Nicotine & Tobacco Research, 2015, 1362–1368 doi:10.1093/ntr/ntv001 Original investigation Advance Access publication January 14, 2015
Original investigation
Smoking Among Sexual Minorities: Are There Racial Differences? Kasim S. Ortiz MS1, Dustin T. Duncan ScD2, John R. Blosnich PhD, MPH3, Ramzi G. Salloum PhD, MA, MBA4, Juan Battle PhD, MA5 Department of Sociology, Vanderbilt University, Nashville, TN; 2Department of Population Health, New York University School of Medicine, New York, NY; 3VA Center for Health Equity Research and Promotion, Pittsburgh, PA; 4 Department of Health Services Policy and Management, University of South Carolina, Columbia, SC; 5Departments of Sociology, Public Health and Urban Education, City University of New York, New York, NY
Downloaded from http://ntr.oxfordjournals.org/ at New York University on November 14, 2015
1
Corresponding Author: Kasim S. Ortiz, MS. Department of Sociology, Garland Hall, PMB 351811, Nashville, TN37235-1811, Vanderbilt University; E-mail: [email protected]
Abstract Introduction: Smoking prevalence is higher among sexual minorities compared to their heterosexual peers. However, very little is known about potential racial differences in smoking among sexual minority populations. We examined differences by race in smoking status among a robust sample of sexual minorities. Methods: We used data from the 2010 Social Justice Sexuality project, a large national convenience sample of sexual minority adults that oversampled individuals from racial minority groups. LogPoisson multivariable regression models were employed to determine the risk of current smoking among sexual minority individuals by race after controlling for socio-demographic characteristics. Results: Among smokers, 22.35% identified as White, 26.98% identified as Black, 19.38% identified as Latino/Hispanic, 5.58% identified as Asian American, and 25.67% were other/multiracial. In fully adjusted gender stratified models, Black men (adjusted risk ratio [aRR] = 0.61, 95% confidence interval [CI] = 0.50, 0.75) and Asian American men (aRR = 0.61, 95% CI = 0.50, 0.75) were at lower risk of smoking compared to White men. Black women were the only to remain statistically significant for decreased risk of smoking in fully adjusted gender stratified models (aRR = 0.78, 95 % CI = 0.65, 0.95). Conclusions: Among sexual minorities, Black and Asian American individuals consistently were at decreased risk of current smoking compared to their White peers. Future research should seek to understand the mechanisms that contribute to decreased smoking status among racial sexual minorities.
Introduction Cigarette smoking is the leading cause of preventable disease and death in the United States with approximately 443,000 attributable deaths and $193 billion in direct health-care expenditures and yearly productivity losses.1 Research consistently shows that sexual minority populations (i.e., lesbian, gay, bisexual and transgender [LGBT] individuals) have significantly higher prevalence of cigarette smoking compared with heterosexuals.2–5 For example, the National Adult Tobacco Survey (ATS) by the Centers for Disease Control and
Prevention (CDC) found 32.8% of LGBT adults indicated cigarette use compared to 19.5% of heterosexual adults.6–8 Pooling 7 years of data from the National Health and Nutrition Examination Surveys, Cochran, Bandiera, and Mays2 demonstrated that lesbian and bisexual women had higher rates of tobacco use than their heterosexual counterparts. Research has consistently revealed that racial minorities are less likely to smoke than their White counterparts in the general population regardless of sexual orientation.6,7,9,10 A majority of the research examining differences in smoking status among sexual minorities compared to their heterosexual counterparts has
© The Author 2015. Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. For permissions, please e-mail: [email protected].
1362
Nicotine & Tobacco Research, 2015, Vol. 17, No. 11
Methods
collected from over 5,500 respondents in all 50 states (including Washington, DC) and Puerto Rico from January 2010 to December 2010. The survey was fielded in both English and Spanish. Several data collection strategies were used including venue-based sampling, snowball sampling, internet-based outreach, and partnerships with community-based organizations, activists and opinion leaders. The total sample consists of 4,953 complete surveys. We focused our analyses on those who reported their smoking status (N = 3,982). Data collection for the SJS originally was approved by the institutional review board of the City University of New York.
Dependent Variable Smoking status was assessed by the following question, “Do you now smoke cigarettes?” Response options included: not at all; some days; every day. This was dichotomized to represent those that currently smoked and those that did not smoke. This question is the same question utilized within the Health Information National Trends Survey.20 This question has demonstrated strong validity in many settings and in conjunction with a myriad of diseases and outcomes,21,22 as well as among sexual minority populations.23
Independent Variable Race was assessed using the question, “Which of the following racial groups comes closest to which you identify (choose all that apply)?” Respondents had the choice of self-identifying with the following categories: Black, Latino/Hispanic, Asian American/Pacific Islander, Multiracial, Native Americans, White, and other. These categories were coded as: White; Black; Latino/Hispanic; Asian American/ Pacific Islander; and other.
Covariates We adjusted for several covariates known to be associated with smoking, including gender at birth (male, female),7 educational attainment (high school diploma/General Education Diploma (GED) or less; some college/associate’s degree; bachelor’s degree or higher),24 relationship status (partnered, single),25 and age (in years).7 We also adjusted for health insurance status (yes, no), because having health care coverage has been associated with factors related to smoking status, such as access to cessation treatments and quitting smoking.26 Income is also strongly associated with smoking status,10 and in the present study, household income was gathered using a 12-category ordinal response categorization (i.e., under $8,500; $8,500–10,999; $11,000–13,499; $13,500–14,999; $15,000–17,499; $17,500– 19,999; $20,000–29,999; $30,000–39,999; $40,000–49,999; $50,000–74,999; $75,000–99,999; $100,000 and over). The ordinal categories were transformed into distinct categories ranging from 1–12 corresponding with the original ordinal categories. Thus, a household income with a mean of 4 would correspond with the range of $13,500–14,999.
Data
Analytic Strategy
Data are from the 2010 SJS Project, which is a convenience sample of collected data on the experiences of sexual minority people of color concerning five themes: identity (both racial and sexual), physical and mental health, family, religion and spirituality, and sociopolitical involvement. Data collection efforts were employed to oversample racial minority individuals15–19 who identify as lesbian, gay, bisexual or transgender for which self-identification was utilized as a selection criteria for inclusion in the study. Surveys were
We used chi-square tests for categorical variables and t tests for continuous variables to compare smokers to nonsmokers. Models were conducted for all individuals and then stratified by gender to assess whether race was associated with smoking status, as it has been demonstrated gender differences do exist among sexual minority populations with respect to smoking behaviors.3,27 Model 1 was unadjusted. Model 2 was adjusted for age, gender (only in aggregate models), educational attainment, health insurance status, relationship status,
Downloaded from http://ntr.oxfordjournals.org/ at New York University on November 14, 2015
only been able to descriptively compare differences in smoking status stratified by race/ethnicity. Thus, these studies have not been explicitly able to establish whether race is a significant predictor for current smoking status but rather examined prevalence of tobacco use among sexual minority populations or general populations. The 2011 Institute of Medicine’s report on LGBT health recommended more research incorporating intersectional perspectives of individuals who are both sexual minorities and racial minorities.11 Among the few studies that have explored racial heterogeneity in smoking among sexual minorities, most relied on relatively small sample sizes, even after pooling data across years12; were descriptive or among a specific age group13; or were conducted in limited geographies.14 For example, Sanchez, Meacher, and Beil14 showed that in a sample of lesbian and bisexual women from Bronx, NY, 55% of Black women and 62% of Hispanic women were current smokers. Blosnich et al.13 demonstrated among a national sample of college-aged adults that Black LGBs had the lowest prevalence of cigarette smoking, Asian American LGBs had significantly lower smoking prevalence of cigarette smoking, and Hispanic LGBs did not significantly differ in cigarette smoking compared to their White LGB counterparts. While these studies examined racial differences in the prevalence of smoking behaviors among sexual minorities, to our knowledge, it is unclear whether race independently is associated with smoking behavior among sexual minority populations. The interaction of sexual orientation and race in regards to smoking specifically may be epidemiologically complex. For example, the most recent data from the CDC ATS suggests no significant differences in current cigarette smoking among White non-Hispanic, Black non-Hispanic, and Hispanic groups, while Asian non-Hispanic individuals had significantly lower prevalence of smoking and Other non-Hispanic individuals had higher prevalence than the other racial groups. Since LGB status is so robustly associated with higher rates of smoking,6,7 it is unclear whether the patterns of smoking observed among racial/ethnic minority persons may differ when taking account of sexual orientation. For instance, would the general population patterns of racial/ethnic smoking prevalence overlay sexual minority populations (e.g., no differences among White, Black, and Hispanic LGB individuals). To fill this important gap in the literature, we used data from a national sample of sexual minority individuals that included oversampling of individuals with minority racial identities. The 2010 Social Justice Sexuality (SJS) project is one of the largest national surveys of Black, Latino, Asian American/Pacific Islander, and multiracial sexual minority adults. We aimed to explore racial differences in current smoking among sexual minority individuals. Based on past research, we hypothesized that among this sample of sexual minority adults, racial minority individuals will have lower risk of current smoking than their White peers.
