Latent class analysis of substance use typologies associated with mental and sexual health outcomes among sexual and gender minority youth

Little is known about sexual and gender minority youth (SGMY) who have sex with men's unique patterns of substance use, even though they are at risk for substance use and adverse mental and other health outcomes. We used latent class analysis to examine typologies of substance use and multinomial logistic regression to investigate mental health outcomes (depression and anxiety) and HIV/STI testing correlates associated with different classes of substance use in a sample of SGMY who have sex with men in the USA and use substances (n = 414) who participated in an online survey. The average age was 22.50 years old (SD = 3.22). A four-class solution was identified representing: 'depressant and stimulant use' (3.4%), 'high polysubstance use' (4.6%), 'low substance use with moderate cannabis use' (79.2%), and 'high cannabis, stimulant and alcohol use' (12.8%). Membership to a specified substance use class varied by age, previous arrest, gender identity, anxiety, and lifetime HIV testing. Multivariate logistic regression results indicated that participants in the high polysubstance use (AOR = 5.48, 95% CI 1.51, 19.97) and high cannabis use class (AOR = 3.87, 95% CI 1.25, 11.94) were significantly more likely than those in the low substance use with moderate cannabis use class to report previous arrest. Those in the high polysubstance use class were also significantly less likely to have been tested for HIV than those in the low substance use with moderate cannabis use class (AOR = 0.21, 95% CI 0.05, 0.93). Findings will guide the development and implementation of tailored approaches to addressing the intersection of substance use and HIV risk among SGMY.