Abstract The APOE ϵ4 allele is the most important genetic risk factor for late onset Alzheimer's disease (AD). APOE ϵ4 may contribute to AD risk by altering inflammation, lipid homeostasis and/or amyloid clearance. Recent genetics studies have implicated multiple pathways in innate immunity in late-onset AD, including the triggering receptor expressed on the myeloid cells 2 (TREM2) gene. In contrast, exercise and physical activity (PA) produce anti-inflammatory changes in the periphery and neurogenic, angiogenic, and anti-inflammatory brain changes in animals. The mechanistic relationships between APOE, innate immunity, and AD, and the potential moderating influence of PA, remain to be established. Our published data on cognitively intact, healthy elders followed for 18 months indicate that sedentary ϵ4 carriers demonstrate significantly lower fMRI activation, poorer episodic memory performance, smaller hippocampal volumes, and abnormal white matter diffusion compared to ϵ4 carriers who engage in regular PA. Most importantly, these group differences were not observed between low and high PA non-carriers, suggesting that the neuroprotective effects of PA are especially potent for persons at genetic risk for AD. Additional preliminary studies demonstrate that TREM2+ innate immune cells play a direct role in regulating AD pathologies in both animal models of AD and potentially in human AD patients. For the proposed interdisciplinary project, our overall hypothesis posits that PA counteracts the negative inflammatory effects of the ϵ4 allele, affecting TREM2 and other innate immune pathways implicated in AD, thereby reducing the risk of cognitive decline and AD in ϵ4 carriers. This project has two specific aims. Aim 1 will recruit 150 cognitively intact, healthy elders (ages 65-80): 75 APOE ϵ4 carriers (ϵ3/ϵ4) and 75 ϵ4 non-carriers (ϵ3/ϵ3). Participants will undergo state-of-the-art measurements of PA and fitness; structural and functional 3T MRI; amyloid PET imaging; CSF/blood biomarkers related to AD, inflammation, and exercise; and comprehensive memory/cognition testing on two occasions separated by 24- months. Aim 2 is analogous to the human project and will determine the impact of voluntary wheel running PA across age (3 months, 6 months, and 9 months) in novel transgenic mouse models of AD humanized for ϵ3 and ϵ4 and crossed with APPPS1 (APPPS1; APOE3/3 and APPPS1; APOE4/4) and control strains. Key indices include: TREM2+ cells, expression of proinflammatory markers, brain Aβ42, and spatial memory performance. Together, these complementary human and animal studies will provide key insights into potential mechanisms linking APOE genotype, exercise, inflammation, AD brain pathology, and cognition that could ultimately be targeted therapeutically.