56 (0.33–0.97) | 0.69 (0.35–1.36) | 0.49 (0.21–1.10) |
Non-Hispanic Black | |||
Heterosexual (straight) | Ref | Ref | Ref |
Gay or lesbian | 3.86 (1.61–9.24) | 4.98 (1.03–24.21) | 3.16 (1.19–8.42) |
Bisexual | 3.31 (1.32–8.30) | 1.47 (0.35–6.30) | 4.33 (1.56–12.07) |
Not sure | 1.25 (0.52–3.03) | 1.51 (0.71–3.24) | 0.84 (0.13–5.57) |
Hispanic | |||
Heterosexual (straight) | Ref | Ref | Ref |
Gay or lesbian | 1.04 (0.47–2.32) | 1.40 (0.41–4.82) | 0.79 (0.32–1.92) |
Bisexual | 1.12 (0.59–2.13) | 1.27 (0.54–2.98) | 1.13 (0.55–2.32) |
Not sure | 0.75 (0.40–1.40) | 0.64 (0.26–1.60) | 0.81 (0.38–1.70) |
Non-Hispanic Other | |||
Heterosexual (straight) | Ref | Ref | Ref |
Gay or lesbian | 2.58 (1.19–5.61) | 2.89 (1.20–6.97) | 2.23 (0.73–6.84) |
Bisexual | 0.68 (0.31–1.49) | 0.75 (0.15–3.68) | 0.65 (0.29–1.48) |
Not sure | 0.32 (0.13–0.79) | 0.14 (0.03–0.64) | 0.52 (0.18–1.56) |
Sexual identity . | Full Male and Female . | Male . | Female . |
---|---|---|---|
Non-Hispanic white | |||
Heterosexual (straight) | Ref | Ref | Ref |
Gay or lesbian | 1.06 (0.70–1.60) | 1.56 (0.84–2.88) | 0.77 (0.38–1.55) |
Bisexual | 1.06 (0.82–1.37) | 0.95 (0.54–1.64) | 1.07 (0.80–1.43) |
Not sure | 0.56 (0.33–0.97) | 0.69 (0.35–1.36) | 0.49 (0.21–1.10) |
Non-Hispanic Black | |||
Heterosexual (straight) | Ref | Ref | Ref |
Gay or lesbian | 3.86 (1.61–9.24) | 4.98 (1.03–24.21) | 3.16 (1.19–8.42) |
Bisexual | 3.31 (1.32–8.30) | 1.47 (0.35–6.30) | 4.33 (1.56–12.07) |
Not sure | 1.25 (0.52–3.03) | 1.51 (0.71–3.24) | 0.84 (0.13–5.57) |
Hispanic | |||
Heterosexual (straight) | Ref | Ref | Ref |
Gay or lesbian | 1.04 (0.47–2.32) | 1.40 (0.41–4.82) | 0.79 (0.32–1.92) |
Bisexual | 1.12 (0.59–2.13) | 1.27 (0.54–2.98) | 1.13 (0.55–2.32) |
Not sure | 0.75 (0.40–1.40) | 0.64 (0.26–1.60) | 0.81 (0.38–1.70) |
Non-Hispanic Other | |||
Heterosexual (straight) | Ref | Ref | Ref |
Gay or lesbian | 2.58 (1.19–5.61) | 2.89 (1.20–6.97) | 2.23 (0.73–6.84) |
Bisexual | 0.68 (0.31–1.49) | 0.75 (0.15–3.68) | 0.65 (0.29–1.48) |
Not sure | 0.32 (0.13–0.79) | 0.14 (0.03–0.64) | 0.52 (0.18–1.56) |
The final sample was restricted to 16 633 high school students with valid responses to the sexual identity question. Interpret the stratified analysis with caution because of small cell sizes. Weighted multivariable logistic regression models were used to test the association between sexual identity and current e-cigarette use, adjusting for grade, tobacco use by household members, other nicotine products use, and survey year. The c statistics, which measure the goodness of fit for binary outcomes in a logistic regression model, ranged from 0.68 to 0.82, indicating good models. Ref, reference.
Discussion
This study examined e-cigarette use among youth by focusing on sexual orientation and race and ethnicity and using a nationally representative sample of high school students. We examined patterns among four sexual identity groups, including gay or lesbian, bisexual, heterosexual, and not sure groups, and across 4 race and ethnic groups: Hispanic, non-Hispanic white, non-Hispanic Black, and non-Hispanic other.
Our findings revealed that sexual minority youth have a higher prevalence of e-cigarette use than heterosexual youth across races and ethnicities. These findings are in keeping with previous studies showing that disparities persist in substance use in general among sexual minority youth and adults.10,12,29–33 Studies examining sexual minority populations commonly draw from the minority stress theory, which posits that disparities are largely because of social and cultural stressors that are disproportionately experienced by nonheterosexual groups.34 Resulting from the uneven distribution of stressors, including prejudice and discrimination34 ; harassment25 ; psychological stress, victimization, and lack of supportive environment35 ; and the negative impacts of associated stigma,8 minority populations may use substances more frequently as a coping mechanism, although e-cigarette use might have long-term adverse consequences for distress.36 Additionally, some studies point to sexual minority youth being more likely to experience bullying,13,37 and at the same time, preliminary data show that bullying victims were more likely to use e-cigarettes.38 The impact of these stressors might be magnified by the targeted marketing of tobacco companies toward sexual minority populations.21,39,40 Additionally, we found no significant differences between bisexuals and gay or lesbian groups although increased vulnerability among bisexuals (particularly females) has been documented across various substances, including e-cigarettes.10 Across all analyses, those who were unsure about their sexual identity had a lower prevalence of e-cigarettes compared to those who self-identified as gay or lesbian or bisexual. It is unclear what protective factors may exist for this population, and qualitative research could elicit a better understanding of this finding.
