Substance Use Behaviors Among LGBTQ Youth of Color: Applying a Novel Method in Two Large Samples to Identify Intersecting Social Positions Bearing the Greatest Burden

ABSTRACT

Purpose: The global COVID pandemic, social uprisings, and a wave of discriminatory policy proposals have highlighted the ways in which structural oppression contributes to health disparities facing youth of color and those identifying as LGBTQ. Young people living at the intersection of multiple types of oppression face the greatest burden, yet also have unique strengths and supports. Existing research has demonstrated persistent substance use disparities across sexual orientation, gender identity, and racial/ethnic groups – as individual categories. However, very little research has examined substance use among those with multiple stigmatized identities. Capitalizing on two very large datasets and a novel analytic technique, this study seeks to identify groups with the highest prevalence of past 30-day alcohol, e-cigarette, and marijuana use. This first step in a larger project will determine key intersecting identities for qualitative interviews regarding interpersonal and community supports that can reduce health disparities. Methods: Data come from the 2019 Minnesota Student Survey and the 2017-2019 California Healthy Kids Survey, two surveillance programs with a combined sample of 892,664 students in grades 6-12. Data were harmonized across sources to create compatible variables including race/ethnicity (non-Hispanic Native American, Asian/Pacific Islander, Black/African/African American, White, Multiracial;Latina/x/o), sexual orientation (straight, gay/lesbian, bisexual, questioning, something else [e.g. pansexual, queer]), gender identity (cisgender, transgender/gender diverse [TGD], questioning), sex assigned at birth (male, female), state, and past 30-day substance use (yes/no for alcohol, e-cigarettes, marijuana). Exhaustive Chi-square Automatic Interaction Detection (CHAID) analysis, a decision tree approach, was used to examine all interactions among social positions with the goal of identifying distinct groups with significantly different rates of substance use behaviors (Bonferroni adjusted p<.05). The groups with the highest prevalence for each substance were examined. Results: The overall prevalence of past 30-day substance use was 10.4% for alcohol, 9.7% for e-cigarettes, and 9.7% for marijuana, with substantial disparities across intersecting groups. For example, although 10.5% of Latina/x/o-identified youth and 20.8% of TGD-identified youth reported drinking alcohol, Latina/x/o TGD youth were among those with the highest prevalence of use, particularly those who also identified with a newer sexual orientation label (e.g. pansexual, queer) and were assigned male at birth (26.2%) or Latina/x/o TGD youth who did not indicate their sexual orientation (31.7%). This pattern was also evident for e-cigarette and marijuana use. Similarly, Black TGD youth had significantly higher rates of alcohol (26.9%), e-cigarette (29.2%, in California), and marijuana use (24.4%, straight-identified;29.5%, missing sexual orientation). Conclusions: Using the power and diversity of large population-based datasets and an innovative analytic technique specifically recommended for studies of intersectionality, we found significant disparities in substance use, with the burden varying by unique intersecting marginalized identities. This approach is recommended to examine disparities in groups often treated as homogeneous, as a precursor to developing relevant and appropriate prevention strategies. Further research is needed to identify structural factors contributing to these high rates. Clinicians, educators, and others working with youth should address intersecting types of stigma and oppression that may contribute to substance use. Sources of Support National Institute of Minority Health and Health Disparities grant #R01MD015722.