Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Jun 1.
Published in final edited form as: Psychol Addict Behav. 2019 Apr 8;33(4):371–381. doi: 10.1037/adb0000465

Explaining Sexual Minority Young Adult Cigarette Smoking Disparities

Josephine T Hinds 1, Alexandra Loukas 2, Cheryl L Perry 3
PMCID: PMC6584955  NIHMSID: NIHMS1016657  PMID: 30958013

Abstract

Objective:

Sexual minority (SM) young adults, such as those who identify as lesbian, gay, or bisexual (LGB), have well documented smoking disparities compared to heterosexual young adults. However, no studies have simultaneously tested the role of three risk factors (depressive symptoms, recalling tobacco marketing in bars, and cigarette-related social norms) to explain SM tobacco use disparities. Longitudinal structural equation modeling was used to explore if the association between SM identity and past 30-day cigarette smoking one year later was mediated by these three risk factors.

Methods:

Starting in fall 2015, three surveys were administered every six months to 3,972 young adult college students. Surveys assessed SM identity, depressive symptoms, tobacco marketing in bars, normative perceptions of cigarettes, and tobacco use behaviors. Greater depressive symptoms, recalling more tobacco marketing in bars, and more accepting cigarette-related social norms were each hypothesized to explain a unique portion of the association between SM identity and subsequent cigarette use.

Results:

SM young adults reported higher prevalence of cigarette use, depressive symptom scores reflecting elevated risk for major depressive disorder, and more accepting cigarette-related social norms than their heterosexual peers. Results indicated that only cigarette-related social norms mediated the association between SM identity and subsequent past 30-day smoking, while controlling for depressive symptoms, recalling tobacco marketing in bars, sociodemographic factors, and previous tobacco use.

Conclusions:

Findings reflect a need for tailored tobacco control messaging that shift the more accepting cigarette-related norms of SM young adults, as doing so may ultimately lead to decreased smoking for this high-risk subgroup.

Keywords: LGB smoking disparities, young adults, risk for tobacco use

High smoking rates among sexual minority (SM) individuals, such as those who identify as lesbian, gay, or bisexual (LGB), are well documented. In the United States (U.S.), over 20% of SM adults reported smoking cigarettes “every day or some days” on the 2016 National Health Interview Survey, compared to 15.3% of heterosexual adults (Jamal et al., 2018). These findings align with a number of nationally representative studies which consistently identify the larger lesbian, gay, and bisexual SM community as smoking more than their heterosexual peers (Agaku et al., 2014; Emory et al., 2015; Hu et al., 2016; Phillips et al., 2017). Considering additional variables like biological sex and SM subgroup adds complexity to these comparisons. Bisexual women have emerged as the subgroup with the highest risk for smoking, compared to other SM men and women, and to heterosexual women (Emory et al., 2015; Johnson et al., 2016). The same cannot be said for bisexual men, where gay (not bisexual) sexual orientation typically places men at an elevated risk for cigarette use compared to their heterosexual peers (Emory et al., 2015; Hoffman, Delahanty, Johnson, & Zhao, 2018; Johnson et al., 2016; Schuler, Rice, Evans-Polce, & Collins, 2018; Wheldon, Kaufman, Kasza, & Moser, 2018). However, a few studies also identify bisexual men at elevated risk for smoking compared to heterosexual men, though not more than their gay peers (Max, Stark, Sung, & Offen, 2016; McCabe et al., 2018).

Alarming smoking prevalence rates among SM individuals are evident in young adulthood, the developmental period when most regular smoking begins (U.S. Surgeon General, 2014). In a 2013 nationally representative study, 34.8% of gay and lesbian young adults and 27.5% of bisexual young adults smoked cigarettes in the past 30 days, compared to 18.5% of their heterosexual peers (Rath, Villanti, Rubenstein, & Vallone, 2013). These prevalence estimates translate to an almost fivefold increased likelihood of regular cigarette use among bisexual young adult women compared to their heterosexual peers, and just over twice the likelihood of regular cigarette smoking among gay men than their heterosexual peers in the 2013–2014 Population Assessment of Tobacco and Health (PATH) study (Wheldon et al., 2018). Alarming prevalence estimates aside, studies that explore reasons for the increased prevalence of smoking among SM young adults are less common. While SM individuals typically have an earlier age of smoking initiation than their heterosexual peers (Corliss et al., 2013), and report unique risk factors that increase the likelihood of smoking, such as attendance at a bar or club, and a proximity to other smokers (Nguyen et al., 2018), it is important to further understand what may be causing these behaviors. The purpose of this study was to explore why smoking disparities exist by examining the role of three potential mediators in the association between SM identity and cigarette smoking one year later: depressive symptoms, tobacco marketing in bars and clubs, and cigarette-related social norms.

The primary explanation for the elevated prevalence of smoking in SM populations involves Meyer and colleagues’ minority stress model, which posits that members of a marginalized group engage in behaviors like tobacco use to cope with stressful experiences associated with SM identity (Meyer, 2003). These experiences include prejudice or discrimination in interactions with others, and internalized stressors such as expectations of rejection and internalized homophobia (Meyer, 2003). There is a growing body of research that documents psychiatric morbidity stemming from these stressors, including depression and anxiety (Hatzenbuehler, 2009; Ross et al., 2018). In their 2008 systematic review, King and colleagues reported a twofold increase in the likelihood of a depressive episode in the past 12 months for LGB individuals compared to heterosexual people (King et al., 2008). In their representative sample of New Zealand young adults, Fergusson and colleagues noted a significant association between non-heterosexual sexual orientations and negative mental health outcomes (Fergusson, Horwood, Ridder, & Beautrais, 2005). Specifically, the proportion of SM young adults meeting the diagnostic criteria for major depression was between 1.5 and 2.1 times greater than that of heterosexuals for SM women, and between 2.9 and 4.9 times greater for SM men (Fergusson et al., 2005). Additionally, the association between depressive symptoms and smoking outcomes is well established. While directionality is not entirely clear, there is a robust body of research demonstrating that individuals with depressive disorders or symptoms use tobacco at particularly high rates (Fluharty, Taylor, Grabski, & Munafo, 2017; Weinberger et al., 2017). Taken together, it is likely that elevated depressive symptoms characteristic of SM young adults experiencing minority stress may contribute to their smoking behaviors.

Another explanation for the elevated prevalence of smoking in SM populations involves targeted marketing by tobacco companies in LGB-focused spaces (Offen, Smith, & Malone, 2003; Stevens, Carlson, & Hinman, 2004; Washington, 2002). Bars and clubs that are LGB-focused have a well-documented and important role as safe gathering spaces for SM individuals (Heffernan, 1998). In a 2008 study, LGB adults reported significantly more exposure to tobacco marketing, and greater receptivity to tobacco marketing through tobacco-sponsored social events and promotional merchandise than heterosexual adults (Dilley, Spigner, Boysun, Dent, & Pizacani, 2008). Furthermore, attendance at LGB bars is associated with a greater odds of smoking compared to attending non-LGB bars (Gruskin, Byrne, Kools, & Altschuler, 2006; Matthews, Hotton, DuBois, Fingerhut, & Kuhns, 2011; Stall, Greenwood, Acree, Paul, & Coates, 1999). Longitudinal studies demonstrate that recalling tobacco marketing and receptivity to that marketing each independently predict an increased likelihood of future smoking among young adults (Gilpin, White, Messer, & Pierce, 2007). Thus, increased tobacco marketing in LGB bars may be likely to influence the smoking behaviors of SM young adults.