1363
Nicotine & Tobacco Research, 2015, Vol. 17, No. 11
1364
Results Table 1 provides our descriptive statistics by smoking status. The overall smoking prevalence in our sample was 28.75%. Among smokers, 22.35% identified as White, 26.98% identified as Black, 19.38% identified as Latino/Hispanic, 5.58% identified as Asian
American, and 25.67% were other/multiracial. Women represented a slightly higher percentage of smokers (51.70%) than men (48.30%). In our bivariate analyses, all demographic characteristics, except for relationship status, were statistically different between smokers and nonsmokers. Table 2 provides results of the risk of smoking by racial identity among the overall sample. In unadjusted models, Blacks (RR = 0.64, 95 % CI = 0.56, 0.74), Asian Americans (RR = 0.74, 95% CI = 0.59, 0.93), and Other (RR = 0.82, 95 % CI = 0.71, 0.95) had significantly lower risk of smoking compared to their White peers. In the fully adjusted model, Blacks still remained at decreased risk of smoking compared to their White counterparts (aRR = 0.70, 95 % CI = 0.61, 0.80) as well as Asian Americans (aRR = 0.74, 95% CI = 0.58, 0.95). In gender stratified models (unadjusted), Black men (RR = 0.60, 95 % CI = 0.49, 0.73) and Asian American men (RR = 0.67, 95 % CI = 0.47, 0.95) and Other men (RR = 0.76, 95 % CI = 0.62, 0.94) were at lower risk of smoking compared to White men. In fully adjusted gender stratified models Black men (aRR = .61, 95% CI = 0.50, 0.75) and Asian American men (aRR = .61, 95% CI = 0.50, 0.75) were significantly at decreased risk of smoking compared to their White counterparts. In unadjusted gender stratified models, Black women (RR = 0.70, 95 % CI = 0.58, 0.84) were the only subgroup exhibiting significant relationship compared to their White peers in that they were at decreased risk of smoking. In the fully adjusted model, Black women remained to be the only group to reveal statistical significance in that they were at decreased risk for smoking (aRR = 0.78, 95 % CI = 0.65, 0.95) compared to White women. Throughout all models, Latinos in general and Latino women and men were at increased risk of smoking yet did not reveal to be statistically significant.
Table 1. Socio-Demographic Characteristics of Participants by Smoking Status: Social Justice Sexuality Study, 2010 (N = 3,982)
Race White Black Latino/Hispanic Asian American Other Gender Male Female Educational attainment High school diploma/GED or less Some college/associates degree Bachelor’s degree or higher Health insurance Yes No Relationship status Partnered Single Household income (M, SD) Age, (M, SD)
Smokers (n = 1,145)
Nonsmokers (n = 2,837)
n (%)
n (%)
X2 or t value statistic
x2 = 59.59*** 256 (22.35%) 309 (26.98%) 222 (19.38%) 64 (5.58%) 294 (25.67%)
639 (22.52%) 1,069 (37.68%) 388 (13.67%) 185 (6.52%) 556 (19.59%)
553 (48.30%) 592 (51.70%)
1,496 (52.73%) 1,341 (47.27%)
628 (54.84%) 301 (26.30%) 216 (18.86%)
994 (35.04%) 934 (32.92%) 909 (32.04%)
827 (72.23%) 318 (27.77%)
2,306 (81.29%) 531 (18.71%)
616 (53.80%) 529 (46.20%) 8.04 (9.76) 33.67 (11.83)
1,574 (55.48%) 1,263 (44.52%) 9.07 (9.80) 37.06 (12.97)
x2 = 6.43*
p value
< .001
.011
x2 = 140.29***
< .001
x2 = 39.88***
< .001
x2 = 0.93
t = 27.49*** t = −1.70***
.334
< .001 < .001
Household income: categories transformed into 12 categories to correspond with the following ranges—$8,500; $8,500–10,999; $11,000–13,499; $13,500– 14,999; $15,000–17,499; $17,500–19,999; $20,000–29,999; $30,000–39,999; $40,000–49,999; $50,000–74,999; $75,000–99,999; $100,000. Thus, a mean household income of 8.04 equates to the range of $30,000–39,999. P-values: *p < .05; **p < .01; ***p < .001.
Downloaded from http://ntr.oxfordjournals.org/ at New York University on November 14, 2015
and household income. Missing data among covariates, specifically educational attainment and household income only, ranged from 9%–11% for the entire analytic sample. Subsequent sensitivity analyses were conducted to assess whether missing data among our covariates could produce different findings in our modeling when predicting smoking status. These results indicated an inability of these missing data to statistically contribute to the prediction of smoking status. This supported our decision to use listwise deletion among covariates and not conduct multiple imputation on missing data among our covariates.28 We employed Log-Poisson regression models for binary outcomes in which results were exponentiated to RRs,29,30 with corresponding 95% confidence intervals (CIs) and p values. Risk ratios (RR) were calculated to assess the risk of smoking by racial identity rather than odds ratios because odds ratios are likely to overestimate the effect when the dependent variable is common.31–35 When the prevalence of an outcome (e.g., tobacco use) is high (e.g., >20%), use of odds ratio will over-estimate the effect.31–33,35–40 Since our sample included a smoking prevalence of roughly 28.75%, exhibiting a relatively high prevalence of smoking, we decided to utilize relative risk ratios to ensure that we were most sufficiently utilizing an estimate that would not render an overestimate. STATA 13.0 was utilized for all analyses in which we employed STATA’s GLM package (for the binomial family with robust standard error estimates) for all Log-Poisson regression models.41
ref 0.64 (0.56, 0.74)**** 1.05 (0.91, 1.21) 0.74 (0.59, 0.93)** 0.82 (0.71, 0.95)**
ref 0.70 (0.61, 0.80)*** 1.02 (0.89, 1.17) 0.74 (0.58, 0.95)* 0.91 (0.79, 1.05)
n = 3,810, aRR (95% CI)
n = 3,982, RR (95% CI) ref 0.60 (0.49, 0.73)*** 1.05 (0.86, 1.28) 0.67 (0.47, 0.95)* 0.78 (0.62, 0.94)*
n = 2,049, RR (95% CI)
Model 1
Men
ref 0.62 (0.51, 0.76)*** 1.01 (0.83, 1.23) 0.68 (0.44, 0.87)* 0.84 (0.68, 1.04)
n = 1,957, aRR (95% CI)
Model 2
Model 2 n = 1,853, aRR (95% CI) ref 0.78 (0.65, 0.95)* 1.04 (0.85, 1.27) 0.83 (0.59, 1.16) 0.99 (0.82, 1.19)
n = 1,933, RR (95% CI) ref 0.70 (0.58, 0.84)*** 1.05 (0.86, 1.28) 0.81 (0.59, 1.10) 0.88 (0.73, 1.06)
Women Model 1
Downloaded from http://ntr.oxfordjournals.org/ at New York University on November 14, 2015
aRR = adjusted risk ratio; CI = confidence interval. Model 1 is unadjusted. Model 2 adjusted for age, educational attainment, income, gender (only in aggregate models), relationship status and health insurance status. *p < .05; **p < .01; ***p < .001, significance tests between those reporting smoking versus those reporting not smoking. P-values: *p < .05; **p < .01; ***p < .001.
Race White (Ref) Black Latino/Hispanic Asian American Other
Model 2
Model 1
Overall
Table 2. Adjusted Risk Ratios of Current Smoking: Social Justice Sexuality Study, 2010 (N = 3,982)
Nicotine & Tobacco Research, 2015, Vol. 17, No. 11 1365
1366
Discussion
populations. Besides race, there may be other demographic characteristics or social dimensions (e.g., nativity/acculturation) useful to examine when looking at heterogeneity among sexual minority populations. Previous research on smoking behavior among Latino/ Hispanic and Asian American populations supports the importance of considering different acculturation processes that might impact smoking behaviors,49,50 but this work has not yet been conducted among sexual minority populations at the population level. Such research could possibly explain the differences in smoking patterns illuminated in our findings. Also, future work should consider variation in smoking status, stratified by race, with respect to sexual orientation identity as research is starting to illuminate a “bisexual paradox,” wherein those that identify as bisexual are demonstrating poorer health outcomes than others within sexual minority subgroups. Post-hoc analyses were conducted further stratifying by sexual orientation and the relative risk of smoking did not change appreciably across the racial groups. Additional future work in this area should examine both tobacco use and tobacco dependence while also considering types of tobacco utilized, which may be differentially distributed among population subgroups. For example, Blacks are more likely than Whites to smoke menthol cigarettes,51 which leads to a high risk of lung cancer in the population as well as other problematic respiratory conditions.52–54 It may be helpful to know if there are racial differences among sexual minorities in use of menthol cigarettes, which is difficult to ascertain as national population-based datasets often include relatively small samples where such questions are probed. More commonly oversampling racial minority persons on federal health surveys, along with incorporating sexual orientation and gender identity items, would be potential steps in providing the data critical for further research. Additionally, future research should seek to extrapolate the influence of socioeconomic status variables in contributing to smoking patterns. Health disparities research has consistently attempted to disentangle the effects of race and socioeconomic status (SES) factors as they both are highly correlated along the casual pathway for explaining health outcomes.55–58 While we sought to explicitly examine racial differences, and in-turn adjusted for SES variables, future research might seek to explore the impact of such variables in explaining smoking status variations among sexual minorities. Disentangling the contributions of SES in understanding racial differences in smoking among sexual minority populations is paramount for future research. Especially future work which seeks to decrease disparities recognizing the diverse heterogeneity comprising sexual minority populations. For example, it has been demonstrated that racial minorities are burdened with disparate targeted tobacco advertising which is even more concentrated in lower SES neighborhoods.59–63 Also recently researchers have identified that sexual minority populations are also experiencing unusually exposure to high levels of tobacco advertising, particularly in venues that are of great importance for development of social capital and strengthening a sense of community and community belonging.63–66 The combination of such social environments potentially can thwart, if not at minimum jeopardize, medical and public health interventions aimed at decreasing smoking and improving smoking cessation via improvement in individual and community resilience. Thus, research uncovering greater clarity into the dynamics of SES, as well as distinguishing their effects above and beyond racial differences, are warranted for tailoring highly effective interventions and producing highly effective policy implications to address racial differences in smoking among sexual minority populations.
