Factors Associated with Drug Use in LGB Adolescents: A Structural Equation Model for Exploring Minority Stress Theory

Introduction: Lesbian, Gay and Bisexual (LGB) adolescents are at disproportionate risk for substance use (Marshal, et al., 2008), including the use of cigarettes, alcohol, marijuana, cocaine and ecstasy (Moon et al., 2007), multiple substances at once (Garofalo, et al., 1998) and increase their use more rapidly as they progress through life (Marshal, et al., 2008). Most prevention programs focus on reducing risk factors and increasing resilience (Hawkins & Catalano, 1992), and have more recently focused on culturally based risk (for example, Cervantes, Goldbach & Santos, 2011).  Despite a documented need, little research exists to explore culturally based risk factors for substance use in LGB youth and their combined impact on substance use behaviors (Goldbach, under review) and no interventions currently exist in the market of evidence-based programs (NREPP, 2011).

Methods: The current study is a secondary analysis of a national online survey (OutProud, 2000) of 2,237 adolescents (11-17 years old).  Youth were recruited through online websites and magazine advertisements that target LGB youth.  The survey covered a variety of topics including substance use behaviors, school, family, and peer experiences.  Constructs of culturally based risk for substance use at both the individual and community level in LGB youth and identified through previous research (Goldbach, under review) were used as a framework for identifying indicator variables in the dataset.  Structural equation modeling (SEM) with latent variables was the overarching approach to analysis (Kline, 2004), beginning with a confirmatory factor analysis (CFA) to verify indicators loaded onto interpretable latent factors.  Regression paths were systematically added among latent and observed variables to explore interrelationships.  Non-significant paths were removed until a parsimonious model was reached.

Results: The CFA resulted in a measurement model with all indicators loading significantly (P < 0.01) onto their respective latent factors and excellent fit (CFI=0.983, TLI=0.978, RMSEA=0.032).  The structural model also indicated good fit.  Two latent factors were associated with marijuana use, internalized homophobia (b=0.116, t=2.096, p<0.05) and community connectedness (b=0.314, t=5.24, p<0.001).  Community connectedness was also associated with lower rates of internalized homophobia (b=-0.331, t=-4.636, p<0.001).  Three factors were predictive of psychological distress, violence/victimization (b=0.215, t=6.857, p<0.001), internalized homophobia (b=0.155, t=-2.359, p<0.05), and outness (b=0.155, t=2.796, p<0.01).

Conclusions: Findings suggest that LGB youth experience minority including violence/victimization, psychological distress, and internalized homophobia, which are associated with marijuana use.  Connection to the gay community was associated with reduced psychological distress but increased marijuana use, indicating that the community provides opportunities for both risk and support.  Further research is needed to explore the relationships between the community connection, psychological distress and substance use.