Another factor that may explain the elevated smoking prevalence in SM communities relates to more accepting cigarette-related social norms. Norms are a broad category of social phenomena describing what people do, or “standard behaviors” (descriptive norms), perceptions of what people “should” do, or what is socially acceptable (injunctive norms) (Cialdini, Reno, & Kallgren, 1990; Lapinski & Rimal, 2005) and personal appraisals of certain behaviors, irrespective of social standards (personal norms) (Elek, Miller-Day, & Hecht, 2006). Modeling and reinforcement of behaviors by close family members and friends are thought to influence norms (Akers & Lee, 1996). Among young adults, friend and intimate partner smoking, as well as friend and romantic partner approval of smoking each independently increase the odds of smoking (Etcheverry & Agnew, 2008). The influence of shared tobacco use behaviors in social networks (Christakis & Fowler, 2008) may be even stronger among SM groups, where tobacco-related norms are considered more accepting (Austin et al., 2004; Hatzenbuehler, Corbin, & Fromme, 2008; Trocki, Drabble, & Midanik, 2009), and where safe spaces such as LGB bars may be less restrictive about tobacco use (Gruskin et al., 2006). As such, more accepting cigarette-related norms may help explain why SM young adults are at increased risk for smoking.

Tobacco use during young adulthood is particularly concerning, given that this is the developmental period when addiction is more likely to be solidified and lifelong tobacco use is established, compared with younger and older age groups (Chen & Kandel, 1995). Furthermore, earlier smoking age of initiation is linked to greater tobacco-related morbidity and mortality including cancer and cardiovascular and pulmonary diseases, independent of sociodemographic factors and other health behaviors (Choi & Stommel, 2017). There is limited evidence that SM identity is related to excess risk for breast, anal, and lung cancer (Matthews, Breen, & Kittiteerasack, 2018), and determining why SM young adults smoke more than their heterosexual peers can ultimately inform tailored interventions aimed at reducing this tobacco-related morbidity and mortality. No studies have simultaneously tested the role of depressive symptoms, the recall of tobacco marketing in bars, and cigarette-related social norms as explanatory factors for SM tobacco use disparities. Thus, the purpose of this study was to determine the role of three mediators in the association between SM identity and subsequent cigarette use, while controlling for sociodemographic factors and tobacco use. Longitudinal structural equation modeling was used to explore if the association between SM identity and past 30-day cigarette smoking one year later was mediated by the three risk factors, each of which was assessed six months after SM identity (see hypothesized model in Figure 1). Greater depressive symptoms, recalling more tobacco marketing in bars, and more accepting cigarette-related norms were hypothesized to each explain a unique portion of the association between SM identity and subsequent cigarette use. Previous research indicates that SM women may be at particularly high risk for tobacco use (Emory et al., 2015; Fallin, Goodin, Lee, & Bennett, 2015; Johnson et al., 2016; Matthews et al., 2011). Thus, multi-group analyses were used to test if the effects of these potential mediators were consistent across sex.

Figure 1.

Figure 1

Open in a new tab

Hypothesized Mediation Model

a Wave 3 covariates include race/ethnicity (non-Hispanic White vs. non-White), attendance at a 2-year or 4-year college/university, age, and any tobacco use.

b Wave 4 covariate is any tobacco use.

Methods

Participants

Participants were 3,972 young adults who took part in waves 3, 4, and 5 of the Marketing and Promotions across Colleges in Texas project (Project M-PACT). Project M-PACT assesses the tobacco use behaviors of a cohort of young adults, initially 18 to 29 years old at baseline (fall 2014), attending one of 24 4-year universities or 2-year colleges in the five largest metropolitan areas of Texas (Austin, Dallas/Ft. Worth, Houston, and San Antonio). Wave 3 was conducted in October/November of 2015, and marked the first wave in which the question regarding sexual orientation was asked. Waves 4 and 5 were each conducted six months thereafter, in April/May and October/November 2016, respectively. These 3,972 participants were 35.4% male, 88.7% heterosexual/straight, 35.7% non-Hispanic White, 30.6% Hispanic, 18.5% Asian, 7.7% African American/Black, and 7.5% other or multiple ethnicities, and were approximately 21.8 years old (SD=2.3) at wave 3.

Since the focus of this study was on sexual (not gender) minority identity, participants who self-identified as transgender (n=11) were excluded. Similarly, participants who identified as “other” (n=77) were excluded, since it is unclear whether “other” meant the participant had a sexual or gender minority identity (or both). Additionally, participants missing predictor variables (age, race/ethnicity, school type) (n=107) and those missing every variable except predictors were dropped (n=161). Attrition analyses indicated that the 356 excluded participants were more likely to be past 30-day smokers than the larger wave 3 Project M-PACT sample, but did not differ in sex, race/ethnicity, age, or school type.

Procedure

Prior to conducting any study procedures, the Institutional Review Board (IRB) at the university leading the study approved all study procedures. To participate in Project M-PACT, participants were required to be 18 to 29 years old at baseline, and full- or part-time, degree- or certificate-seeking undergraduate students attending a participating university/college. Eligible students responded to email invitations, which included a hyperlink that brought them to the informed consent page. After reading and agreeing to the study terms, participants indicated that they understood the study and provided informed consent, then completed the baseline online survey in fall 2014. More than 13,000 students were eligible to participate in the study and, of these, 5,482 students (40%) participated in wave 1. Over 79% of the original cohort participated in wave 3. Additional information regarding study procedures is published elsewhere (Bandiera, Loukas, Wilkinson, & Perry, 2016; Loukas et al., 2016).

Measures

Wave 3 Independent Variable.

SM identity was ascertained by asking “Do you consider yourself to be…?” with mutually exclusive answer choices of a) “heterosexual or straight,” b) “gay or lesbian,” c) “bisexual,” d) “queer,” e) “transgender,” or f) “other.” Those who identified as gay, lesbian, bisexual, or queer were included in the sexual minority identity category (coded 1), while those who chose heterosexual/straight comprised the comparison group (coded 0).

Wave 3 and 4 Covariates.

Four wave 3 covariates and one wave 4 covariate were included in the model. First was race/ethnicity, represented by a dichotomous variable with non-Hispanic White coded as 1, and non-White or multiple races/ethnicities coded as 0. School type was also dichotomous, with attendance at a 4-year college/university coded as 1, and 2-year college attendance coded as 0. Age was treated as a continuous variable. Any tobacco use at wave 3 and wave 4 were included to control for potential confounding between these variables, the proposed mediators, and the outcome of wave 5 cigarette smoking. At waves 3 and 4, any participant who responded that they had used, on any day in the past 30 days, cigarettes, e-cigarettes, cigars (including large cigars, cigarillos, or little filtered cigars), hookah, or smokeless tobacco (including dip, snuff, chew, or snus) was coded as 1, and all others as 0.

Wave 4 Mediators.

Depressive symptoms.

