PROJECT SUMMARY Major depressive disorder is one of the top 5 contributors to years lived with disability among adults in sub- Saharan Africa. Numerous studies have suggested a link between social networks and depression, but none have measured actual changes in social network structure in relation to depression or have examined the potential mediating role of inflammation. This is an important gap in the literature because negative social interactions increase production of pro-inflammatory cytokines, and inflammation is believed to play an important role in depression pathogenesis. Our scientific objective is to estimate the causal effect of social networks on depression and to assess the extent to which inflammation causally mediates this effect. The cytokine theory of depression is consistent with our strong preliminary data from rural Uganda showing that: elevated C-reactive protein and interleukin-6 values correlate strongly with depression; reversals in tryptophan catabolism, mediated by the activity of indoleamine 2,3-dioxygenase-1, explain improvements in depression symptom severity over the course of early HIV antiretroviral treatment; and social networks provide a moderating, or “buffering,” effect against the adverse mental health impacts of food insecurity. The proposed study will leverage the resources of an ongoing, whole-population social network cohort study in Mbarara and a long-term partnership with Mbarara University that is presently focused on characterizing the contributions of social networks and HIV stigma to HIV care cascade outcomes. The sample of opportunity includes 3 waves of existing survey data and biological specimens from this whole-population sample; the plasma samples are currently in -80C storage. During the three-year study we propose here, we will test our currently held 3 years of frozen plasma samples for inflammatory markers of interest, and collect an additional 3 years of survey and biomarker data (total N=1,776, T=6). We will accomplish two specific aims: (1) Estimate the causal effect of social networks on depression symptom severity; and (2) assess the extent to which inflammatory activity mediates the causal effect of social networks on depression. Our expected contribution is detailed mechanistic insight into the causal pathways linking social networks, inflammation, & major depressive disorder. Our key innovation is that we will apply state-of-the-art methods of causal inference to whole-population social network cohort data in rural Uganda to identify the causal effect of social networks on depression & the causal mediation effects of inflammation. This contribution will have significant public health impact by providing novel data linking social network structure to depression and therefore inform future social network-based interventions; investigation of inflammatory mediators will inform the use of anti-inflammatory depression treatment strategies or of inflammatory testing to predict depress...