The Effect of Social Networks and Social Constructions on HIV Risk Perceptions.

Roughly 1 in 7 people living with HIV in the United States is unaware of their sero-status, signaling that individuals may be underestimating their risk for HIV. Few studies have examined the effect of socio-structural and socio-cognitive factors on HIV risk perceptions. This analysis identifies individual, interpersonal and network influences on HIV risk perceptions among high-risk heterosexuals. Data come from the Colorado Springs study, a CDC-funded project focused on HIV transmission among high-risk heterosexuals. Using social network data, analyses were first conducted at the individual-level using a partial proportional odds regression to identify predictors of self-perceived HIV risk. Next, multivariate binary logistic regression using GEE was used to examine  predictors of perceptions of network member's  HIV risk. Interpersonal characteristics such as perceptions of network member's HIV risk, racial homophily, and engagement in multiplexity (co-occurrence of drug-use, needle sharing and sex within relationships) were significantly associated with respondents' self-perceived HIV risk. Factors associated with perceptions of network member's HIV risk include self-perceived HIV risk, emotional closeness within relationships, and density of drug ties. Analyses found HIV risk perception is the product of not only individual-level factors, but also interpersonal and social network processes. We also found a reciprocal relationship between individuals' perceptions of their own risk and the risk of their associates/network members. Findings highlight the need for understanding risk perception as a function of interpersonal relationships, social constructions, including socio-cognitive processes.