Downloaded from http://ntr.oxfordjournals.org/ at New York University on November 14, 2015
Researchers increasingly have emphasized the need to apply nuanced approaches that shed light on socio-demographic variation among sexual minorities.11,42–44 In this study, we embrace intersectionality recognizing heterogeneity among sexual minority populations by assessing the importance of race in explaining smoking behavior among sexual minorities. Our analyses revealed differences by race in current cigarette smoking among sexual minorities, specifically among Black men, Black women, and Asian American men, all of whom exhibited lower risk of smoking compared to their White sexual minority counterparts. Additionally, we employed Log-Poisson regression models to calculate adjusted relative risks rather than traditional odds ratios as an attempt to provide a more conservative estimate of smoking risk. This is a methodological improvement over previous studies as we recognize the high prevalence of smoking patterns among sexual minority populations compared to their heterosexual peers substantiating a need to account for this, including the high prevalence of smoking in the SJS dataset. The overall smoking rate of our sample (28.75%) is slightly lower compared to national estimates of smoking prevalence among sexual minorities (32.80%),7 although the sample included a large enough percentage of smokers calling for a need to account for such overestimation. Our findings supports previous research which indicated that Black and Asian American LGB young adults had a lower prevalence of smoking than their White LGB peers13; yet our results extends this work by demonstrating that such lower prevalence rates also translates into lower risk of smoking. It should be emphasized that racial differences in smoking in the general population reflects that racial minorities are at decreased odds of smoking compared to their White counterparts in general.7 Often, racial minorities, in general, experience poorer health outcomes and health behaviors compared to their White counterparts.45 However, our findings are counterintuitive to such trends and further confirms previous research on racial differences in smoking behaviors among sexual minorities.13 One possible explanation is that, among sexual minority populations, racial minority individuals could have unique forms of resiliency.46,47 For example, Meyer48 suggests that individuals who are both racial and sexual minority may develop different forms of community, leading to varying forms of resiliency. This view contrasts with deficit-based approaches that suggest being both racial and sexual minorities’ exhibit poorer health as a result of conflicting identity conflict. A resiliency perspective counteracts narratives of racial sexual minorities being at increased risk for “double jeopardy” (the state of having multiple marginalized identities resulting in increased stress that manifests in more risky health behaviors). It is important to recognize that resiliency may differentially manifest depending on contextual factors and while our study includes a large sample, our findings are not generalizable. Resilience is a relatively nascent area of inquiry in sexual minority health,47 and unfortunately the SJS did not collect information on resiliency constructs used within the literature previously. Future research is needed not only in terms of resiliency and smoking in general, but also if there may be specific elements of resilience among racial minority communities that can inform the general population. Moreover, in applying concepts of resilience it is extremely important to approach such discussions with appropriate cultural sensitivity so as to not perpetuate pathologizing ideas concerning racial minorities. Future research also should discern other demographic variations that may further elucidate heterogeneity among sexual minority
Nicotine & Tobacco Research, 2015, Vol. 17, No. 11
Nicotine & Tobacco Research, 2015, Vol. 17, No. 11
Conclusions Using data from a national sample of sexual minorities, we found that Blacks and Asian Americans had decreased risks for smoking than Whites. Our findings highlight greater need in attentiveness to heterogeneity among sexual minority populations with respect to understanding smoking and a need for learning about resiliency against smoking among certain sexual minority individuals. These findings contribute to a small, but growing, body of knowledge concerning racial heterogeneity among sexual minority populations.
Funding None declared.
Declaration of Interests None declared.
References 1. Hoyert DL, Xu JQ. Deaths: Preliminary Data for 2011. In: N. V. S. Reports ed. Vol. 61. Hyattsville, MD: National Center for Health Statistics; 2012. 2. Cochran SD, Bandiera FC, Mays VM. Sexual orientation–related differences in tobacco use and secondhand smoke exposure among US adults aged 20 to 59 years: 2003–2010 National Health and Nutrition Examination Surveys. Am J Public Health. 2013;103:1837–1844.
3. Corliss HL, Wadler BM, Jun H-J, et al. Sexual-orientation disparities in cigarette smoking in a longitudinal cohort study of adolescents. Nicotine Tob Res. 2013;15:213–222. 4. Lee JGL, Griffin GK, Melvin CL. Tobacco use among sexual minorities in the USA, 1987 to May 2007: a systematic review. Tob Control. 2009;18:275–282. 5. Rosario M, Corliss HL, Everett BG, et al. Sexual orientation disparities in cancer-related risk behaviors of tobacco, alcohol, sexual behaviors, and diet and physical activity: pooled youth risk behavior surveys. Am J Public Health. 2013;104:245–254. 6. Agaku IT, King BA, Dube SR. Current cigarette smoking among adults—United States, 2005–2012. MMWR Morb Mortal Wkly Rep. 2014;63:29–34. 7. King BA, Dube SR, Tynan MA. Current tobacco use among adults in the United States: findings from the National Adult Tobacco Survey. Am J Public Health. 2012;102:e93–e100. 8. King BA, Dube SR, Tynan MA. Flavored cigar smoking among U.S. adults: findings from the 2009–2010 National Adult Tobacco Survey. Nicotine Tob Res. 2013;15:608–614. 9. Dube SR, Asman K, Malarcher A, Carabollo R. Cigarette smoking among adults and trends in smoking cessation-United States, 2008. Morb Mortal Wkly Rep. 2009;58:1227–1232. 10. Garrett BE, Dube SR, Winder C, Caraballo RS. Cigarette smoking— United States, 2006–2008 and 2009–2010. CDC Health Disparities and Inequalities Report—United States. 2013:62:81. 11. Institute of Medicine. The Health of Lesbian, Gay, Bisexual, and Transgender People: Building a Foundation for Better Understanding. Washington, DC: The National Academies Press; 2011. 12. Hahm HC, Wong FY, Huang ZJ, Ozonoff A, Lee J. Substance use among Asian Americans and Pacific Islanders sexual minority adolescents: findings from the National Longitudinal Study of Adolescent Health. J Adolescent Health. 2008;42:275–283. 13. Blosnich JR, Jarrett T, Horn K. Racial and ethnic differences in current use of cigarettes, cigars, and hookahs among lesbian, gay, and bisexual young adults. Nicotine Tob Res. 2011;13:487–491. 14. Sanchez JP, Meacher P, Beil R. Cigarette smoking and lesbian and bisexual women in the bronx. J Commun Health. 2005;30:23–37. 15. Battle J, DeFreece A. The impact of community involvement, religion, and spirituality on happiness and health among a National Sample of Black Lesbians. Women, Gender, and Families of Color. 2014;2:1–31. 16. Battle J, Harris A. Belonging and acceptance: examining the correlates of sociopolitical involvement among bisexual and lesbian Latinas. J Gay Lesbian Soc Serv. 2013;25:141–157. 17. Battle J, Harris A. Connectedness and the sociopolitical involvement of same-gender-loving Black Men. Men Masc. 2013;16:260–267. 18. Battle J, Pastrana AJ, Daniels J. Social Justice Sexuality Project: 2010 National Survey, including Puerto Rico. 2013. doi: http://doi.org/10.3886/ ICPSR34363.v1. 19. Harris A, Battle J. Unpacking civic engagement: the sociopolitical involvement of same-gender loving Black women. J Lesbian St. 2013;17:195–207. 20. Hesse B, Moser R. Health Information National Trends Survey (HINTS), 2005. 2009. doi: http://doi.org/10.3886/ICPSR24383.v1. 21. Etter J-F, Le Houezec J, Huguelet P, Etter M. Testing the cigarette dependence scale in 4 samples of daily smokers: psychiatric clinics, smoking cessation clinics, a smoking cessation website and in the general population. Addict Behav. 2009;34:446–450. 22. Yawn BP, Mapel DW, Mannino DM, et al. Development of the Lung Function Questionnaire (LFQ) to identify airflow obstruction. Int J COPD. 2010;5:1–10. 23. Newcomb ME, Heinz AJ, Birkett M, Mustanski B. A longitudinal examination of risk and protective factors for cigarette smoking among lesbian, gay, bisexual, and transgender youth. J Adolescent Health [published online ahead of print 2014]. doi: http://dx.doi.org/10.1016/j. jadohealth.2013.10.208. 24. Gilman SE, Martin LT, Abrams DB, et al. Educational attainment and cigarette smoking: a causal association? Int J Epidemiol. 2008;37:615–624.
Downloaded from http://ntr.oxfordjournals.org/ at New York University on November 14, 2015
The current study has several limitations. First, this was a convenience sample, which limits generalizability due to potential selection bias due to voluntary participation. The convenience sampling may also result in overestimation of smoking prevalence since the SJS sampled from social venues (e.g., gay bars) which are settings that may facilitate smoking behaviors,67 although such specific sampling was minimal. However, most health studies among sexual minorities use convenience sampling techniques because of difficulties in assessing this population.42,63,68,69 The large sample size included within the SJS was greatly influenced by maximization of convenience sampling techniques, however. Second, the SJS may have found respondents who were more “out” about their sexual orientation and used a self-identity measure of sexual orientation, which may influence estimates smoking behaviors. Third, self-report of smoking may be a limitation; studies show trends of underestimation when comparing self-report smoking prevalence with biological samples to detect smoking status.70 Additionally, the item to gauge smoking in the SJS was not the item typically used to derive the most common definition of current smoking (i.e., smoked ≤100 cigarettes in one’s lifetime and currently smokes some or every day).7,9,10 Thus, the more liberally worded smoking item in the SJS may have resulted in an overestimate of smoking status in this sample, for which we attempted to redress in part by utilizing Log-Poisson models. Fourth, the sample of some racial groups (e.g., Native Americans) was small and thus precluded analysis for these groups respectively. Large enough sample sizes among Native Americans who are sexual minorities has generally been problematic in previous research13 and substantiated our decision to group them with those identifying with being multiracial resulting in the racial group of other. Finally, residual confounding may be a potential concern since the SJS did not include measures that may be associated with smoking among racial minority individuals (e.g., acculturation).
