HIV Risk Behaviors in Context

Background and Purpose: Research regarding HIV and drug use has focused primarily on risky substance use associated with impaired judgment resulting in risky sex and unsafe injection drug practices.  These individual-level behaviors have led to an increase in HIV among subgroups of the population however, multiple studies suggest that underlying social-structural and contextual conditions inextricably constrain and enable individual agency in HIV transmission and HIV prevention.  Reducing HIV transmission behaviors among drug users continues to be a challenge. Substance users who accessed alcohol and drug treatment programs were less likely to test positive for HIV and a significant percentage of injection drug users who accessed harm reduction services lowered their injection risk. However few individuals who use alcohol and/or illicit substances access programs. Modest efficacy has been reported for interventions aimed at reducing HIV sexual risk behaviors among drug users. Promoting the use of condoms for sexually active individuals has been the focus of prevention efforts yet only one quarter of adults reported using condoms at the time of their most recent sexual encounter. Identifying the contributions of social-structural and contextual conditions on HIV transmission will clarify the importance of these factors for prevention, inform policy and help to set priorities for achieving reduction in the disparities of  the HIV burden. The purpose of this study was to employ a socioecological approach to examine longitudinally the social and physical factors that contributed to changes in drug and sexual HIV-risk behaviors among out-of-treatment substance users in an urban community.

Methods: Data were collected at two time points from a cohort of 537 out-of-treatment substance users with NIDA’s Risk Behavior Assessment at baseline and the Risk Behavior Follow-up Assessment at 3-11 months. Scales were developed using conceptual indices of HIV risk behavior for the 30 days prior to first interview and 30 days prior to their last interview. 

 Social and ecological factors included housing, income, employment, cohabiting, treatment, sexual partners, drugs and sexual activity, and jail time. Factors significant at the bivariate level were entered into two regression models; a Poisson model for drug risks and a Zero-inflated Negative Binomial model for sex risks.

Results: Two variables significantly predicted reduced sex-risk behaviors: income and not living with a sexual partner. Four variables were found to significantly reduce drug risk at follow up: income, employment, housing, and prior treatment. Participant perception acquiring HIV at baseline predicted a significant decrease in drug-related risk at follow-up (p=.0158, d=.72).  However, perception of risk did not significantly predict sexual or injection risks. 

Conclusions and Implications: To achieve prevention goals, we need multilevel interventions through direct service practitioners, policy reform, individual behaviors, and community planning. These findings replicate prior research regarding the influence of income, employment, shelter, and drug treatment experience associated with a reduction of HIV-related risks among drug users.  Cumulative environmental disadvantage may influence HIV risk behaviors and further research is needed. Interventions aimed at reducing sexual risk behaviors must address the contextual factors of economic vulnerability, as well as those of individual agency.