Network Homophily and Substance Use Among Young Adults Experiencing Homelessness

Background and Purpose: Substance use (SU) among young adults experiencing homelessness (YAEH) has been shown to be associated with violence, HIV risk and poor housing outcomes. Prior work has shown networks impact SU. Guided by homophily theory, which suggests people tend to form and maintain relationships with similar others, this study extends prior research by identifying how individual SU profiles (non-users, alcohol- and cannabis-only users, and polysubstance users) reflect homophilous behaviors across home-, street-, and family-based ties.

Methods: Data were drawn from a cross‑sectional study of YAEH recruited from drop‑in centers in Los Angeles, California and Saint Louis, Missouri (N = 686), ages 16-24, who had experienced homelessness in the past 12 months. Interviewer-administered surveys assessed participants’ past-30-day substance use, categorized as no use, alcohol or cannabis use, and polysubstance use (PSU; three or more substances). Social network data captured up to 20 alters with whom participants interacted in the past 12 months, classified as home-, street-, or family-based. Participants reported each alter’s past-year substance use, and binary indicators reflected the presence of non-using, alcohol/cannabis-using, and polysubstance-using (PSU) alters within each network type. Multinomial logistic regression models examined associations between youth past-30-day substance use and network composition, adjusting for lifetime homelessness, gender identity, race, ethnicity, and sexual orientation; results are presented as relative risk ratios (RRRs).

Results: Having non-using home-based alters was associated with lower relative risk of alcohol/cannabis-only use (RRR = 0.53, p < .001) and polysubstance use (RRR = 0.33, p < .001), while having alcohol/cannabis-using home alters was associated with higher risk of alcohol/cannabis-only use (RRR = 3.06, p < .001) and polysubstance use (RRR = 4.80, p < .001). Polysubstance-using home alters were associated with higher risk of polysubstance use (RRR = 2.02, p < .001). Having non-using street alters was associated with lower risk of alcohol/cannabis-only use (RRR = 0.40, p < .001) and polysubstance use (RRR = 0.25, p < .001), while having alcohol/cannabis-using street alters was associated with higher risk of alcohol/cannabis-only use (RRR = 2.69, p < .001) and polysubstance use (RRR = 4.32, p < .001). Polysubstance-using street alters were associated with higher risk of polysubstance use (RRR = 3.35, p < .001). Having non-using family alters were associated with lower risk of polysubstance use (RRR = 0.73, p < .001), while alcohol/cannabis-using family alters were associated with higher risk of alcohol/cannabis-only use (RRR = 1.63, p < .001) and polysubstance use (RRR = 3.28, p < .001). Polysubstance-using family alters were associated with lower relative risk of alcohol/cannabis-only use (RRR = 0.64, p = .031).

Conclusions and Implications: Findings suggest homophily shapes SU within YAEH networks and highlight how relationship patterns influence where substance‑related risks and supports are concentrated. Street-based networks appear to concentrate and reinforce polysubstance use, whereas family- and home-based networks are more often associated with alcohol/cannabis-only use, reflecting comparatively lower-risk contexts. Social network-driven, peer‑led harm reduction interventions engaging youth as central actors may reduce SU-related harms within their networks and promote safer SU behaviors.