Substance Use Typologies Among Young People Experiencing Homelessness in Seven Cities across the United States: A Latent Class Analysis

Background:

Over 3.5 million young adults in the United States experience some form of homelessness annually. A large body of research shows high rates of lifetime substance use among young adults experiencing homelessness (YEH); for example, 89% report alcohol use, 77% marijuana use, and 48% report methamphetamine use. In fact, substance use among YEH is 2 to 3 times more prevalent compared to their housed counterparts. Substance use among YEH is a serious public health issue, with deleterious impacts in multiple domains of health and wellness. Indeed, YEH with substance use experience longer periods of homelessness, trauma, depression, anxiety, and engagement in high-risk behaviors such as injection drug use or sexual intercourse while under the influence of substances. Despite robust relationships between substance use and myriad risk factors, less is known about the use of multiple substances (polysubstance use) and associated risk factors among this vulnerable population. Therefore, the present study aimed to identify subgroups of YEH based on polysubstance use and the linkages to sociodemographic and behavioral-health characteristics.

Methods:

From June 2016 to July 2017, 1,426 YEH (aged 18–26 years) were recruited from seven cities (Houston, Los Angeles, Denver, Phoenix, New York City, St. Louis, San Jose). Participants provided information via a self-administered electronic survey on substance use, mental health, trauma, and sexual-risk behaviors. The majority of YEH identified as Black (37.3%), cisgender (92.8%), and heterosexual (69.2%). On average, YEH were 20.9 years (SD=2.1). This study employed latent class analysis to identify subgroups of YEH according to their substance use. Multinomial logistic regression analyses were then conducted to identify the sociodemographic and risk characteristics associated with substance use class membership.

Results:

With regard to substance use, 17.5% of participants reported methamphetamine use, 17.7% cocaine, 8.2% crack, 15.4% ecstasy, 51.8% alcohol, 60.7% marijuana, 20% prescription drug use, and 8.6% had engaged in injection drug use. Four latent classes of YEH substance use were identified: (1) high alcohol and marijuana use/ moderate stimulant and prescription drug use; (2) high alcohol and marijuana use; (3) high substance use; and (4) low or no substance use. Multinomial logistic regression models indicated that geographic location, gender, race/ethnicity, mental health, trauma history, and sexual risk behavior were significant correlates of classes of substance use among YEH.

Conclusion:

Findings demonstrate that distinct groups of YEH can be identified by their polysubstance use patterns and that certain sociodemographic and risk characteristics are significantly associated with substance use class membership. For example, trauma and mental health were significant correlates of substance use among YEH, which may indicate that YEH may engage in polysubstance use as a way to cope with trauma experiences and mental health. Thus, findings offer important implications for the prevention and treatment of substance use among YEH. Screening protocols should consider traumatic experiences, sexual risk behaviors, and mental health history as a useful indicator of polysubstance use. In addition, social work practice should prioritize incorporating trauma-informed care when working with YEH.