Evidence of depressive symptoms was treated as a latent factor, comprised of three indicator variables: depressed affect, somatic retardation, and positive affect. Each of the indicator variables was derived from questions from the 10-item Center for Epidemiologic Studies Depression 10 Scale (CES-D-10). The CES-D-10 assesses the frequency of symptoms occurring in the past week on a 4-point scale from 0 “rarely (less than 1 day)” to 3 “most of the time (5–7 days)” (Andresen, Malmgren, Carter, & Patrick, 1994), where higher scores reflect higher levels of depressive symptoms. The depressed affect indictor variable was an average of four items asking about feelings like being “bothered” and “fearful.” The somatic retardation indicator variable was an average of four items inquiring about experiences like “trouble focusing” or “everything requiring effort.” The positive affect indicator variable was an average of two reverse-coded items inquiring about feelings of being “hopeful about the future” and “happy,” with higher scores indicating lower positive affect (higher depressive symptoms). The CES-D-10 has acceptable reliability and validity, and is an appropriate screener for depression among older adults (Irwin, Artin, & Oxman, 1999; O’Halloran, Kenny, & King-Kallimanis, 2014), and psychiatric samples (Bjorgvinsson, Kertz, Bigda-Peyton, McCoy, & Aderka, 2013). In the present study, internal consistency reliability (alpha coefficients) for depressed affect, somatic retardation, and positive affect were 0.77, 0.74, and 0.74, respectively.

Recall of tobacco marketing.

The recall of tobacco marketing in bars was treated as a latent factor indicated by three items, each scored on a 4-point Likert-type scale ranging from zero to three (never, rarely, sometimes, and frequently). Higher scores indicated seeing more tobacco marketing in bars/clubs. Participants who reported attending bars in the past six months were asked how often they: received free samples of tobacco/e-cigarette products, saw tobacco/e-cigarette product advertisements, or saw/interacted with tobacco/e-cigarette representatives at a bars/clubs. Non bar-goers were assigned 0 (“never”) for all three marketing recall items.

Cigarette-related social norms.

Cigarette-related social norms was treated as a latent factor indicated by three items that assessed injunctive, descriptive, and personal norms regarding cigarettes. Higher scores indicated more accepting cigarette-related norms. Each item was scored on a 5-point Likert-type scale ranging from one to five. The first item asked about cigarettes’ “social acceptability,” with answer options ranging from “totally unacceptable” to “totally acceptable.” Next, participants were asked their number of close friends who smoked cigarettes, with answer options of “none,” “a few,” “some,” “most,” or “all.” The last item stated, “I would date someone who smokes cigarettes,” with five responses ranging from “strongly disagree” to “strongly agree.”

Wave 5 Dependent Variable

Past 30-Day Cigarette Use.

Participants were asked, “On how many of the past 30 days did you smoke cigarettes?” Those who smoked on one or more days were coded 1 while those who reported smoking on zero days were coded 0.

Data Analysis

Structural equation modeling and MPLUS version 7.4 (Muthen & Muthen, 1998–2017) were used to examine the role of depressive symptoms, recalling tobacco marketing in bars, and cigarette-related norms as potential mediators in the association between SM identity and cigarette use one year later. A number of fit indices were used to assess the fit of data to the hypothesized model. A non-significant chi-square (p>.05) indicates good fit, though this test is sensitive to large sample sizes (Siddiqui, 2013). Additionally, the root-mean square error of approximation (RMSEA) was used a measure of absolute fit, with desired values of ≤.08 (Bentler & Yuan, 1999) as was the Comparative Fit Index (CFI), which assesses the improvement of the hypothesized model over a “null” model, where values >.90 are considered acceptable (Bentler, 1990). The robust weighted least squares (WLSMV) estimator was used because predictor variables were continuous and categorical, and the outcome variable was dichotomous. This estimator uses all available data in converging on a solution (Muthén, Muthén, & Asparouhov, 2015).

First, the fit of the measurement model was tested, comprised of three latent factors (depressive symptoms, recalling tobacco marketing in bars, and cigarette-related norms) and their nine indicator variables. Next, the structural model tested the associations between SM identity, the three latent mediating variables measured six months later, and past 30-day cigarette use observed one year later, while including a direct path from SM identity to the outcome variable. The model controlled for the wave 3 covariates (race/ethnicity, school type, age and any tobacco use) by including paths from each to the mediating variables and to the outcome variable (see Figure 1).

Multi-group models were used to determine if associations between variables were equivalent across females and males. To do so, first a model was fit where the seven paths between variables could freely vary across the sex groups (i.e., a fully unconstrained model). Next, a completely constrained model was fit, where all seven paths were constrained to be equal (invariant) across sex. The fit of the constrained model was compared to that of the unconstrained model. A significant chi square in this comparison indicates a deleterious effect on model fit, meaning one or more paths differ significantly between females and males and should not be constrained equal. In this case, each path is examined individually to determine the source of invariance. That is, a series of comparisons are made where only one path at a time is constrained to be equal across sex, and results from this model are compared to the fully unconstrained model. Significant chi square tests in these comparisons identify paths that differ across sex.

Results

Descriptive Analyses

Preliminary analyses were conducted to examine differences between SM and heterosexual young adults on the study variables of interest. As displayed in Table 1, significantly more SM females than heterosexual females used tobacco and smoked in waves 3, 4, and 5. Significantly more SM males than heterosexual males used any tobacco in wave 4 and smoked in wave 5. All SM participants had significantly higher scores on the wave 4 depressive symptoms subscales than their heterosexual peers, with the exception of low positive affect for males, which approached significance at p=.054. It is also important to note that both female and male SM participants reported, on average, a CES-D-10 score higher than 10, which reflects elevated risk for major depressive disorder (SM females: m=10.86; heterosexual females: m=8.25; SM males: m=10.28; heterosexual males: m=7.42, data not shown) (Irwin et al., 1999). There were no differences between SM and heterosexual participants in their wave 4 recall of tobacco marketing at bars. All SM participants had significantly more accepting wave 4 cigarette-related norms than their heterosexual peers, with the exception of social acceptability of cigarettes among males. Table 2 displays zero-order correlations, means, and standard deviations for all study variables.

Table 1:

Differences Between SM and Heterosexual Females and Males on Main Study Variablesa (N=3,972)

SM
Females
(n=287)		 Heterosexual
Females
(n=2,276)		 Significance SM
Males
(n=163)		 Heterosexual
Males
(n=1,246)		 Significance
Wave 3 Past 30-day any tobacco use (%) 45.6% 23.2% X2 (1, N=2,560)=66.98** 44.7% 37.1% X2 (1, N=1,405)=3.48
Wave 4 Past 30-day any tobacco use (%) 41.98% 23.8% X2 (1, N=2,455)=42.00** 46.4% 34.8% X2 (1, N=1,355)=7.770*
Wave 4 Depressive symptoms latent variables
   -  Depressed affect, m (SD) 0.90 (0.74) 0.62 (0.64) t(−6.76)=2,472** 0.85 (0.73) 0.52 (0.61) t(−6.09)=1,362**
   -  Low positive affect, m (SD) 1.33 (0.89) 1.09 (0.89) t(−4.29)=2,472** 1.27 (0.93) 1.12 (0.93) t(−1.93)=1,362
   -  Somatic retardation, m (SD) 1.15 (0.83) 0.89 (0.71) t(−5.60)=2,472** 1.09 (0.78) 0.78 (0.67) t(−5.27)=1,362**
Wave 4 Marketing recall latent variables
   -  Saw or received free samples, m (SD) 0.13 (0.40) 0.11 (0.42) t(−0.61)=2,465 0.19 (0.52) 0.13 (0.44) t(−1.48)=1,360
   -  Saw branded ads and merchandise, m (SD) 0.34 (0.66) 0.29 (0.65) t(−1.07)=2,465 0.30 (0.62) 0.28 (0.64) t(−0.47)=1,360
   -  Saw a tobacco/ENDS representative, m (SD) 0.17 (0.48) 0.13 (0.46) t(−1.08)=2,465 0.21 (0.54) 0.14 (0.47) t(−1.71)=1,360
Wave 4 cigarette social norms latent variable
   -  Social acceptability, m (SD) 2.94 (1.22) 2.75 (1.32) t(−2.28)=2,472* 2.84 (1.24) 2.76 (1.31) t(−0.73)=1,362
   -  Number of friends who use, m (SD) 2.23 (0.96) 1.83 (0.88) t(−7.01)=2,473** 2.22 (0.90) 1.98 (0.93) t(−3.03)=1,362*
   -  Would date, m (SD) 2.37 (1.26) 1.71 (1.02) t(−9.70)=2,472** 2.35 (1.16) 1.89 (1.10) t(−4.90)=1,362**
Wave 5 Past 30-day cigarette use (%) 27.0% 11.5% X2 (1, N=2,439)=50.71** 32.2% 19.2% X2 (1, N=1,319)=13.43**
Open in a new tab
a

