Identifying Social-Epidemiological Roles Associated with Viral Exposure Using Regular Equivalence Blockmodeling

Identifying individuals at high risk of infection is critical to guiding interventions during infectious disease outbreaks. Network centrality can characterize infection risk, but its utility varies across pathogens and social systems. We applied a new approach – equivalence-based blockmodeling – to identify social-epidemiological roles associated with viral exposure among 1,297 adults in northeast Madagascar. Roles were determined based on social networks derived from reported shared free time and exchanges of food and farmwork. We identified three distinct roles, including individuals with many reciprocated social ties (Popular), individuals who sent many ties with few reciprocated (Core), and individuals who had few connections (Periphery). To assess whether the roles covaried with viral exposure, Phage ImmunoPrecipitation Sequencing was performed using dried blood spot samples and VirScan to determine exposure to 337 virus species and subtypes. Membership in the Popular ( ß [95% CI]: 0.29 [-0.06-0.63]) and Core ( ß [95% CI]: 0.36 [0.12-0.60]) roles was associated with greater viral exposure than membership in the Periphery role. Roles performed better at predicting exposure than single measures of centrality. Equivalence-based blockmodeling extends the utility of network centrality for characterizing infection risk and provides new insights into how social roles may relate to both pathogen exposure and susceptibility to infection.