Heart failure (HF) disproportionately affects the elderly who predominantly develop HF with preserved left ventricular (LV) ejection fraction (HFpEF), for which no efficacious therapies exist. Pulmonary hypertension (PH) – the most clinically recognized expression of pulmonary vascular dysfunction (PVD) – is a risk factor for incident HF and is present in upto 83% of patients with prevalent HFpEF, among whom it portends worse outcomes. PVD is therefore an attractive therapeutic target in HFpEF, but its pathophysiology is complex with variable contributions from elevated left atrial pressure, pulmonary parenchymal injury, and intrinsic pulmonary vascular dysfunction. A critical barrier to understanding PVD in HFpEF is a lack of knowledge regarding the anatomic alterations in the pulmonary vasculature underlying abnormal hemodynamics. The investigative team has pioneered development and validation of advanced image processing pipelines to quantify pulmonary venous and arterial remodeling on non-contrast chest computerized tomography (CT) scans. Their published and preliminary data from smokers in the NHLBI-funded COPDGene study show that pulmonary vascular remodeling associates with RV dysfunction and worse functional capacity, and is more frequently observed in HF. They now propose to extend these findings to a community-based cohort of older adults to define the role of pulmonary vascular remodeling in the development of HFpEF. This proposal’s central hypothesis is that activation of pro-inflammatory and pro-fibrotic pathways promotes pulmonary vascular remodeling, partially via concomitant LV dysfunction and pulmonary parenchymal injury, leading to PH, RV dysfunction, and ultimately HF. This project will leverage recently completed chest CT (for CAC) and echocardiography in 1,579 Atherosclerosis Risk in Communities (ARIC) study participants at the 7th study visit (2/2018-11/2019; age ~81±4 yrs). Novel CT-based measures of pulmonary vascular remodeling and parenchymal injury (fibrotic, emphysematous), and advanced 3D and strain-based echo measures of RV function will be performed. These data will be integrated with clinical assessments, outcomes surveillance, aptamer-based proteomics, and genomics to help define the most relevant targets to prevent progressive PVD in the very elderly. Specific aims include: (1) Define the extent to which LV dysfunction and pulmonary parenchymal injury promote pulmonary venous and arterial remodeling in the very elderly; (2) Determine the extent to which pulmonary vascular remodeling predicts RV dysfunction, reduced functional capacity, and incident HFpEF; and (3) Identify proteins and protein networks that predict pulmonary vascular remodeling. Replication will occur in COPDGene and the Framingham Heart Study, and Mendelian randomization analyses will identify the subset of potentially causal associations. Quantifying pulmonary vascular remodeling will identify pathophysiologically distinct morphologic PVD sub-phenotypes enabling more precise application of existing therapies, while discovery of associated molecular pathways may inform novel therapeutic targets.