Cell sizes ranged from 2,439 to 2,560 for female participants and 1,319 to 1,405 male participants, reflecting changing participation rates from wave to wave and within-wave missing data.

**

p<.001

*

p<.05

Table 2:

Zero-order Correlations, Means, and Standard Deviations for Main Study Variablesa (N=3,972)

Variable 1 2 3 4 5 6 7 8 9 10 11
1. W3b Sexual minority identity 1.00 0.16** 0.14** 0.05 0.02 0.13** 0.08** 0.04 0.05 0.01 0.10**
2. W4 Depressed affect 0.14** 1.00 0.72** 0.29** 0.04 0.13** 0.13** 0.06* 0.03 0.03 0.07*
3. W4 Somatic retardation 0.11** 0.71** 1.00 0.14** 0.03 0.13** 0.12** 0.08** 0.04 0.08** 0.07*
4. W4 Low positive affect 0.09** 0.37** 0.26** 1.00 −0.01 0.04 −0.04 −0.02 −0.02 −0.07** 0.02
5. W4 Social acceptability 0.05* 0.07** 0.08** −0.03 1.00 0.35** 0.34** 0.08** 0.12** 0.10** 0.23**
6. W4 Willingness to date 0.19** 0.13** 0.14** 0.04 0.36** 1.00 0.40** 0.13** 0.13** 0.09** 0.46**
7. W4 Number of friends who smoke 0.14** 0.13** 0.17** −0.01 0.36** 0.51** 1.00 0.21** 0.21** 0.23** 0.39**
8. W4 Recall free samples in bars 0.01 0.04* 0.07** −0.03 0.07** 0.13** 0.19** 1.00 0.60** 0.52** 0.12**
9. W4 Recall industry representatives 0.02 0.01 0.06** −0.04 0.08** 0.15** 0.21** 0.57** 1.00 0.52** 0.14**
10. W4 Recall branded ads in bars 0.02 0.06** 0.09** −0.04* 0.10** 0.13** 0.20** 0.45** 0.47** 1.00 0.12**
11. W5 Past 30-day cigarette use 0.14** 0.05* 0.06** 0.02 0.18** 0.45** 0.38** 0.16** 0.17** 0.14** 1.00
Female Mean 0.11 0.65 0.92 1.11 2.77 1.79 1.88 0.12 0.14 0.30 0.13
(Female Standard Deviation) (0.32) (0.66) (0.73) (0.89) (1.31) (1.07) (0.90) (0.42) (0.47) (0.65) (0.34)
Male Mean 0.12 0.56 0.81 1.13 2.77 1.94 2.00 0.14 0.15 0.28 0.21
(Male Standard Deviation) (0.32) (0.63) (0.69) (0.93) (1.30) (1.12) (0.93) (0.45) (0.48) (0.64) (0.41)
Open in a new tab
a

Zero-order correlations for female participants are below the diagonal; male participants are above the diagonal in shaded boxes. Cell sizes ranged from 2,348 to 2,560 for female participants and 1,276 to 1,405 male participants, reflecting changing participation rates from wave to wave and within-wave missing data.

b

“W3” indicates a wave 3 variable, W4 indicates wave 4, etc.

**

p<.01 (2-tailed)

*

p<.05 (2-tailed)

Measurement Model

The measurement model had good fit (see Figure 2). Although the chi square was significant, X2(22, N=4,357)=78.61, p≤.001, the RMSEA was 0.024 [0.019, 0.030] and the CFI was 0.994. Standardized factor loadings for each of the indicator variables exhibited significance at p<.001. The lowest factor loading (0.223) existed where low positive affect loaded on the latent variable of depressive symptoms. The factor loading for low positive affect was lower than desirable (Siddiqui, 2013), thus, we tested a second model with low positive affect removed. In the revised measurement model, chi square difference testing indicated overall fit was slightly improved: RMSEA=0.018 [0.011, 0.025], CFI=0.997, p<.001. We chose to use the original measurement model in order to maintain the originally conceptualized structure of the depression latent factor, and to use as many indicator variables as we had available in the survey.

Figure 2.

Figure 2

Open in a new tab

Final Measurement Modela

a Standardized path coefficients are reported in the figure. Model fit was good: X2(22, N=4,357)=78.61, p<.001, RMSEA=.024 [.019-.030], CFI=.994.

All paths significant at p<.001.

Structural Model

Fitting the hypothesized structural model (without separation by sex) yielded adequate fit: X2(68, N=3,974)=437.284, p<.001, RMSEA=0.037 [0.034, 0.040], CFI=0.955. Examination of the multi-group model allowed determination equivalence across sex. The omnibus test for a decrease in overall model fit (DIFFTEST) indicated no significant difference between the fully unconstrained and fully constrained models X2(155, N=3,972)=491.87, p=<.001). However, to ensure individual path differences were not obscured by the overall test, each of the seven individual paths was successively fixed to be invariant across sex, and then compared to the fully unconstrained model. Findings indicated that paths from wave 4 cigarette-related norms to wave 5 cigarette smoking should be allowed to vary across sex. Thus, the final structural model in our study constrained six paths invariant across sex, while allowing the path from wave 4 cigarette norms to wave 5 smoking to freely vary. This final model fit was excellent: X2(133, N=3,972)=490.38, p=<.001, RMSEA=.033 [.030-.036], CFI=.958. Moreover, this model explained 57.4% and 62.0% of the variance in past 30-day smoking for females and males, respectively.

