PROJECT SUMMARY/ABSTRACT Heart failure (HF) with preserved ejection fraction (HFpEF) represents ≈50% of HF, with limited methods for prevention and treatment. Current approaches to detect/treat HFpEF rely on cardiac phenotypes obtained at rest, missing reserve capacity impairments in multiple organ systems central to HFpEF that are uniquely revealed through exercise. In the first R01 period, we performed 3117 cardiopulmonary exercise tests (CPETs) in the Framingham Heart Study (FHS) to quantify cardiorespiratory fitness, individual exercise responses, and their metabolomic signatures. We defined a “normal” metabolic response to exercise and reported novel measures of impaired fitness and HF risk. Strikingly, ≈1 in 4 asymptomatic community-dwelling FHS participants displayed fitness impairment that overlapped with individuals with hemodynamically confirmed HFpEF studied in our clinical referral lab. Metabolites related to HFpEF physiology were also associated with exercise responses in FHS. These results underscore a critical molecular and phenotypic overlap between clinical HFpEF and the early stages of its development, providing a unique opportunity to define the origins of HFpEF and reveal new targets for its screening/prevention. Proteins are potentially targetable biomarkers that provide a readout of specific pathways relevant to multiple organs, have direct correlation with genetics, and have a well-defined process for assay development. The hypothesis of this renewal is that molecular pathways related to precise HFpEF phenotypes captured during exercise will specify mechanisms underlying HFpEF susceptibility long before its usual clinical detection. We leverage 2 unique samples developed in the first period: (1) deeply-phenotyped HFpEF patients with hemodynamic measures during CPET (MGH-ExS, N=500) and (2) FHS participants with CPET and plasma samples at rest/peak exercise (N=1500). We will study a broad circulating proteome (>3000 proteins at rest/peak exercise) in relation to precise exercise phenotypes, in silico bioinformatics, and human genetics to specify proteomic signatures of exercise response in 3 aims. In Aim 1, we identify pathways of organ- specific responses to exercise in HFpEF (using the rest proteome) and their relation to subclinical phenotypes central to HFpEF susceptibility in the community. In Aim 2, we will quantify changes in the circulating proteome with acute exercise and evaluate how exercise-induced changes differ in the presence of HFpEF (MGH-ExS vs. FHS) and HFpEF risk factors. In Aim 3, we measure association of proteomic signatures of HFpEF phenotypes with incident HF and cardiovascular disease (CVD) in racially diverse primary prevention cohorts. We also construct genetic instruments of implicated proteins (pQTLs) for association with HF/CVD in large biobanks using Mendelian randomization. This application unites profiling of broad pathways with relevance to HFpEF with unique, precise exercise phenotypes coll...