1367
1368
48. Meyer IH. Identity, stress, and resilience in lesbians, gay men, and bisexuals of color. Couns Psychol. 2010;38:442–454. 49. Iglesias-Rios L, Bromberg JE, Moser RP, Augustson EM. Food insecurity, cigarette smoking, and acculturation among Latinos: data from NHANES 1999–2008. J Immigr Minor Health. 2013;1–9. doi: 10.1007/ s10903-013-9957-7. 50. Li S, Kwon SC, Weerasinghe I, Rey MJ, Trinh-Shevrin C. Smoking among Asian Americans: acculturation and gender in the context of tobacco control policies in New York City. Health Promot Pract. 2013;14(suppl):18S–28S. 51. Cubbin C, Soobader M-J, LeClere FB. The intersection of gender and race/ ethnicity in smoking behaviors among menthol and non-menthol smokers in the United States. Addiction. 2010;105:32–38. 52. Blot WJ, Cohen SS, Aldrich M, McLaughlin JK, Hargreaves MK, Signorello LB. Lung cancer risk among smokers of menthol cigarettes. J Natl Cancer I. 2011;103:810–816. 53. Jones MR, Apelberg BJ, Tellez-Plaza M, Samet JM, Navas-Acien A. Menthol cigarettes, race/ethnicity, and biomarkers of tobacco use in U.S. Adults: the 1999–2010 National Health and Nutrition Examination Survey (NHANES). Cancer Epidem Biomar. 2013;22:224–232. 54. Willis DN, Liu B, Ha MA, Jordt S-E, Morris JB. Menthol attenuates respiratory irritation responses to multiple cigarette smoke irritants. FASEB J. 2011;25:4434–4444. 55. Adler NE, Rehkopf DH. U.S. disparities in health: descriptions, causes, and mechanisms. Annu Rev Public Health. 2008;29:235–252. 56. Shavers VL. Measurement of socioeconomic status in health disparities research. J Natl Med Assoc. 2007;99:1013. 57. Williams DR, Jackson PB. Social sources of racial disparities in health. Health Affair. 2005;24:325–334. 58. Williams DR, Kontos EZ, Viswanath K, et al. Integrating multiple social statuses in health disparities research: the case of lung cancer. Health Serv Res. 2012;47:1255–1277. 59. Barbeau EM, Wolin KY, Naumova EN, Balbach E. Tobacco advertising in communities: associations with race and class. Prev Med. 2005;40:16–22. 60. Cohen EL, Caburnay CA, Rodgers S. Alcohol and tobacco advertising in Black and general audience newspapers: targeting with message cues? J Health Commun. 2011;16:566–582. 61. McCandless PM, Yerger VB, Malone RE. Quid Pro Quo: tobacco companies and the Black Press. Am J Public Health. 2011;102:739–750. 62. Seidenberg AB, Caughey RW, Rees VW, Connolly GN. Storefront cigarette advertising differs by community demographic profile. Am J Health Promot. 2010;24:e26–e31. 63. Blosnich J, Lee JGL, Horn K. A systematic review of the aetiology of tobacco disparities for sexual minorities. Tob Control. 2013;22:66–73. 64. Dilley JA, Spigner C, Boysun MJ, Dent CW, Pizacani BA. Does tobacco industry marketing excessively impact lesbian, gay and bisexual communities? Tob Control. 2008;17:385–390. 65. Matthews AK, Balsam K, Hotton A, Kuhns L, Li C-C, Bowen DJ. Awareness of media-based antitobacco messages among a community sample of LGBT individuals. Health Promot Pract. 2014; 15(6):857–866. doi: 10.1177/1524839914533343. 66. Stevens P, Carlson LM, Hinman JM. An analysis of tobacco industry marketing to lesbian, gay, bisexual, and transgender (LGBT) populations: strategies for mainstream tobacco control and prevention. Health Promot Pract. 2004;5(suppl):129S–134S. 67. Gruskin E, Byrne K, Kools S, Altschuler A. Consequences of frequenting the lesbian bar. Women Health. 2007;44:103–120. 68. DeBlaere C, Brewster ME, Sarkees A, Moradi B. Conducting research with LGB people of color: methodological challenges and strategies. Couns Psychol. 2010;38:331–362. 69. Meyer I, Wilson P. Sampling lesbian, gay, and bisexual populations. J Couns Psychol. 2009;56:23. 70. Gorber SC, Schofield-Hurwitz S, Hardt J, Levasseur G, Tremblay M. The accuracy of self-reported smoking: a systematic review of the relationship between self-reported and cotinine-assessed smoking status. Nicotine Tob Res. 2009;11:12–24.
Downloaded from http://ntr.oxfordjournals.org/ at New York University on November 14, 2015
25. Schoenborn CA. Marital Status and Health, United States 1999–2002: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics; 2004. 26. Malarcher A, Dube S, Shaw L, Babb S, Kaufmann R. Cancer-related news from the CDC: quitting smoking among adults. Oncology Times. 2011;33:62–66. 27. Levinson AH, Hood N, Mahajan R, Russ R. Smoking cessation treatment preferences, intentions, and behaviors among a large sample of Colorado gay, lesbian, bisexual, and transgendered smokers. Nicotine Tob Res. 2012;14:910–918. 28. Thomeer MB. Sexual minority status and self-rated health: the importance of socioeconomic status, age, and sex. Am J Public Health. 2013;103:881–888. 29. Blizzard L, Hosmer W. Parameter estimation and goodness-of-fit in log binomial regression. Biometrical J. 2006;48:5–22. 30. Localio AR, Margolis DJ, Berlin JA. Relative risks and confidence intervals were easily computed indirectly from multivariable logistic regression. J Clin Epidemiol. 2007;60:874–882. 31. Behrens T, Taeger D, Wellmann J, Keil U. Different methods to calculate effect estimates in cross-sectional studies a comparison between prevalence odds ratio and prevalence ratio. Method Inform Med. 2004;43:505–509. 32. Duncan D, Johnson R, Molnar B, Azrael D. Association between neighborhood safety and overweight status among urban adolescents. BMC Public Health. 2009;9:289. 33. Duncan D, Wolin K, Scharoun-Lee M, Ding E, Warner E, Bennett G. Does perception equal reality? Weight misperception in relation to weightrelated attitudes and behaviors among overweight and obese US adults. Int J Behav Nutr Phys Act. 2011;8:20. 34. Spiegelman D, Hertzmark E. Easy SAS calculations for risk or prevalence ratios and differences. Am J Epidemiol. 2005;162:199–200. 35. Zou G. A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159:702–706. 36. Axelson O, Fredriksson M, Ekberg K. Use of the prevalence ratio v the prevalence odds ratio as a measure of risk in cross sectional studies. Occup Environ Med. 1994;51:574. 37. Krieger N, Chen JT, Ware JH, Kaddour A. Race/ethnicity and breast cancer estrogen receptor status: impact of class, missing data, and modeling assumptions. Cancer Cause Control. 2008;19:1305–1318. 38. McNutt L-A, Wu C, Xue X, Hafner JP. Estimating the relative risk in cohort studies and clinical trials of common outcomes. Am J Epidemiol. 2003;157:940–943. 39. Schmidt CO, Kohlmann T. When to use the odds ratio or the relative risk? Int J Public Health. 2008;53:165–167. 40. Thompson ML, Myers JE, Kriebel D. Prevalence odds ratio or prevalence ratio in the analysis of cross sectional data: what is to be done? Occup Environ Med. 1998;55:272–277. 41. StataCorp. Stata Statistical Software (Version Release 13). College Station, TX: StatCorp LP; 2013. 42. Gates GJ. Demographics and LGBT health. J Health Soc Behav. 2013;54:72–74. 43. Kertzner RM, Meyer IH, Frost DM, Stirratt MJ. Social and psychological weil-being in lesbians, gay men, and bisexuals: the effects of race, gender, age, and sexual identity. Am J Orthopsychiat. 2009;79:500–510. 44. Wong CF, Schrager SM, Holloway IW, Meyer IH, Kipke MD. Minority stress experiences and psychological well-being: the impact of support from and connection to social networks within the Los Angeles house and ball communities. Prevention Science. 2013;15:1–12. doi: 10.1007/ s11121-012-0348-4. 45. Williams DR, Sternthal M. Understanding racial-ethnic disparities in health: sociological contributions. J Health Soc Behav. 2010;51(suppl):S15–S27. 46. Earnshaw VA, Bogart LM, Dovidio JF, Williams DR. Stigma and racial/ethnic HIV disparities: moving toward resilience. Am Psychol. 2013;68:225. 47. Herrick AL, Stall R, Goldhammer H, Egan JE, Mayer KH. Resilience as a research framework and as a cornerstone of prevention research for gay and bisexual men: theory and evidence. AIDS Behav. 2014;18:1–9.