As shown in Figure 3, for both females and males, wave 3 SM identity was positively associated with wave 4 depressive symptoms. Unexpectedly, wave 4 depressive symptoms were negatively associated with wave 5 past 30-day cigarette smoking for both sexes. Prior to testing the significance of the indirect effects, additional analyses were conducted to determine the cause of the unexpected negative association between depressive symptoms and cigarette use. In a simple regression of wave 5 smoking on the wave 4 depressive symptoms latent factor, higher depression was positively associated with smoking (β=4.70, SE=0.40, Z=11.79, p=0.00, R2=0.66). However, when the latent factor of cigarette-related norms was added to the model, the overall explanatory power decreased, and the association between depression and wave 5 smoking changed to negative: β=−0.117, SE=0.047, Z=−2.51, p=.012, R2=.545. In a separate model, wave 5 smoking was regressed onto wave 4 cigarette-related norms. More accepting norms were positively and significantly associated with smoking (β=1.28, SE=0.075, Z=17.06, p=.000, R2=0.523). Adding the latent factor of depressive symptoms to this latter model improved the explanatory power and increased social norms’ coefficient: β=1.33, SE=0.078, Z=17.09, p=.000, R2=0.545, although the association between depression and cigarette use was once again negative. Given these findings, the negative association may be caused by a “cross-over suppression” effect, which occurs when the introduction of a predictor variable (i.e., social norms) reverses the sign of the initial predictor (i.e., depressive symptoms) (Paulhus, Robins, Trzesniewski, & Tracy, 2004). As this finding was unexpected and requires additional research, the role of depressive symptoms as a mediator of the association between SM identity and later cigarette use was not explored further. The roles of the covariates on the mediating and outcome variables are displayed in Table 3.

Figure 3.

Figure 3

Open in a new tab

Final Structural Mediation Model with Significant Paths Displayeda

a Model fit was good: X2(154, N=3,972)=490.38, p<.001, RMSEA=.033 [.030-.036], CFI=.958. Showing only significant standardized coefficients for female (before parens) and male (inside parens) participants. Not shown: Paths controlling for non-Hispanic White race/ethnicity, school type (2-year vs. 4-year), age, and Waves 3 and 4 any tobacco use. For standardized path values for all variables in the model (including covariates), see Table 3. *p<.05, **p<.001

Table 3:

Standardized Path Estimates for the Role of Five Covariates on the Wave 4 Latent Mediator and Wave 5 Observed Outcome Variables (N=3,972)

Females Males
W4 Depression W4 Cigarette Norms W4 Marketing Recall W5 Cigarette Smoking W4 Depression W4 Cigarette Norms W4 Marketing Recall W5 Cigarette Smoking
Wave 3 Covariates
- Past 30-day any tobacco use .082** .483** .183** .137** .072* .577** .224** .129*
- Non-Hispanic White (vs. non-White) .010 .068* .069* .044 −.049 −.018 .010 .005
- 4-year school (vs. 2-year) −.021 −.053* .020 −.021 .004 .002 .047 .027
- Age −.045* .141** .199** .052 −.042 .101** .248** −.018
Wave 4 Covariate
- Past 30-day any tobacco use .046* .290** .120** .240** .066* .338** .086* .198**
Open in a new tab
*

p<.05

**

p<.001

For both females and males, wave 3 SM identity was not a significant predictor of wave 4 recalling tobacco marketing in bars, nor was wave 4 marketing recall associated with wave 5 past 30-day smoking. Finally, wave 3 SM identity predicted higher wave 4 cigarette-related norms for both females and males. In turn, wave 4 cigarette-related norms was significantly predictive of wave 5 past 30-day cigarette smoking for both sexes, though it was higher for males. Only cigarette-related norms was considered a potential mediator in the model, as it was associated with both predictor and outcome variables. Thus, the significance of this indirect effect was tested. The standardized indirect effect of SM identity on subsequent cigarette smoking through social norms was significant for females (β=0.062, p<.001), and for males (β=0.075, p<.001).

Discussion

The present study extends the growing body of SM-focused research by examining factors that may explain the elevated smoking prevalence among SM young adults. Consistent with previous research, all SM participants in this study reported higher scores than their heterosexual peers on almost all study variables, including a higher prevalence of cigarette use (Jamal et al., 2018; Rath et al., 2013), concerning levels of depressive symptoms (Fergusson et al., 2005; Hatzenbuehler, 2009; McKirnan, Tolou-Shams, Turner, Dyslin, & Hope, 2006), and more accepting cigarette-related social norms (Hamilton & Mahalik, 2009; Matthews et al., 2011) than their heterosexual peers. Findings also indicated that only social norms clearly explained the disparity in cigarette smoking for SM young adults.

More accepting cigarette-related social norms explained a significant portion of the association between SM identity and past 30-day cigarette use one year later for both females and males. This mediating effect existed while simultaneously accounting for tobacco use, depressive symptoms, recalling tobacco marketing in bars, and socio-demographic factors that influence tobacco use. It is possible that groups who are not afforded mainstream social status adopt norms unique to their own culture, as a way of rejecting the larger status quo of a society that stigmatizes their existence (d’Emilio, 1983). According to social learning theory, groups that are similar to an individual are viewed as the most influential (Hornstein, Fisch, & Holmes, 1968), and it is possible that this study is tapping into the close social environment of LGB communities and its norms. Indeed, research among SM women indicates that those who are more immersed in the LGB culture may be more susceptible to adopting that culture’s norms and attitudes surrounding cigarettes (Matthews et al., 2011). Given the explanatory role of social norms in the association between SM identity and subsequent cigarette use, more research is needed to determine exactly why these norms are more accepting, and whether these norms differ across subgroups of other identities in the LGBTQ umbrella, such as transgender young adults.

Contrary to study hypotheses, SM identity was not associated with recalling tobacco marketing in bars, which in turn was not associated with subsequent past 30-day cigarette smoking. Marketing has a demonstrated effect on influencing tobacco-related norms (Gunther, Bolt, Borzekowski, Liebhart, & Dillard, 2006; Wakefield, Flay, Nichter, & Giovino, 2003), and as such perhaps its effect on smoking is not direct, but rather indirect through social norms. It is also possible that SM participants were not attending LGB bars, and thus were not the recipients of the targeted tobacco marketing that is characteristic of these spaces (Dilley et al., 2008). Additionally, it may be that the bar marketing environment in the large metropolitan areas of our study is relatively homogenous, and the tobacco marketing environments do not differ between heterosexual and LGB bars. Finally, the marketing recall assessed related to general tobacco marketing, and was not specific to cigarette use, which was our outcome of interest. Future studies that objectively assess tobacco advertising in LGB bars will help researchers understand if targeted tobacco marketing in these spaces persists, and if that marketing influences the tobacco use behaviors of its bar patrons.

Finally, the negative association between depressive symptoms and subsequent smoking behavior for SM young adults warrants further attention. We hypothesized a positive association between the depressive symptoms mediator and subsequent smoking, such that more severe symptoms would be associated with a greater likelihood of smoking. Instead, there was a negative direct association between depressive symptoms and smoking when included alongside the tobacco marketing and social norms mediators. One possibility is the suppressor effect, which reverses the sign of the initial predictor (depressive symptoms mediator) in the presence of another predictor variable (social norms) (Paulhus et al., 2004). Another possibility for the negative association may be due to the interrelatedness of depressive symptoms, social norms, and smoking behavior. In the present study, social norms were measured in part by peer use of cigarettes, and prior research focused on adolescent substance use indicates that peer use is a mediator of the association between depressive symptoms and substance use. These studies suggest that when peer use is taken into account, elevated depressive symptoms can lead to social withdrawal or avoidance, and in turn to less substance use behaviors (Colder et al., 2017). As an example, Audrain-McGovern, Rodriguez, and Kassel found that peer smoking completely mediated the association between depression and cigarette smoking, while the direct association between depression and smoking level was negative and approached significance (Audrain-McGovern, Rodriguez, & Kassel, 2009). Given the findings from the present study and prior research on adolescents, additional research examining the nature of the associations between depressive symptoms, social norms, peer use, and cigarette use among young adults is warranted.

