Evaluation of individual attributes associated with shared HIV risk behaviors among two network-based studies of people who inject drugs

TL;DR

This study investigates how social and individual attributes influence HIV risk behaviors among people who inject drugs, revealing that similarities in demographics and social factors increase the likelihood of risk behaviors within networks.

Contribution

It identifies key individual and social network factors associated with HIV risk behaviors among PWID using exponential random graph models across two datasets.

Findings

Similarity in sex, race/ethnicity, and living conditions increases risk behavior likelihood.

Individual risk factors like housing instability are linked to network structure.

Social influences and community context are crucial for intervention strategies.

Abstract

Social context plays an important role in perpetuating or reducing HIV risk behaviors. This study analyzed the network and individual attributes that were associated with the likelihood that people who inject drugs (PWID) will engage in HIV risk behaviors with one another. We analyze data collected in the Social Risk Factors and HIV Risk Study (SFHR) and Transmission Reduction Intervention Project (TRIP) to perform the analysis. Exponential random graph models were used to determine which attributes were associated with the likelihood of people engaging in HIV risk behaviors, such as injection behaviors that are associated with one another, among PWID. Results across all models and across both data sets indicated that people were more likely to engage in risk behaviors with others who were similar to them in some way (e.g., were the same sex, race/ethnicity, living conditions). In both SFHR and TRIP, we explore the effects of missingness at individual and network levels on the likelihood of individuals to engage in HIV risk behaviors among PWID. In this study, we found that known individual-level risk factors, including housing instability and race/ethnicity, are also important factors in determining the structure of the observed network among PWID. Future development of interventions should consider not only individual risk factors, but communities and social influences leaving individuals vulnerable to HIV risk.