Nicotine & Tobacco Research, 2015, Vol. 17, No. 11
Original investigation
Smoking Among Sexual Minorities: Are There Racial Differences? Kasim S. Ortiz MS1, Dustin T. Duncan ScD2, John R. Blosnich PhD, MPH3, Ramzi G. Salloum PhD, MA, MBA4, Juan Battle PhD, MA5 Department of Sociology, Vanderbilt University, Nashville, TN; 2Department of Population Health, New York University School of Medicine, New York, NY; 3VA Center for Health Equity Research and Promotion, Pittsburgh, PA; 4 Department of Health Services Policy and Management, University of South Carolina, Columbia, SC; 5Departments of Sociology, Public Health and Urban Education, City University of New York, New York, NY
Downloaded from http://ntr.oxfordjournals.org/ at New York University on November 14, 2015
1
Corresponding Author: Kasim S. Ortiz, MS. Department of Sociology, Garland Hall, PMB 351811, Nashville, TN37235-1811, Vanderbilt University; E-mail: [email protected]
Abstract Introduction: Smoking prevalence is higher among sexual minorities compared to their heterosexual peers. However, very little is known about potential racial differences in smoking among sexual minority populations. We examined differences by race in smoking status among a robust sample of sexual minorities. Methods: We used data from the 2010 Social Justice Sexuality project, a large national convenience sample of sexual minority adults that oversampled individuals from racial minority groups. LogPoisson multivariable regression models were employed to determine the risk of current smoking among sexual minority individuals by race after controlling for socio-demographic characteristics. Results: Among smokers, 22.35% identified as White, 26.98% identified as Black, 19.38% identified as Latino/Hispanic, 5.58% identified as Asian American, and 25.67% were other/multiracial. In fully adjusted gender stratified models, Black men (adjusted risk ratio [aRR] = 0.61, 95% confidence interval [CI] = 0.50, 0.75) and Asian American men (aRR = 0.61, 95% CI = 0.50, 0.75) were at lower risk of smoking compared to White men. Black women were the only to remain statistically significant for decreased risk of smoking in fully adjusted gender stratified models (aRR = 0.78, 95 % CI = 0.65, 0.95). Conclusions: Among sexual minorities, Black and Asian American individuals consistently were at decreased risk of current smoking compared to their White peers. Future research should seek to understand the mechanisms that contribute to decreased smoking status among racial sexual minorities.
Introduction Cigarette smoking is the leading cause of preventable disease and death in the United States with approximately 443,000 attributable deaths and $193 billion in direct health-care expenditures and yearly productivity losses.1 Research consistently shows that sexual minority populations (i.e., lesbian, gay, bisexual and transgender [LGBT] individuals) have significantly higher prevalence of cigarette smoking compared with heterosexuals.2–5 For example, the National Adult Tobacco Survey (ATS) by the Centers for Disease Control and
Prevention (CDC) found 32.8% of LGBT adults indicated cigarette use compared to 19.5% of heterosexual adults.6–8 Pooling 7 years of data from the National Health and Nutrition Examination Surveys, Cochran, Bandiera, and Mays2 demonstrated that lesbian and bisexual women had higher rates of tobacco use than their heterosexual counterparts. Research has consistently revealed that racial minorities are less likely to smoke than their White counterparts in the general population regardless of sexual orientation.6,7,9,10 A majority of the research examining differences in smoking status among sexual minorities compared to their heterosexual counterparts has
© The Author 2015. Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. For permissions, please e-mail: [email protected].
1362
Nicotine & Tobacco Research, 2015, Vol. 17, No. 11
Methods
collected from over 5,500 respondents in all 50 states (including Washington, DC) and Puerto Rico from January 2010 to December 2010. The survey was fielded in both English and Spanish. Several data collection strategies were used including venue-based sampling, snowball sampling, internet-based outreach, and partnerships with community-based organizations, activists and opinion leaders. The total sample consists of 4,953 complete surveys. We focused our analyses on those who reported their smoking status (N = 3,982). Data collection for the SJS originally was approved by the institutional review board of the City University of New York.
Dependent Variable Smoking status was assessed by the following question, “Do you now smoke cigarettes?” Response options included: not at all; some days; every day. This was dichotomized to represent those that currently smoked and those that did not smoke. This question is the same question utilized within the Health Information National Trends Survey.20 This question has demonstrated strong validity in many settings and in conjunction with a myriad of diseases and outcomes,21,22 as well as among sexual minority populations.23
Independent Variable Race was assessed using the question, “Which of the following racial groups comes closest to which you identify (choose all that apply)?” Respondents had the choice of self-identifying with the following categories: Black, Latino/Hispanic, Asian American/Pacific Islander, Multiracial, Native Americans, White, and other. These categories were coded as: White; Black; Latino/Hispanic; Asian American/ Pacific Islander; and other.
Covariates We adjusted for several covariates known to be associated with smoking, including gender at birth (male, female),7 educational attainment (high school diploma/General Education Diploma (GED) or less; some college/associate’s degree; bachelor’s degree or higher),24 relationship status (partnered, single),25 and age (in years).7 We also adjusted for health insurance status (yes, no), because having health care coverage has been associated with factors related to smoking status, such as access to cessation treatments and quitting smoking.26 Income is also strongly associated with smoking status,10 and in the present study, household income was gathered using a 12-category ordinal response categorization (i.e., under $8,500; $8,500–10,999; $11,000–13,499; $13,500–14,999; $15,000–17,499; $17,500– 19,999; $20,000–29,999; $30,000–39,999; $40,000–49,999; $50,000–74,999; $75,000–99,999; $100,000 and over). The ordinal categories were transformed into distinct categories ranging from 1–12 corresponding with the original ordinal categories. Thus, a household income with a mean of 4 would correspond with the range of $13,500–14,999.
Data
Analytic Strategy
Data are from the 2010 SJS Project, which is a convenience sample of collected data on the experiences of sexual minority people of color concerning five themes: identity (both racial and sexual), physical and mental health, family, religion and spirituality, and sociopolitical involvement. Data collection efforts were employed to oversample racial minority individuals15–19 who identify as lesbian, gay, bisexual or transgender for which self-identification was utilized as a selection criteria for inclusion in the study. Surveys were
We used chi-square tests for categorical variables and t tests for continuous variables to compare smokers to nonsmokers. Models were conducted for all individuals and then stratified by gender to assess whether race was associated with smoking status, as it has been demonstrated gender differences do exist among sexual minority populations with respect to smoking behaviors.3,27 Model 1 was unadjusted. Model 2 was adjusted for age, gender (only in aggregate models), educational attainment, health insurance status, relationship status,
Downloaded from http://ntr.oxfordjournals.org/ at New York University on November 14, 2015
only been able to descriptively compare differences in smoking status stratified by race/ethnicity. Thus, these studies have not been explicitly able to establish whether race is a significant predictor for current smoking status but rather examined prevalence of tobacco use among sexual minority populations or general populations. The 2011 Institute of Medicine’s report on LGBT health recommended more research incorporating intersectional perspectives of individuals who are both sexual minorities and racial minorities.11 Among the few studies that have explored racial heterogeneity in smoking among sexual minorities, most relied on relatively small sample sizes, even after pooling data across years12; were descriptive or among a specific age group13; or were conducted in limited geographies.14 For example, Sanchez, Meacher, and Beil14 showed that in a sample of lesbian and bisexual women from Bronx, NY, 55% of Black women and 62% of Hispanic women were current smokers. Blosnich et al.13 demonstrated among a national sample of college-aged adults that Black LGBs had the lowest prevalence of cigarette smoking, Asian American LGBs had significantly lower smoking prevalence of cigarette smoking, and Hispanic LGBs did not significantly differ in cigarette smoking compared to their White LGB counterparts. While these studies examined racial differences in the prevalence of smoking behaviors among sexual minorities, to our knowledge, it is unclear whether race independently is associated with smoking behavior among sexual minority populations. The interaction of sexual orientation and race in regards to smoking specifically may be epidemiologically complex. For example, the most recent data from the CDC ATS suggests no significant differences in current cigarette smoking among White non-Hispanic, Black non-Hispanic, and Hispanic groups, while Asian non-Hispanic individuals had significantly lower prevalence of smoking and Other non-Hispanic individuals had higher prevalence than the other racial groups. Since LGB status is so robustly associated with higher rates of smoking,6,7 it is unclear whether the patterns of smoking observed among racial/ethnic minority persons may differ when taking account of sexual orientation. For instance, would the general population patterns of racial/ethnic smoking prevalence overlay sexual minority populations (e.g., no differences among White, Black, and Hispanic LGB individuals). To fill this important gap in the literature, we used data from a national sample of sexual minority individuals that included oversampling of individuals with minority racial identities. The 2010 Social Justice Sexuality (SJS) project is one of the largest national surveys of Black, Latino, Asian American/Pacific Islander, and multiracial sexual minority adults. We aimed to explore racial differences in current smoking among sexual minority individuals. Based on past research, we hypothesized that among this sample of sexual minority adults, racial minority individuals will have lower risk of current smoking than their White peers.
