Access to housing subsidies, housing status, drug use and HIV risk among low-income U.S. urban residents

BACKGROUND Much research has shown an association between homelessness and unstable housing and HIV risk but most has relied on relatively narrow definitions of housing status that preclude a deeper understanding of this relationship. Fewer studies have examined access to housing subsidies and supportive housing programs among low-income populations with different personal characteristics. This paper explores personal characteristics associated with access to housing subsidies and supportive housing, the relationship between personal characteristics and housing status, and the relationship between housing status and sexual risk behaviors among low-income urban residents. METHODS Surveys were conducted with 392 low-income residents from Hartford and East Harford, Connecticut through a targeted sampling plan. We measured personal characteristics (income, education, use of crack, heroin, or cocaine in the last 6 months, receipt of welfare benefits, mental illness diagnosis, arrest, criminal conviction, longest prison term served, and self-reported HIV diagnosis); access to housing subsidies or supportive housing programs; current housing status; and sexual risk behaviors. To answer the aims above, we performed univariate analyses using Chi-square or 2-sided ANOVA's. Those with significance levels above (0.10) were included in multivariate analyses. We performed 2 separate multiple regressions to determine the effects of personal characteristics on access to housing subsidies and access to supportive housing respectively. We used multinomial main effects logistic regression to determine the effects of housing status on sexual risk behavior. RESULTS Being HIV positive or having a mental illness predicted access to housing subsidies and supportive housing, while having a criminal conviction was not related to access to either housing subsidies or supportive housing. Drug use was associated with poorer housing statuses such as living on the street or in a shelter, or temporarily doubling up with friends, acquaintances or sex partners. Living with friends, acquaintances or sex partners was associated with greater sexual risk than those living on the street or in other stable housing situations. CONCLUSIONS Results suggest that providing low-income and supportive housing may be an effective structural HIV prevention intervention, but that the availability and accessibility of these programs must be increased.