We also found significantly higher odds of e-cigarette use among non-Hispanic Black SMY, particularly gay males, gay females, and bisexual females. Although not directly comparable with previous studies because of a dearth of literature on SMY e-cigarette use disparities in relation to race and ethnicity and sex, there is evidence of race and ethnicity differences in e-cigarette use. For example, Dai and colleagues found that non-Hispanic Black and Hispanic youth were more likely to be occasional e-cigarette users (defined as ≤5 days in the past 30), but they were less likely to be frequent users (defined as ≥ 20 days in the past 30) when compared to non-Hispanic white youth.15 Elsewhere, a study found that Black adults were less likely to have ever used an e-cigarette but were significantly more likely to endorse plans to continue e-cigarette use when compared to white adults.41 Unlike our current study, these analyses did not examine SMY as well as differences by race and ethnicity and sex.
Several reasons may partially explain why non-Hispanic Black SMY were more likely to use e-cigarettes than Black heterosexuals. First, racism and homophobia may intersect to bear psychological impacts on marginalized individuals, including increased stress, despair, and negative affect.26 Studies have shown that African American sexual minorities have a heightened risk of poor mental health than their white sexual minority counterparts.42,43 Second and related to the first, an array of social identities affect individuals’ lives in different ways,26 and some of these can bear compounded consequences for marginalized groups. For example, African American lesbian, gay, and bisexual adolescents may face victimization because of their sexual orientation and discrimination because of race and ethnicity, and together, these may exacerbate mental health risks.44 Notably, multiple minoritized identities may produce intensified stress, resulting in intensified coping among non-Hispanic Black sexual minority youth.34 Third, not only does the tobacco industry disproportionately market its products to the sexual minority community, but it also disproportionately targets the Black community.45,46 Thus, identities involving sexual minority status and non-Hispanic Black race and ethnicity may be at heightened risk than either category taken separately.
We note the limitations of this study. First, study measures are self-reported, resulting in social desirability bias from study participants (eg, sexual orientation status). Second, the data come from a school-based survey and may not be generalizable to youth not attending schools. Third, we combined 2 NYTS data years to have enough cell sizes for SMY subgroup analysis; some race and ethnic groups may not have had enough power, and collapsed race and ethnic groups might hide unique contributing or protective factors. Fourth, the analytical design of the study precludes us from making causal inferences. Additionally, we could not obtain the use of products containing nicotine or nicotine salts in the past 30 days, because this information was not available in the 2020 NYTS, in which these class products contain different harms compared to nonnicotine vaping devices.9
Future research would benefit from exploring how other stressors among sexual minority youth differentially affect race and ethnic groups. Certain stressors (eg, discrimination and prejudice, parental disapproval) may affect unique groups and qualitatively identifying these stressors, including discerning their disproportionate impact, could improve current prevention efforts. Second, research examining the intersection of other social identities with race and ethnicity and sexual orientation could yield important findings regarding e-cigarette use. For example, differences in socioeconomic position or mental health conditions may present significant challenges specific to certain racial and ethnic and sexual orientation subgroups. Gaining a better understanding of such intersections would reshape intervention efforts to be tailored to meet the needs of diverse groups.
Conclusions
Several studies point to the increased risk of cigarette use among sexual minority populations because of multiple factors, including social and cultural stressors that they disproportionately experience.23,31,47 However, there is limited research about the differences in e-cigarette use within sexual minority youth populations, particularly by race and ethnicity and sex. This study was among the first to consider the intersection of race and ethnicity with sexual orientation regarding e-cigarette use among youth in the United States. Informed by the intersectionality approach,26 this study used a nationally representative sample of high school students to examine the intersection of race and ethnicity and sex for e-cigarette use by youth sexual orientation status. Findings suggest significant differences in SMY e-cigarette use across race and ethnic groups and sex. In particular, non-Hispanic Black gay males, gay females, and bisexual females had elevated e-cigarette use compared to their heterosexual racial group counterparts. Intersections of sexual orientation with race and ethnicity and sex are important to understand e-cigarette use differences among youth, which is beneficial for tailored intervention targeting.
Dr Azagba conceptualized and designed the study, drafted the initial manuscript, and reviewed and revised the manuscript; Dr Ebling drafted the initial manuscript and reviewed the manuscript; Mr Shan conducted the data analysis and revised it critically for important intellectual content; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
CONFLICT OF INTEREST DISCLOSURES: The authors have no conflicts of interest relevant to disclose.
FUNDING: No external funding.
Comments