1363
Nicotine & Tobacco Research, 2015, Vol. 17, No. 11
1364
Results Table 1 provides our descriptive statistics by smoking status. The overall smoking prevalence in our sample was 28.75%. Among smokers, 22.35% identified as White, 26.98% identified as Black, 19.38% identified as Latino/Hispanic, 5.58% identified as Asian
American, and 25.67% were other/multiracial. Women represented a slightly higher percentage of smokers (51.70%) than men (48.30%). In our bivariate analyses, all demographic characteristics, except for relationship status, were statistically different between smokers and nonsmokers. Table 2 provides results of the risk of smoking by racial identity among the overall sample. In unadjusted models, Blacks (RR = 0.64, 95 % CI = 0.56, 0.74), Asian Americans (RR = 0.74, 95% CI = 0.59, 0.93), and Other (RR = 0.82, 95 % CI = 0.71, 0.95) had significantly lower risk of smoking compared to their White peers. In the fully adjusted model, Blacks still remained at decreased risk of smoking compared to their White counterparts (aRR = 0.70, 95 % CI = 0.61, 0.80) as well as Asian Americans (aRR = 0.74, 95% CI = 0.58, 0.95). In gender stratified models (unadjusted), Black men (RR = 0.60, 95 % CI = 0.49, 0.73) and Asian American men (RR = 0.67, 95 % CI = 0.47, 0.95) and Other men (RR = 0.76, 95 % CI = 0.62, 0.94) were at lower risk of smoking compared to White men. In fully adjusted gender stratified models Black men (aRR = .61, 95% CI = 0.50, 0.75) and Asian American men (aRR = .61, 95% CI = 0.50, 0.75) were significantly at decreased risk of smoking compared to their White counterparts. In unadjusted gender stratified models, Black women (RR = 0.70, 95 % CI = 0.58, 0.84) were the only subgroup exhibiting significant relationship compared to their White peers in that they were at decreased risk of smoking. In the fully adjusted model, Black women remained to be the only group to reveal statistical significance in that they were at decreased risk for smoking (aRR = 0.78, 95 % CI = 0.65, 0.95) compared to White women. Throughout all models, Latinos in general and Latino women and men were at increased risk of smoking yet did not reveal to be statistically significant.
Table 1. Socio-Demographic Characteristics of Participants by Smoking Status: Social Justice Sexuality Study, 2010 (N = 3,982)
Race White Black Latino/Hispanic Asian American Other Gender Male Female Educational attainment High school diploma/GED or less Some college/associates degree Bachelor’s degree or higher Health insurance Yes No Relationship status Partnered Single Household income (M, SD) Age, (M, SD)
Smokers (n = 1,145)
Nonsmokers (n = 2,837)
n (%)
n (%)
X2 or t value statistic
x2 = 59.59*** 256 (22.35%) 309 (26.98%) 222 (19.38%) 64 (5.58%) 294 (25.67%)
639 (22.52%) 1,069 (37.68%) 388 (13.67%) 185 (6.52%) 556 (19.59%)
553 (48.30%) 592 (51.70%)
1,496 (52.73%) 1,341 (47.27%)
628 (54.84%) 301 (26.30%) 216 (18.86%)
994 (35.04%) 934 (32.92%) 909 (32.04%)
827 (72.23%) 318 (27.77%)
2,306 (81.29%) 531 (18.71%)
616 (53.80%) 529 (46.20%) 8.04 (9.76) 33.67 (11.83)
1,574 (55.48%) 1,263 (44.52%) 9.07 (9.80) 37.06 (12.97)
x2 = 6.43*
p value
< .001
.011
x2 = 140.29***
< .001
x2 = 39.88***
< .001
x2 = 0.93
t = 27.49*** t = −1.70***
.334
< .001 < .001
Household income: categories transformed into 12 categories to correspond with the following ranges—$8,500; $8,500–10,999; $11,000–13,499; $13,500– 14,999; $15,000–17,499; $17,500–19,999; $20,000–29,999; $30,000–39,999; $40,000–49,999; $50,000–74,999; $75,000–99,999; $100,000. Thus, a mean household income of 8.04 equates to the range of $30,000–39,999. P-values: *p < .05; **p < .01; ***p < .001.
Downloaded from http://ntr.oxfordjournals.org/ at New York University on November 14, 2015
and household income. Missing data among covariates, specifically educational attainment and household income only, ranged from 9%–11% for the entire analytic sample. Subsequent sensitivity analyses were conducted to assess whether missing data among our covariates could produce different findings in our modeling when predicting smoking status. These results indicated an inability of these missing data to statistically contribute to the prediction of smoking status. This supported our decision to use listwise deletion among covariates and not conduct multiple imputation on missing data among our covariates.28 We employed Log-Poisson regression models for binary outcomes in which results were exponentiated to RRs,29,30 with corresponding 95% confidence intervals (CIs) and p values. Risk ratios (RR) were calculated to assess the risk of smoking by racial identity rather than odds ratios because odds ratios are likely to overestimate the effect when the dependent variable is common.31–35 When the prevalence of an outcome (e.g., tobacco use) is high (e.g., >20%), use of odds ratio will over-estimate the effect.31–33,35–40 Since our sample included a smoking prevalence of roughly 28.75%, exhibiting a relatively high prevalence of smoking, we decided to utilize relative risk ratios to ensure that we were most sufficiently utilizing an estimate that would not render an overestimate. STATA 13.0 was utilized for all analyses in which we employed STATA’s GLM package (for the binomial family with robust standard error estimates) for all Log-Poisson regression models.41
ref 0.64 (0.56, 0.74)**** 1.05 (0.91, 1.21) 0.74 (0.59, 0.93)** 0.82 (0.71, 0.95)**
ref 0.70 (0.61, 0.80)*** 1.02 (0.89, 1.17) 0.74 (0.58, 0.95)* 0.91 (0.79, 1.05)
n = 3,810, aRR (95% CI)
n = 3,982, RR (95% CI) ref 0.60 (0.49, 0.73)*** 1.05 (0.86, 1.28) 0.67 (0.47, 0.95)* 0.78 (0.62, 0.94)*
n = 2,049, RR (95% CI)
Model 1
Men
ref 0.62 (0.51, 0.76)*** 1.01 (0.83, 1.23) 0.68 (0.44, 0.87)* 0.84 (0.68, 1.04)
n = 1,957, aRR (95% CI)
Model 2
Model 2 n = 1,853, aRR (95% CI) ref 0.78 (0.65, 0.95)* 1.04 (0.85, 1.27) 0.83 (0.59, 1.16) 0.99 (0.82, 1.19)
n = 1,933, RR (95% CI) ref 0.70 (0.58, 0.84)*** 1.05 (0.86, 1.28) 0.81 (0.59, 1.10) 0.88 (0.73, 1.06)
Women Model 1
Downloaded from http://ntr.oxfordjournals.org/ at New York University on November 14, 2015
aRR = adjusted risk ratio; CI = confidence interval. Model 1 is unadjusted. Model 2 adjusted for age, educational attainment, income, gender (only in aggregate models), relationship status and health insurance status. *p < .05; **p < .01; ***p < .001, significance tests between those reporting smoking versus those reporting not smoking. P-values: *p < .05; **p < .01; ***p < .001.
Race White (Ref) Black Latino/Hispanic Asian American Other
Model 2
Model 1
Overall
Table 2. Adjusted Risk Ratios of Current Smoking: Social Justice Sexuality Study, 2010 (N = 3,982)
Nicotine & Tobacco Research, 2015, Vol. 17, No. 11 1365
1366
Discussion
populations. Besides race, there may be other demographic characteristics or social dimensions (e.g., nativity/acculturation) useful to examine when looking at heterogeneity among sexual minority populations. Previous research on smoking behavior among Latino/ Hispanic and Asian American populations supports the importance of considering different acculturation processes that might impact smoking behaviors,49,50 but this work has not yet been conducted among sexual minority populations at the population level. Such research could possibly explain the differences in smoking patterns illuminated in our findings. Also, future work should consider variation in smoking status, stratified by race, with respect to sexual orientation identity as research is starting to illuminate a “bisexual paradox,” wherein those that identify as bisexual are demonstrating poorer health outcomes than others within sexual minority subgroups. Post-hoc analyses were conducted further stratifying by sexual orientation and the relative risk of smoking did not change appreciably across the racial groups. Additional future work in this area should examine both tobacco use and tobacco dependence while also considering types of tobacco utilized, which may be differentially distributed among population subgroups. For example, Blacks are more likely than Whites to smoke menthol cigarettes,51 which leads to a high risk of lung cancer in the population as well as other problematic respiratory conditions.52–54 It may be helpful to know if there are racial differences among sexual minorities in use of menthol cigarettes, which is difficult to ascertain as national population-based datasets often include relatively small samples where such questions are probed. More commonly oversampling racial minority persons on federal health surveys, along with incorporating sexual orientation and gender identity items, would be potential steps in providing the data critical for further research. Additionally, future research should seek to extrapolate the influence of socioeconomic status variables in contributing to smoking patterns. Health disparities research has consistently attempted to disentangle the effects of race and socioeconomic status (SES) factors as they both are highly correlated along the casual pathway for explaining health outcomes.55–58 While we sought to explicitly examine racial differences, and in-turn adjusted for SES variables, future research might seek to explore the impact of such variables in explaining smoking status variations among sexual minorities. Disentangling the contributions of SES in understanding racial differences in smoking among sexual minority populations is paramount for future research. Especially future work which seeks to decrease disparities recognizing the diverse heterogeneity comprising sexual minority populations. For example, it has been demonstrated that racial minorities are burdened with disparate targeted tobacco advertising which is even more concentrated in lower SES neighborhoods.59–63 Also recently researchers have identified that sexual minority populations are also experiencing unusually exposure to high levels of tobacco advertising, particularly in venues that are of great importance for development of social capital and strengthening a sense of community and community belonging.63–66 The combination of such social environments potentially can thwart, if not at minimum jeopardize, medical and public health interventions aimed at decreasing smoking and improving smoking cessation via improvement in individual and community resilience. Thus, research uncovering greater clarity into the dynamics of SES, as well as distinguishing their effects above and beyond racial differences, are warranted for tailoring highly effective interventions and producing highly effective policy implications to address racial differences in smoking among sexual minority populations.