Limitations

While this study helps explain the association between SM identity and cigarette use, it does have limitations. First, there were no specific measures of minority stress; rather, depressive symptoms were used as a proxy. Future studies should assess specific constructs from Meyer’s minority stress model (such as measures of “gay-related stress”) to allow a more in-depth exploration of the association between SM identity and tobacco use. Additionally, exploring the role of depressive symptoms and other internalizing problems in the smoking behaviors of SM young adults deserves its own consideration, not merely as a proxy for minority stress. Additionally, while the outcome of interest was cigarette use, control variables at wave 3 and 4 included any past 30-day tobacco use (cigarettes, e-cigarettes, cigars, hookah, and/or smokeless tobacco). Given the wide range of products on the market and the potential for e-cigarettes and other emerging products to renormalize smoking (Loukas, Marti, Cooper, Pasch, & Perry, 2018; Voigt, 2015), future studies are needed that disentangle the influence of different types of tobacco product use on subsequent cigarette use. The present study would be strengthened by a larger, more diverse sample and improved survey measures that account for identities outside of lesbian, gay, bisexual, and queer (though the last is viewed as ambiguous and not always a sexual identity), as well as gender-diverse identities. A larger sample would also enhance researchers’ ability to examine intragroup variation in the wide range of sexual and gender minority identities that was not possible in the current study. This paper groups lesbian, bisexual, and queer-identified females together, which may mask important differences, especially in light of the evidence that bisexual women face disproportionate risk for tobacco use compared to their peers. A larger sample would also allow consideration of other important differences across subgroup, such racial/ethnic differences. Lastly, though biological sex was used to examine differences between males and females, it is essential to note that sex is not synonymous with gender. Thus, to characterize study findings in terms of men and women is not completely accurate, as gender identity does not always correspond to biological sex assignment.

Conclusion

A strength of this longitudinal study is that it includes a large, diverse sample of metropolitan area college-attending young adults from the second most populated state in the U.S. It adds to the existing research by showing that more accepting cigarette-related social norms explain, at least in part, the elevated smoking of SM young adults, while accounting for their depressive symptoms and recall of exposure to marketing at bars and clubs. This study also explores potential differences by sex, though we found no differences in the associations between SM identity, depressive symptoms, marketing recall in bars, and cigarette norms with subsequent smoking variables across SM females and males. Members of the SM community still experience prejudicial interactions and legally sanctioned discrimination related to their sexual orientation (Sears & Mallory, 2011), thus it is plausible that these experiences contribute to stressors that increase the prevalence of smoking behaviors, while also reinforcing culturally specific norms surrounding behaviors that help individuals navigate stress related to their SM identity. These findings lend support to initiatives that focus on tailoring tobacco prevention and cessation messages for SM individuals, with particular emphasis on shifting tobacco-related social norms, in order to reduce the tobacco-related morbidity and mortality for this high-risk subpopulation. Although SM-focused smoking interventions among young adults are relatively new to public health (Baskerville et al., 2018), it is encouraging that campaigns such as the Food and Drug Administration’s This Free Life (U.S. Food and Drug Administration, 2018) use tailored, inclusive messages to target this high-risk group. Future studies should explore the temporality and interrelatedness of norms, depressive symptoms, and the wide range of contemporary marketing strategies, as well as other or new risk factors, to lessen the tobacco burden for this high-risk group of young adults.

Acknowledgments

Research reported in this publication was supported by grant number [1 P50 CA180906] from the National Cancer Institute (NCI) and the Food and Drug Administration (FDA) Center for Tobacco Products (CTP). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the FDA.

Footnotes

Initial cross-sectional findings from these analyses were presented in poster form at the 2018 national meeting of the Society for Research on Nicotine & Tobacco (SRNT). Since then, analyses were changed to a longitudinal design and have not been presented elsewhere.

Contributor Information

Josephine T. Hinds, Department of Kinesiology and Health Education, The University of Texas at Austin

Alexandra Loukas, Department of Kinesiology and Health Education, The University of Texas at Austin.

