In the United States, disparities in HIV between black and white men who have sex with men (MSM) are large and unexplained. Individual behavioral reasons for these disparities have been generally dismissed as possible explanations. Network explanations have proven useful in understanding persistent disparities in STI burdens in other populations, and have been hypothesized to explain racial/ethnic disparities in US MSM. However, these hypotheses have not been tested using modeling approaches for two reasons. First, there have historically been inadequate data to parameterize models with respect to critical network characteristics, such as concurrency, transitivity, and assortative mixing by key characteristics. Second, model structures have not been developed that account for the unique aspects of MSM sexual networks. We propose to develop this understanding by analyzing new data on MSM sexual networks from multiple studies conducted by the investigators in Atlanta, a metropolitan area with large populations of both black and white MSM and significant racial disparities in HIV. We will recall participants in current and recent studies to collect additional data as necessary. Using these new network parameters, we will extend an existing stochastic simulation model of HIV transmission in MSM to reflect these network attributes. This approach will use STERGMs (separable temporal exponential random graph models), which can handle dynamic network data in the form of cross-sectional network structure combined with information on relational duration or dissolution patterns. We will use the expanded, newly parameterized model for two purposes. First, we will use the model to attempt to describe the role of network parameters in black/white disparities in HIV incidence among US MSM. Second, we will use the model to test which modifiable network parameters might have the most potential in reducing HIV incidence, if modified among MSM populations. Completion of these activities is expected to result in the testing of hypotheses about reasons for black/white disparities in HIV among US MSM, modeling tools that will be useful to further evaluation of MSM networks, and specific, proposed targets for new prevention interventions for this most highly impacted community in the US HIV epidemic.
Our Specific Aims are: (1) to develop a comprehensive set of individual, dyadic, and triadic network parameters, accounting for transmission potential, for use in an agent-based HIV transmission model for black and white MSM in metropolitan Atlanta;(2) to expand an existing stochastic, agent-based HIV transmission model for MSM to include code for modeling network features for black and white MSM;(3) to utilize this enhanced model and parameters to (a) determine whether racial disparities for HIV observed among Atlanta-area MSM can be explained using existing hypotheses for their sources in combination with our measured network phenomena;(b) identify the highest- impact network properties for addressing racial/ethnic disparities in HIV among US MSM.
Men who have sex with men (MSM) are the most heavily impacted risk group in the US HIV epidemic;within this community, large racial differences in HIV prevalence exist, which are not explained by obvious differences in sexual behavior. In this study, we will use existing and newly collected network data, in combination with mathematical modeling, to consider whether multiple aspects of the structure of sexual networks can help explain this disparity. This work offers the potential to answer the persistent question of why some MSM communities have high HIV levels, and to identify future strategies for reducing those levels.