Downloaded from http://ntr.oxfordjournals.org/ at New York University on November 14, 2015
Researchers increasingly have emphasized the need to apply nuanced approaches that shed light on socio-demographic variation among sexual minorities.11,42–44 In this study, we embrace intersectionality recognizing heterogeneity among sexual minority populations by assessing the importance of race in explaining smoking behavior among sexual minorities. Our analyses revealed differences by race in current cigarette smoking among sexual minorities, specifically among Black men, Black women, and Asian American men, all of whom exhibited lower risk of smoking compared to their White sexual minority counterparts. Additionally, we employed Log-Poisson regression models to calculate adjusted relative risks rather than traditional odds ratios as an attempt to provide a more conservative estimate of smoking risk. This is a methodological improvement over previous studies as we recognize the high prevalence of smoking patterns among sexual minority populations compared to their heterosexual peers substantiating a need to account for this, including the high prevalence of smoking in the SJS dataset. The overall smoking rate of our sample (28.75%) is slightly lower compared to national estimates of smoking prevalence among sexual minorities (32.80%),7 although the sample included a large enough percentage of smokers calling for a need to account for such overestimation. Our findings supports previous research which indicated that Black and Asian American LGB young adults had a lower prevalence of smoking than their White LGB peers13; yet our results extends this work by demonstrating that such lower prevalence rates also translates into lower risk of smoking. It should be emphasized that racial differences in smoking in the general population reflects that racial minorities are at decreased odds of smoking compared to their White counterparts in general.7 Often, racial minorities, in general, experience poorer health outcomes and health behaviors compared to their White counterparts.45 However, our findings are counterintuitive to such trends and further confirms previous research on racial differences in smoking behaviors among sexual minorities.13 One possible explanation is that, among sexual minority populations, racial minority individuals could have unique forms of resiliency.46,47 For example, Meyer48 suggests that individuals who are both racial and sexual minority may develop different forms of community, leading to varying forms of resiliency. This view contrasts with deficit-based approaches that suggest being both racial and sexual minorities’ exhibit poorer health as a result of conflicting identity conflict. A resiliency perspective counteracts narratives of racial sexual minorities being at increased risk for “double jeopardy” (the state of having multiple marginalized identities resulting in increased stress that manifests in more risky health behaviors). It is important to recognize that resiliency may differentially manifest depending on contextual factors and while our study includes a large sample, our findings are not generalizable. Resilience is a relatively nascent area of inquiry in sexual minority health,47 and unfortunately the SJS did not collect information on resiliency constructs used within the literature previously. Future research is needed not only in terms of resiliency and smoking in general, but also if there may be specific elements of resilience among racial minority communities that can inform the general population. Moreover, in applying concepts of resilience it is extremely important to approach such discussions with appropriate cultural sensitivity so as to not perpetuate pathologizing ideas concerning racial minorities. Future research also should discern other demographic variations that may further elucidate heterogeneity among sexual minority
Nicotine & Tobacco Research, 2015, Vol. 17, No. 11
Nicotine & Tobacco Research, 2015, Vol. 17, No. 11
Conclusions Using data from a national sample of sexual minorities, we found that Blacks and Asian Americans had decreased risks for smoking than Whites. Our findings highlight greater need in attentiveness to heterogeneity among sexual minority populations with respect to understanding smoking and a need for learning about resiliency against smoking among certain sexual minority individuals. These findings contribute to a small, but growing, body of knowledge concerning racial heterogeneity among sexual minority populations.
Funding None declared.
Declaration of Interests None declared.
References 1. Hoyert DL, Xu JQ. Deaths: Preliminary Data for 2011. In: N. V. S. Reports ed. Vol. 61. Hyattsville, MD: National Center for Health Statistics; 2012. 2. Cochran SD, Bandiera FC, Mays VM. Sexual orientation–related differences in tobacco use and secondhand smoke exposure among US adults aged 20 to 59 years: 2003–2010 National Health and Nutrition Examination Surveys. Am J Public Health. 2013;103:1837–1844.
3. Corliss HL, Wadler BM, Jun H-J, et al. Sexual-orientation disparities in cigarette smoking in a longitudinal cohort study of adolescents. Nicotine Tob Res. 2013;15:213–222. 4. Lee JGL, Griffin GK, Melvin CL. Tobacco use among sexual minorities in the USA, 1987 to May 2007: a systematic review. Tob Control. 2009;18:275–282. 5. Rosario M, Corliss HL, Everett BG, et al. Sexual orientation disparities in cancer-related risk behaviors of tobacco, alcohol, sexual behaviors, and diet and physical activity: pooled youth risk behavior surveys. Am J Public Health. 2013;104:245–254. 6. Agaku IT, King BA, Dube SR. Current cigarette smoking among adults—United States, 2005–2012. MMWR Morb Mortal Wkly Rep. 2014;63:29–34. 7. King BA, Dube SR, Tynan MA. Current tobacco use among adults in the United States: findings from the National Adult Tobacco Survey. Am J Public Health. 2012;102:e93–e100. 8. King BA, Dube SR, Tynan MA. Flavored cigar smoking among U.S. adults: findings from the 2009–2010 National Adult Tobacco Survey. Nicotine Tob Res. 2013;15:608–614. 9. Dube SR, Asman K, Malarcher A, Carabollo R. Cigarette smoking among adults and trends in smoking cessation-United States, 2008. Morb Mortal Wkly Rep. 2009;58:1227–1232. 10. Garrett BE, Dube SR, Winder C, Caraballo RS. Cigarette smoking— United States, 2006–2008 and 2009–2010. CDC Health Disparities and Inequalities Report—United States. 2013:62:81. 11. Institute of Medicine. The Health of Lesbian, Gay, Bisexual, and Transgender People: Building a Foundation for Better Understanding. Washington, DC: The National Academies Press; 2011. 12. Hahm HC, Wong FY, Huang ZJ, Ozonoff A, Lee J. Substance use among Asian Americans and Pacific Islanders sexual minority adolescents: findings from the National Longitudinal Study of Adolescent Health. J Adolescent Health. 2008;42:275–283. 13. Blosnich JR, Jarrett T, Horn K. Racial and ethnic differences in current use of cigarettes, cigars, and hookahs among lesbian, gay, and bisexual young adults. Nicotine Tob Res. 2011;13:487–491. 14. Sanchez JP, Meacher P, Beil R. Cigarette smoking and lesbian and bisexual women in the bronx. J Commun Health. 2005;30:23–37. 15. Battle J, DeFreece A. The impact of community involvement, religion, and spirituality on happiness and health among a National Sample of Black Lesbians. Women, Gender, and Families of Color. 2014;2:1–31. 16. Battle J, Harris A. Belonging and acceptance: examining the correlates of sociopolitical involvement among bisexual and lesbian Latinas. J Gay Lesbian Soc Serv. 2013;25:141–157. 17. Battle J, Harris A. Connectedness and the sociopolitical involvement of same-gender-loving Black Men. Men Masc. 2013;16:260–267. 18. Battle J, Pastrana AJ, Daniels J. Social Justice Sexuality Project: 2010 National Survey, including Puerto Rico. 2013. doi: http://doi.org/10.3886/ ICPSR34363.v1. 19. Harris A, Battle J. Unpacking civic engagement: the sociopolitical involvement of same-gender loving Black women. J Lesbian St. 2013;17:195–207. 20. Hesse B, Moser R. Health Information National Trends Survey (HINTS), 2005. 2009. doi: http://doi.org/10.3886/ICPSR24383.v1. 21. Etter J-F, Le Houezec J, Huguelet P, Etter M. Testing the cigarette dependence scale in 4 samples of daily smokers: psychiatric clinics, smoking cessation clinics, a smoking cessation website and in the general population. Addict Behav. 2009;34:446–450. 22. Yawn BP, Mapel DW, Mannino DM, et al. Development of the Lung Function Questionnaire (LFQ) to identify airflow obstruction. Int J COPD. 2010;5:1–10. 23. Newcomb ME, Heinz AJ, Birkett M, Mustanski B. A longitudinal examination of risk and protective factors for cigarette smoking among lesbian, gay, bisexual, and transgender youth. J Adolescent Health [published online ahead of print 2014]. doi: http://dx.doi.org/10.1016/j. jadohealth.2013.10.208. 24. Gilman SE, Martin LT, Abrams DB, et al. Educational attainment and cigarette smoking: a causal association? Int J Epidemiol. 2008;37:615–624.
Downloaded from http://ntr.oxfordjournals.org/ at New York University on November 14, 2015
The current study has several limitations. First, this was a convenience sample, which limits generalizability due to potential selection bias due to voluntary participation. The convenience sampling may also result in overestimation of smoking prevalence since the SJS sampled from social venues (e.g., gay bars) which are settings that may facilitate smoking behaviors,67 although such specific sampling was minimal. However, most health studies among sexual minorities use convenience sampling techniques because of difficulties in assessing this population.42,63,68,69 The large sample size included within the SJS was greatly influenced by maximization of convenience sampling techniques, however. Second, the SJS may have found respondents who were more “out” about their sexual orientation and used a self-identity measure of sexual orientation, which may influence estimates smoking behaviors. Third, self-report of smoking may be a limitation; studies show trends of underestimation when comparing self-report smoking prevalence with biological samples to detect smoking status.70 Additionally, the item to gauge smoking in the SJS was not the item typically used to derive the most common definition of current smoking (i.e., smoked ≤100 cigarettes in one’s lifetime and currently smokes some or every day).7,9,10 Thus, the more liberally worded smoking item in the SJS may have resulted in an overestimate of smoking status in this sample, for which we attempted to redress in part by utilizing Log-Poisson models. Fourth, the sample of some racial groups (e.g., Native Americans) was small and thus precluded analysis for these groups respectively. Large enough sample sizes among Native Americans who are sexual minorities has generally been problematic in previous research13 and substantiated our decision to group them with those identifying with being multiracial resulting in the racial group of other. Finally, residual confounding may be a potential concern since the SJS did not include measures that may be associated with smoking among racial minority individuals (e.g., acculturation).
