Recent Violence, Violent Peers and HIV Risk Behavior of Homeless Youth

Background: Homeless youth (HY), a large and growing population (an estimated 1.6 million per year), remain disproportionately susceptible to violent victimization and perpetration. While previous studies have documented a linkage between experiencing violence and HIV risk among this population, very few studies have tried to understand how experience of different types of violent episodes as well as involvement with violent peers and partners over the life course might exacerbate their exposure to HIV risk.  Therefore, this current study attempts to assess how childhood trauma, interpersonal violence, intimate partner violence, and violent peer networks are related to the HIV risk behaviors of homeless of HY.

Method: A sample of 360 Los Angeles area drop-in service seeking HY completed a self-administered questionnaire and social network interview. Participating youth were asked a series of questions regarding HIV risk behaviors including recent unprotected sex, concurrent sex, sex under the influence, exchange sex, methamphetamine (meth) use and Injection drug use (IDU). Univariable logistic regressions tested the relationship between violence variables (childhood trauma, interpersonal violence, intimate partner violence, and proportion of violent peer in personal networks) and HIV risk outcome variables. Multivariable logistic regressions tested the overall effect of violence controlling for age, gender, race, sexual identity, current living situation and data collection site.

Results:Univariable logistic regressions revealed that experience of IPV was related to unprotected sex (OR=1.96, p<.01) and IDU (OR-1.96, p<.05); childhood trauma was related to meth use (OR=2.34, P<.001), IDU (OR=2.74, p<.01), sex under the influence (OR=2.43, p<.001) and exchange sex (OR=1.86, p<.05); interpersonal violence was related to unprotected sex (OR=2.39, p<.001), concurrent sex (OR=2.88, p<.001), meth use  (OR=2.85, p<.001), IDU (OR=3.49, p<.01) and sex under the influence (OR=1.87, p<.05). Recent peer violence was related to unprotected sex (1.67, p<.05). Multivariable logistic regressions revealed that interpersonal violence remained significant for unprotected sex (OR=2.08, p<.01), concurrent sex (OR=2.98, p<.001), meth use (OR=2.86, p<.05) and IDU (OR=2.86, p<.01). Childhood trauma also remained significant for concurrent sex (OR=1.81, p<.05), sex under the influence (OR=1.71, p<.01) and exchange sex (OR=2.04, p<.05).

Discussion: Violent experiences throughout the life course are closely related to current HIV risk behaviors among homeless youth. Results suggest that while most types of violence have individual effects; childhood trauma and interpersonal violence are the most salient. This research suggests that reducing violence and addressing trauma from violent experiences may impact homeless youth HIV risk behaviors. HIV prevention interventions would benefit from considering the effects of these experiences. Overall, violence prevention interventions within homeless youth should be prioritized.