Theses Doctoral

Structural Factors and Sexual Orientation Disparities in Adolescent Substance Use: A Multi-Level Model

Frazer, Melanie Somjen

Studies consistently find sexual orientation disparities in tobacco, alcohol and marijuana use among adolescents in the United States (Goldbach, Tanner-Smith, Bagwell, & Dunlap, 2014; Hatzenbuehler, Jun, Corliss, & Austin, 2015; Kann et al., 2018; Marshal et al., 2008; Mustanski, Van Wagenen, Birkett, Eyster, & Corliss, 2014). Having documented this elevated risk, the field has turned to identifying mediators that may explain the sexual orientation disparity in substance use behaviors. There is growing evidence that one of the mediators of sexual orientation-based health disparities is structural stigma—defined as “societal-level conditions, cultural norms, and institutional policies that constrain the opportunities, resources, and well-being of the stigmatized” (Hatzenbuehler & Link, 2014, p.2). Structural stigma can in turn lead to greater stress and subsequent negative coping behaviors among sexual minorities (Hatzenbuehler et al., 2015; Hatzenbuehler & Link, 2014; Hatzenbuehler & McLaughlin, 2014; Pachankis, Hatzenbuehler, & Starks, 2014). However, most of this work has been conducted among adults and has focused on mental health outcomes (e.g., suicide attempts, psychiatric morbidity, mental distress) as opposed to specific substance use behaviors; further, if structural factors are found to affect sexual orientation health disparities and these can be altered, the gap between sexual minority and heterosexual substance use can be narrowed through policy change. State-level substance use policy (e.g., levels of taxation of cigarettes and alcohol, policies that remove adolescent access to driving upon substance use infractions) may also explain the gap in substance use prevalence between sexual minority and heterosexual youth because tobacco, alcohol and marijuana are also stigmatized substances and restrictive substance use environments may enhance the stigma attached to sexual minority youth. However, only one study has explored this research question, and it was conducted with an adult sample (Hatzenbuehler, Keyes, Hamilton, & Hasin, 2014).

This project aims to address these gaps in the literature. It uses multi-level modeling to test the association between (1) state-level sexual minority structural stigma (SMSS) and (2) state-level substance use policy environments (SUPEs) and sexual orientation disparities in recent substance use as measured in the Youth Risk Behavior Surveillance Survey (YRBSS). Analyses were also conducted to assess the size and direction of the association between SMSS and SUPE and state-level prevalence of substance use within heterosexual and sexual minority youth subpopulations. Analyses were conducted in subpopulations of female and male youth.

Support was found for the structural stigma hypothesis among girls but not boys. Higher overall measures of sexual minority structural stigma were associated with larger disparities in tobacco use, alcohol use, and marijuana use among girls. Higher scores on the SMSS scale were associated with tobacco use among sexual minority girls but not heterosexual girls. Individual indicators of structural stigma (e.g., specific state-level policies) were also associated with disparities in these outcomes, as well as with the other outcomes tested (binge drinking and drunk driving). No such associations were found among male youth.

Very little support was found for the hypotheses that restrictive substance use policy environments would be associated with larger sexual orientation disparities in substance use and that heterosexual but not sexual minority youth would demonstrate lower prevalence of substance use in restrictive states. Among girls, no associations were found between scales measuring the restrictiveness of substance use and sexual orientation disparities in substance use. A smaller sexual orientation disparity in recent cigarette smoking was associated with one indicator of restrictiveness: state enforcement of underage tobacco sale laws. Among boys, no associations were found between scales measuring the restrictiveness of substance use and sexual orientation disparities in substance use; two indicators of restrictiveness were associated with smaller sexual orientation disparities in binge drinking and three indicators were associated with smaller sexual orientation disparities in drunk driving. Among heterosexual but not sexual minority boys, higher tobacco taxes were associated with lower prevalence of smoking.

This study raises several questions for future research on structural factors that may explain sexual orientation disparities in substance use behaviors among youth. For instance, future work is needed to understand the gender differences in response to structural stigma among sexual minorities. Sexual minority girls may have greater rejection sensitivity than sexual minority boys and thus may be more likely to use substances in response to structural stigma; however, further research is needed to test this hypothesis. In addition, the current study lacked data on the implementation of substance use policy environments, which may have masked important effects. A study of SUPE that includes measures of implementation of restrictive substance use policies is therefore needed to expand the work reported herein. Finally, a better understanding of gender identity and expression is needed; while the YRBSS measures “sex” and not gender or gender expression in these datasets, more information about these topics will help to understand how these factors may play into experiences of structural stigma and substance use policy environments.


  • thumnail for Frazer_columbia_0054D_15657.pdf Frazer_columbia_0054D_15657.pdf application/pdf 635 KB Download File

More About This Work

Academic Units
Sociomedical Sciences
Thesis Advisors
Hatzenbuehler, Mark L.
Ph.D., Columbia University
Published Here
January 17, 2020