1367
1368
48. Meyer IH. Identity, stress, and resilience in lesbians, gay men, and bisexuals of color. Couns Psychol. 2010;38:442–454. 49. Iglesias-Rios L, Bromberg JE, Moser RP, Augustson EM. Food insecurity, cigarette smoking, and acculturation among Latinos: data from NHANES 1999–2008. J Immigr Minor Health. 2013;1–9. doi: 10.1007/ s10903-013-9957-7. 50. Li S, Kwon SC, Weerasinghe I, Rey MJ, Trinh-Shevrin C. Smoking among Asian Americans: acculturation and gender in the context of tobacco control policies in New York City. Health Promot Pract. 2013;14(suppl):18S–28S. 51. Cubbin C, Soobader M-J, LeClere FB. The intersection of gender and race/ ethnicity in smoking behaviors among menthol and non-menthol smokers in the United States. Addiction. 2010;105:32–38. 52. Blot WJ, Cohen SS, Aldrich M, McLaughlin JK, Hargreaves MK, Signorello LB. Lung cancer risk among smokers of menthol cigarettes. J Natl Cancer I. 2011;103:810–816. 53. Jones MR, Apelberg BJ, Tellez-Plaza M, Samet JM, Navas-Acien A. Menthol cigarettes, race/ethnicity, and biomarkers of tobacco use in U.S. Adults: the 1999–2010 National Health and Nutrition Examination Survey (NHANES). Cancer Epidem Biomar. 2013;22:224–232. 54. Willis DN, Liu B, Ha MA, Jordt S-E, Morris JB. Menthol attenuates respiratory irritation responses to multiple cigarette smoke irritants. FASEB J. 2011;25:4434–4444. 55. Adler NE, Rehkopf DH. U.S. disparities in health: descriptions, causes, and mechanisms. Annu Rev Public Health. 2008;29:235–252. 56. Shavers VL. Measurement of socioeconomic status in health disparities research. J Natl Med Assoc. 2007;99:1013. 57. Williams DR, Jackson PB. Social sources of racial disparities in health. Health Affair. 2005;24:325–334. 58. Williams DR, Kontos EZ, Viswanath K, et al. Integrating multiple social statuses in health disparities research: the case of lung cancer. Health Serv Res. 2012;47:1255–1277. 59. Barbeau EM, Wolin KY, Naumova EN, Balbach E. Tobacco advertising in communities: associations with race and class. Prev Med. 2005;40:16–22. 60. Cohen EL, Caburnay CA, Rodgers S. Alcohol and tobacco advertising in Black and general audience newspapers: targeting with message cues? J Health Commun. 2011;16:566–582. 61. McCandless PM, Yerger VB, Malone RE. Quid Pro Quo: tobacco companies and the Black Press. Am J Public Health. 2011;102:739–750. 62. Seidenberg AB, Caughey RW, Rees VW, Connolly GN. Storefront cigarette advertising differs by community demographic profile. Am J Health Promot. 2010;24:e26–e31. 63. Blosnich J, Lee JGL, Horn K. A systematic review of the aetiology of tobacco disparities for sexual minorities. Tob Control. 2013;22:66–73. 64. Dilley JA, Spigner C, Boysun MJ, Dent CW, Pizacani BA. Does tobacco industry marketing excessively impact lesbian, gay and bisexual communities? Tob Control. 2008;17:385–390. 65. Matthews AK, Balsam K, Hotton A, Kuhns L, Li C-C, Bowen DJ. Awareness of media-based antitobacco messages among a community sample of LGBT individuals. Health Promot Pract. 2014; 15(6):857–866. doi: 10.1177/1524839914533343. 66. Stevens P, Carlson LM, Hinman JM. An analysis of tobacco industry marketing to lesbian, gay, bisexual, and transgender (LGBT) populations: strategies for mainstream tobacco control and prevention. Health Promot Pract. 2004;5(suppl):129S–134S. 67. Gruskin E, Byrne K, Kools S, Altschuler A. Consequences of frequenting the lesbian bar. Women Health. 2007;44:103–120. 68. DeBlaere C, Brewster ME, Sarkees A, Moradi B. Conducting research with LGB people of color: methodological challenges and strategies. Couns Psychol. 2010;38:331–362. 69. Meyer I, Wilson P. Sampling lesbian, gay, and bisexual populations. J Couns Psychol. 2009;56:23. 70. Gorber SC, Schofield-Hurwitz S, Hardt J, Levasseur G, Tremblay M. The accuracy of self-reported smoking: a systematic review of the relationship between self-reported and cotinine-assessed smoking status. Nicotine Tob Res. 2009;11:12–24.
Downloaded from http://ntr.oxfordjournals.org/ at New York University on November 14, 2015
25. Schoenborn CA. Marital Status and Health, United States 1999–2002: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics; 2004. 26. Malarcher A, Dube S, Shaw L, Babb S, Kaufmann R. Cancer-related news from the CDC: quitting smoking among adults. Oncology Times. 2011;33:62–66. 27. Levinson AH, Hood N, Mahajan R, Russ R. Smoking cessation treatment preferences, intentions, and behaviors among a large sample of Colorado gay, lesbian, bisexual, and transgendered smokers. Nicotine Tob Res. 2012;14:910–918. 28. Thomeer MB. Sexual minority status and self-rated health: the importance of socioeconomic status, age, and sex. Am J Public Health. 2013;103:881–888. 29. Blizzard L, Hosmer W. Parameter estimation and goodness-of-fit in log binomial regression. Biometrical J. 2006;48:5–22. 30. Localio AR, Margolis DJ, Berlin JA. Relative risks and confidence intervals were easily computed indirectly from multivariable logistic regression. J Clin Epidemiol. 2007;60:874–882. 31. Behrens T, Taeger D, Wellmann J, Keil U. Different methods to calculate effect estimates in cross-sectional studies a comparison between prevalence odds ratio and prevalence ratio. Method Inform Med. 2004;43:505–509. 32. Duncan D, Johnson R, Molnar B, Azrael D. Association between neighborhood safety and overweight status among urban adolescents. BMC Public Health. 2009;9:289. 33. Duncan D, Wolin K, Scharoun-Lee M, Ding E, Warner E, Bennett G. Does perception equal reality? Weight misperception in relation to weightrelated attitudes and behaviors among overweight and obese US adults. Int J Behav Nutr Phys Act. 2011;8:20. 34. Spiegelman D, Hertzmark E. Easy SAS calculations for risk or prevalence ratios and differences. Am J Epidemiol. 2005;162:199–200. 35. Zou G. A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159:702–706. 36. Axelson O, Fredriksson M, Ekberg K. Use of the prevalence ratio v the prevalence odds ratio as a measure of risk in cross sectional studies. Occup Environ Med. 1994;51:574. 37. Krieger N, Chen JT, Ware JH, Kaddour A. Race/ethnicity and breast cancer estrogen receptor status: impact of class, missing data, and modeling assumptions. Cancer Cause Control. 2008;19:1305–1318. 38. McNutt L-A, Wu C, Xue X, Hafner JP. Estimating the relative risk in cohort studies and clinical trials of common outcomes. Am J Epidemiol. 2003;157:940–943. 39. Schmidt CO, Kohlmann T. When to use the odds ratio or the relative risk? Int J Public Health. 2008;53:165–167. 40. Thompson ML, Myers JE, Kriebel D. Prevalence odds ratio or prevalence ratio in the analysis of cross sectional data: what is to be done? Occup Environ Med. 1998;55:272–277. 41. StataCorp. Stata Statistical Software (Version Release 13). College Station, TX: StatCorp LP; 2013. 42. Gates GJ. Demographics and LGBT health. J Health Soc Behav. 2013;54:72–74. 43. Kertzner RM, Meyer IH, Frost DM, Stirratt MJ. Social and psychological weil-being in lesbians, gay men, and bisexuals: the effects of race, gender, age, and sexual identity. Am J Orthopsychiat. 2009;79:500–510. 44. Wong CF, Schrager SM, Holloway IW, Meyer IH, Kipke MD. Minority stress experiences and psychological well-being: the impact of support from and connection to social networks within the Los Angeles house and ball communities. Prevention Science. 2013;15:1–12. doi: 10.1007/ s11121-012-0348-4. 45. Williams DR, Sternthal M. Understanding racial-ethnic disparities in health: sociological contributions. J Health Soc Behav. 2010;51(suppl):S15–S27. 46. Earnshaw VA, Bogart LM, Dovidio JF, Williams DR. Stigma and racial/ethnic HIV disparities: moving toward resilience. Am Psychol. 2013;68:225. 47. Herrick AL, Stall R, Goldhammer H, Egan JE, Mayer KH. Resilience as a research framework and as a cornerstone of prevention research for gay and bisexual men: theory and evidence. AIDS Behav. 2014;18:1–9.
Nicotine & Tobacco Research, 2015, Vol. 17, No. 11
