Mapping Structural Constraints on HIV Prevention: A Network Analysis of Food Insecurity, Housing Instability, and Barriers to Care in Miami, FL

Communities in Miami face persistent disparities in HIV prevention, including low uptake of pre-exposure prophylaxis (PrEP) and gaps in routine HIV testing. These inequities are shaped by co-occurring structural conditions, such as food insecurity, housing instability, discrimination, transportation barriers, and broader barriers to care. Understanding how these conditions interconnect may inform approaches to sustained engagement in HIV prevention. We examined how structural and sociodemographic conditions and food insecurity were interconnected and reflective of overall structural vulnerability. We conducted a cross-sectional network analysis using data from 2755 surveys collected during Phases 1 and 2 of the Five Point Initiative (FPI), a community-engaged, place-based sexual health implementation strategy in Miami. Partial correlation networks were estimated using the EBICglasso algorithm across key structural barriers to HIV prevention. We calculated network centrality metrics (strength, betweenness, closeness) to characterize the relative connectivity and positioning of conditions within the network. Secondarily, we used traditional mixed correlations to characterize bivariate associations among structural and sociodemographic factors. Multiple regression also examined associations of barriers to care, discrimination, ethnicity, education, housing instability, and income with food insecurity. Additional models examined whether these differed by race. Food insecurity emerged as having the highest strength, closeness, and betweenness centrality within the network and showed the strongest association with barriers to care. Greater food insecurity was also associated with higher discrimination, lower income, unstable housing, and Latino ethnicity. Additional significant connections among variables were also found. Similarly, in secondary bivariate analyses, food insecurity was positively correlated with barriers to care (r = 0.331), discrimination (r = 0.251), ethnicity (r = 0.141), and housing instability (r = 0.125), and negatively correlated with income (r = −0.150) and education (r = −0.079). The multivariable regression model predicting food insecurity was statistically significant, F(6, 2612) = 65.8, p < 0.001, with association directions generally consistent with the network findings. Race did not significantly moderate most individual associations with food insecurity. Race by housing interaction was significant (p = 0.03), indicating that participants who identified as Black or African American with insecure housing were more likely to experience food insecurity. Overall, the findings indicate that food insecurity occupies a prominent position within an interconnected system of social and structural constraints around HIV prevention. Structural barriers to HIV prevention operate as a mutually reinforcing system rather than isolated challenges. Addressing highly connected constraints, particularly food insecurity, housing instability, and barriers to care, through place-based, community-engaged strategies may yield compounded benefits across the prevention continuum. These findings highlight potential priorities for structural intervention in Ending the HIV Epidemic (EHE) priority jurisdictions.