Identifying actionable multilevel intervention strategies to improve cardiovascular health among people with HIV - People with HIV (PWH) are living longer and experiencing similar chronic diseases as seronegative counterparts, but at earlier ages and with greater incidence. While HIV alone has been linked to the acceleration of non-communicable diseases, the increased CVD risk is not fully explained by antiretroviral therapy or detectable HIV viral load. Excess CVD risk among PWH may be related to heterogeneity of social drivers of health, including healthcare access, economic stability, social support, and neighborhood conditions, that span the social-ecological model of societal, community, interpersonal, and individual domains. The complementary frameworks of social drivers and social-ecological model can aid in understanding health outcomes driven by structural and intermediary factors. Clustering of multiple social drivers across social-ecological domains can interact synergistically to disproportionately impact PWH. PWH experience overlapping social, behavioral, and healthcare access barriers that span social-ecological domains. Sparse research has used advanced approaches to identify the complex clustering and relationships among the aspects of social drivers across social-ecological domains that accelerate CVD, qualified as syndemics, to inform multilevel interventions. This gap has limited the development of actionable targets for realistic interventions, particularly those that extend beyond individual behaviors and clinical treatment to also address interpersonal relationships, community contexts, and broader social structures. New advances in Machine Learning (ML)/Artificial Intelligence (AI) can enhance traditional epidemiologic approaches. With multidimensional data – integrated from electronic health records, socio-ecological databases, and physical measurements – AI can be used to efficiently identify syndemic patterns that, if realistically modified, can aid in the design and implementation of interventions tailored to syndemic patterns and contexts. The overarching goal of this project is to systematically identify actionable targets for multilevel interventions that improve cardiovascular health outcomes among PWH through the complementary frameworks of social drivers of health, social-ecological model, and syndemics theory. Therefore, the current aims are to (1) identify syndemic phenotypes of social drivers and quantify the association between these phenotypes and cardiovascular comorbidities among PWH using AI-enhanced epidemiology, (2) propose syndemic-based, actionable multilevel intervention targets across social-ecological domains of influence through qualitative approaches and partnerships with a Community Advisory Panel and stakeholder input and (3) test actionable multilevel interventions using traditional and AI-enhanced counterfactual models. Our team includes experts in HIV syndemics, CVD, social and behavioral epidemiology, data integration, causal inference, ML/AI, qualitative methodology, and community engagement. This work will have impact by identifying and testing actionable, syndemic-focused intervention targets to inform solution-oriented implementation strategies that improve cardiovascular health across affected populations, incorporating community stakeholder expertise.