Cheryl L. Perry, UTHealth School of Public Health

References

  1. Agaku IT, King BA, Husten CG, Bunnell R, Ambrose BK, Hu SS, … Prevention. (2014). Tobacco product use among adults--United States, 2012–2013. MMWR Morb Mortal Wkly Rep, 63(25), 542–547. [PMC free article] [PubMed] [Google Scholar]
  2. Akers RL, & Lee G (1996). A longitudinal test of social learning theory: Adolescent smoking. Journal of Drug Issues, 26(2), 317–343. [Google Scholar]
  3. Andresen EM, Malmgren JA, Carter WB, & Patrick DL (1994). Screening for depression in well older adults: Evaluation of a short form of the CES-D. American Journal of Preventive Medicine, 10(2), 77–84. [PubMed] [Google Scholar]
  4. Audrain-McGovern J, Rodriguez D, & Kassel JD (2009). Adolescent smoking and depression: evidence for self-medication and peer smoking mediation. Addiction, 104(10), 1743–1756. doi: 10.1111/j.1360-0443.2009.02617.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Austin SB, Ziyadeh N, Fisher LB, Kahn JA, Colditz GA, & Frazier AL (2004). Sexual orientation and tobacco use in a cohort study of US adolescent girls and boys. Archives of Pediatrics and Adolescent Medicine, 158(4), 317–322. [DOI] [PubMed] [Google Scholar]
  6. Bandiera FC, Loukas A, Wilkinson AV, & Perry CL (2016). Associations between tobacco and nicotine product use and depressive symptoms among college students in Texas. Addictive Behaviors, 63, 19–22. doi: 10.1016/j.addbeh.2016.06.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Baskerville NB, Wong K, Shuh A, Abramowicz A, Dash D, Esmail A, & Kennedy R (2018). A qualitative study of tobacco interventions for LGBTQ+ youth and young adults: overarching themes and key learnings. BMC Public Health, 18(1), 155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bentler PM (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107(2), 238–246. [DOI] [PubMed] [Google Scholar]
  9. Bentler PM, & Yuan KH (1999). Structural Equation Modeling with Small Samples: Test Statistics. Multivariate Behavioral Research, 34(2), 181–197. doi: 10.1207/S15327906Mb340203 [DOI] [PubMed] [Google Scholar]
  10. Bjorgvinsson T, Kertz SJ, Bigda-Peyton JS, McCoy KL, & Aderka IM (2013). Psychometric properties of the CES-D-10 in a psychiatric sample. Assessment, 20(4), 429–436. doi: 10.1177/1073191113481998 [DOI] [PubMed] [Google Scholar]
  11. Chen K, & Kandel DB (1995). The natural history of drug use from adolescence to the mid-thirties in a general population sample. American Journal of Public Health, 85(1), 41–47. doi: 10.2105/AJPH.85.1.41 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Choi SH, & Stommel M (2017). Impact of Age at Smoking Initiation on Smoking-Related Morbidity and All-Cause Mortality. American Journal of Preventive Medicine, 53(1), 33–41. doi: 10.1016/j.amepre.2016.12.009 [DOI] [PubMed] [Google Scholar]
  13. Christakis NA, & Fowler JH (2008). The collective dynamics of smoking in a large social network. N Engl J Med, 358(21), 2249–2258. doi: 10.1056/NEJMsa0706154 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Cialdini RB, Reno RR, & Kallgren CA (1990). A focus theory of normative conduct: recycling the concept of norms to reduce littering in public places. Journal of Personality and Social Psychology, 58(6), 1015. [Google Scholar]
  15. Colder CR, Shyhalla K, Frndak S, Read JP, Lengua LJ, Hawk LW Jr., & Wieczorek WF (2017). The Prospective Association Between Internalizing Symptoms and Adolescent Alcohol Involvement and the Moderating Role of Age and Externalizing Symptoms. Alcohol Clin Exp Res, 41(12), 2185–2196. doi: 10.1111/acer.13512 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Corliss HL, Wadler BM, Jun HJ, Rosario M, Wypij D, Frazier AL, & Austin SB (2013). Sexual-orientation disparities in cigarette smoking in a longitudinal cohort study of adolescents. Nicotine Tob Res, 15(1), 213–222. doi: 10.1093/ntr/nts114 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. d’Emilio J (1983). Capitalism and gay identity. Families in the US: Kinship and domestic politics, 131–141.
  18. Dilley JA, Spigner C, Boysun MJ, Dent CW, & Pizacani BA (2008). Does tobacco industry marketing excessively impact lesbian, gay and bisexual communities? Tobacco Control, 17(6), 385–390. doi: 10.1136/tc.2007.024216 [DOI] [PubMed] [Google Scholar]
  19. Elek E, Miller-Day M, & Hecht ML (2006). Influences of personal, injunctive, and descriptive norms on early adolescent substance use. Journal of Drug Issues, 36(1), 147–172. [Google Scholar]
  20. Emory K, Kim Y, Buchting F, Vera L, Huang J, & Emery SL (2015). Intragroup Variance in Lesbian, Gay, and Bisexual Tobacco Use Behaviors: Evidence That Subgroups Matter, Notably Bisexual Women. Nicotine & Tobacco Research doi: 10.1093/ntr/ntv208 [DOI] [PMC free article] [PubMed]
  21. Etcheverry PE, & Agnew CR (2008). Romantic partner and friend influences on young adult cigarette smoking: comparing close others’ smoking and injunctive norms over time. Psychology of Addictive Behaviors, 22(3), 313–325. doi: 10.1037/0893-164X.22.3.313 [DOI] [PubMed] [Google Scholar]
  22. Fallin A, Goodin A, Lee YO, & Bennett K (2015). Smoking characteristics among lesbian, gay, and bisexual adults. Preventive Medicine, 74, 123–130. doi: 10.1016/j.ypmed.2014.11.026 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Fergusson DM, Horwood LJ, Ridder EM, & Beautrais AL (2005). Sexual orientation and mental health in a birth cohort of young adults. Psychological Medicine, 35(7), 971–981. [DOI] [PubMed] [Google Scholar]
  24. Fluharty M, Taylor AE, Grabski M, & Munafo MR (2017). The Association of Cigarette Smoking With Depression and Anxiety: A Systematic Review. Nicotine Tob Res, 19(1), 3–13. doi: 10.1093/ntr/ntw140 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Gilpin EA, White MM, Messer K, & Pierce JP (2007). Receptivity to tobacco advertising and promotions among young adolescents as a predictor of established smoking in young adulthood. American Journal of Public Health, 97(8), 1489–1495. doi: 10.2105/AJPH.2005.070359 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Gruskin E, Byrne K, Kools S, & Altschuler A (2006). Consequences of frequenting the lesbian bar. Women’s Health, 44(2), 103–120. [DOI] [PubMed] [Google Scholar]
  27. Gunther AC, Bolt D, Borzekowski DLG, Liebhart JL, & Dillard JP (2006). Presumed influence on peer norms: How mass media indirectly affect adolescent smoking. Journal of Communication, 56(1), 52–68. doi: 10.1111/j.1460-2466.2006.00002.x [DOI] [Google Scholar]
  28. Hamilton CJ, & Mahalik JR (2009). Minority stress, masculinity, and social norms predicting gay men’s health risk behaviors. Journal of Counseling Psychology, 56(1), 132. [Google Scholar]
  29. Hatzenbuehler ML (2009). How does sexual minority stigma “get under the skin”? A psychological mediation framework. Psychological Bulletin, 135(5), 707–730. doi: 10.1037/a0016441 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Hatzenbuehler ML, Corbin WR, & Fromme K (2008). Trajectories and determinants of alcohol use among LGB young adults and their heterosexual peers: results from a prospective study. Developmental Psychology, 44(1), 81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Heffernan K (1998). The nature and predictors of substance use among lesbians. Addictive Behaviors, 23(4), 517–528. [DOI] [PubMed] [Google Scholar]
  32. Hoffman L, Delahanty J, Johnson SE, & Zhao X (2018). Sexual and gender minority cigarette smoking disparities: An analysis of 2016 Behavioral Risk Factor Surveillance System data. Prev Med, 113, 109–115. doi: 10.1016/j.ypmed.2018.05.014 [DOI] [PubMed] [Google Scholar]
  33. Hornstein HA, Fisch E, & Holmes M (1968). Influence of a model’s feeling about his behavior and his relevance as a comparison other on observers’ helping behavior. Journal of Personality and Social Psychology, 10(3), 222. [Google Scholar]
  34. Hu SS, Neff L, Agaku IT, Cox S, Day HR, Holder-Hayes E, & King BA (2016). Tobacco Product Use Among Adults - United States, 2013–2014. MMWR Morb Mortal Wkly Rep, 65(27), 685–691. doi: 10.15585/mmwr.mm6527a1 [DOI] [PubMed] [Google Scholar]
  35. Irwin M, Artin KH, & Oxman MN (1999). Screening for depression in the older adult: criterion validity of the 10-item Center for Epidemiological Studies Depression Scale (CES-D). Archives of Internal Medicine, 159(15), 1701–1704. [DOI] [PubMed] [Google Scholar]
  36. Jamal A, Phillips E, Gentzke AS, Homa DM, Babb SD, King BA, & Neff LJ (2018). Current Cigarette Smoking Among Adults - United States, 2016. Morbidity & Mortality Weekly Report, 67(2), 53–59. doi: 10.15585/mmwr.mm6702a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Johnson SE, Holder-Hayes E, Tessman GK, King BA, Alexander T, & Zhao X (2016). Tobacco Product Use Among Sexual Minority Adults: Findings From the 2012–2013 National Adult Tobacco Survey. American Journal of Preventive Medicine, 50(4), e91–e100. doi: 10.1016/j.amepre.2015.07.041 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. King M, Semlyen J, Tai SS, Killaspy H, Osborn D, Popelyuk D, & Nazareth I (2008). A systematic review of mental disorder, suicide, and deliberate self harm in lesbian, gay and bisexual people. BMC Psychiatry, 8, 70. doi: 10.1186/1471-244X-8-70 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Lapinski MK, & Rimal RN (2005). An explication of social norms. Communication theory, 15(2), 127–147. [Google Scholar]
  40. Loukas A, Chow S, Pasch KE, Li X, Hinds JT, Marti CN, … Perry CL (2016). College students’ polytobacco use, cigarette cessation, and dependence. American Journal of Health Behavior, 40(4), 514–522. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Loukas A, Marti CN, Cooper M, Pasch KE, & Perry CL (2018). Exclusive e-cigarette use predicts cigarette initiation among college students. Addictive Behaviors, 76, 343–347. doi: 10.1016/j.addbeh.2017.08.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Matthews AK, Breen E, & Kittiteerasack P (2018). Social Determinants of LGBT Cancer Health Inequities. Seminars in Oncology Nursing, 34(1), 12–20. doi: 10.1016/j.soncn.2017.11.001 [DOI] [PubMed] [Google Scholar]
  43. Matthews AK, Hotton A, DuBois S, Fingerhut D, & Kuhns LM (2011). Demographic, psychosocial, and contextual correlates of tobacco use in sexual minority women. Research in Nursing and Health, 34(2), 141–152. doi: 10.1002/nur.20427 [DOI] [PubMed] [Google Scholar]
  44. Max WB, Stark B, Sung HY, & Offen N (2016). Sexual Identity Disparities in Smoking and Secondhand Smoke Exposure in California: 2003–2013. American Journal of Public Health, 106(6), 1136–1142. doi: 10.2105/AJPH.2016.303071 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. McCabe SE, Matthews AK, Lee JGL, Veliz P, Hughes TL, & Boyd CJ (2018). Tobacco Use and Sexual Orientation in a National Cross-sectional Study: Age, Race/Ethnicity, and Sexual Identity-Attraction Differences. Am J Prev Med, 54(6), 736–745. doi: 10.1016/j.amepre.2018.03.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. McKirnan DJ, Tolou-Shams M, Turner L, Dyslin K, & Hope B (2006). Elevated risk for tobacco use among men who have sex with men is mediated by demographic and psychosocial variables. Substance use & misuse, 41(8), 1197–1208. [DOI] [PubMed] [Google Scholar]
  47. Meyer IH (2003). Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: conceptual issues and research evidence. Psychological Bulletin, 129(5), 674–697. doi: 10.1037/0033-2909.129.5.674 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Muthén B, Muthén L, & Asparouhov T (2015). Estimator choices with categorical outcomes. MPLUS Technical Appendix, 8. [Google Scholar]
  49. Muthen LK, & Muthen BO (1998–2017). Mplus User’s Guide (8th Edition ed.). Los Angeles, CA: Muthen & Muthen. [Google Scholar]
  50. Nguyen N, McQuoid J, Ramo D, Holmes LM, Ling PM, & Thrul J (2018). Real-time predictors of smoking among sexual minority and heterosexual young adults: An ecological momentary assessment study. Drug Alcohol Depend, 192, 51–58. doi: 10.1016/j.drugalcdep.2018.07.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. O’Halloran A, Kenny R, & King-Kallimanis B (2014). The latent factors of depression from the short forms of the CES-D are consistent, reliable and valid in community-living older adults. European Geriatric Medicine, 5(2), 97–102. [Google Scholar]
  52. Offen N, Smith EA, & Malone RE (2003). From adversary to target market: the ACT-UP boycott of Philip Morris. Tobacco Control, 12(2), 203–207. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Paulhus DL, Robins RW, Trzesniewski KH, & Tracy JL (2004). Two Replicable Suppressor Situations in Personality Research. Multivariate Behavioral Research, 39(2), 303–328. doi: 10.1207/s15327906mbr3902_7 [DOI] [PubMed] [Google Scholar]
  54. Phillips E, Wang TW, Husten CG, Corey CG, Apelberg BJ, Jamal A, … King BA (2017). Tobacco Product Use Among Adults - United States, 2015. MMWR Morb Mortal Wkly Rep, 66(44), 1209–1215. doi: 10.15585/mmwr.mm6644a2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Rath JM, Villanti AC, Rubenstein RA, & Vallone DM (2013). Tobacco use by sexual identity among young adults in the United States. Nicotine Tob Res, 15(11), 1822–1831. doi: 10.1093/ntr/ntt062 [DOI] [PubMed] [Google Scholar]
  56. Ross LE, Salway T, Tarasoff LA, MacKay JM, Hawkins BW, & Fehr CP (2018). Prevalence of Depression and Anxiety Among Bisexual People Compared to Gay, Lesbian, and Heterosexual Individuals:A Systematic Review and Meta-Analysis. J Sex Res, 55(4–5), 435–456. doi: 10.1080/00224499.2017.1387755 [DOI] [PubMed] [Google Scholar]
  57. Schuler MS, Rice CE, Evans-Polce RJ, & Collins RL (2018). Disparities in substance use behaviors and disorders among adult sexual minorities by age, gender, and sexual identity. Drug Alcohol Depend, 189, 139–146. doi: 10.1016/j.drugalcdep.2018.05.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Sears B, & Mallory C (2011). Documented evidence of employment discrimination & its effects on LGBT people Retrieved from https://williamsinstitute.law.ucla.edu/wp-content/uploads/Sears-Mallory-Discrimination-July-20111.pdf
  59. Siddiqui K (2013). Heuristics for sample size determination in multivariate statistical techniques. World Applied Sciences Journal, 27(2), 285–287. [Google Scholar]
  60. Stall RD, Greenwood GL, Acree M, Paul J, & Coates TJ (1999). Cigarette smoking among gay and bisexual men. American Journal of Public Health, 89(12), 1875–1878. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Stevens P, Carlson LM, & Hinman JM (2004). An analysis of tobacco industry marketing to lesbian, gay, bisexual, and transgender (LGBT) populations: strategies for mainstream tobacco control and prevention. Health Promot Pract, 5(3 Suppl), 129S–134S. doi: 10.1177/1524839904264617 [DOI] [PubMed] [Google Scholar]
  62. Trocki KF, Drabble LA, & Midanik LT (2009). Tobacco, marijuana, and sensation seeking: comparisons across gay, lesbian, bisexual, and heterosexual groups. Psychology of Addictive Behaviors, 23(4), 620–631. doi: 10.1037/a0017334 [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. U.S. Food and Drug Administration. (2018). This Free Life Campaign Retrieved from https://www.fda.gov/TobaccoProducts/PublicHealthEducation/PublicEducationCampaigns/ThisFreeLifeCampaign/default.htm
  64. U.S. Surgeon General. (2014). The health consequences of smoking—50 years of progress: a report of the surgeon general Paper presented at the US Department of Health and Human Services. [Google Scholar]
  65. Voigt K (2015). Smoking Norms and the Regulation of E-Cigarettes. American Journal of Public Health, 105(10), 1967–1972. doi: 10.2105/AJPH.2015.302764 [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Wakefield M, Flay B, Nichter M, & Giovino G (2003). Role of the media in influencing trajectories of youth smoking. Addiction, 98 Suppl 1, 79–103. [DOI] [PubMed] [Google Scholar]
  67. Washington HA (2002). Burning Love: big tobacco takes aim at LGBT youths. American Journal of Public Health, 92(7), 1086–1095. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Weinberger AH, Kashan RS, Shpigel DM, Esan H, Taha F, Lee CJ, … Goodwin RD (2017). Depression and cigarette smoking behavior: A critical review of population-based studies. Am J Drug Alcohol Abuse, 43(4), 416–431. doi: 10.3109/00952990.2016.1171327 [DOI] [PubMed] [Google Scholar]
  69. Wheldon CW, Kaufman AR, Kasza KA, & Moser RP (2018). Tobacco Use Among Adults by Sexual Orientation: Findings from the Population Assessment of Tobacco and Health Study. LGBT Health, 5(1), 33–44. doi: 10.1089/lgbt.2017.0175 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES