PROJECT SUMMARY/ABSTRACT Background: As cognitive decline is a mounting health concern for our aging population, it is becoming increasingly important to better characterize the neurobiological changes leading to age-related cognitive impairment and dementia. Although neuroimaging methods have shown promise at detecting microstructural brain changes associated with cognitive decline and Alzheimer's disease (AD), there is need for biomarkers with improved sensitivity to the earliest disease stages, when treatments may be most effective. Emerging research suggests that blood-brain barrier (BBB) breakdown occurs when cognitive impairments are still mild, and may precede brain atrophy. Further research is needed to understand how BBB permeability changes over the course of AD and how it relates to established markers of brain microstructure, morphometry and molecular pathology. Preliminary Studies: Our pilot study demonstrated sensitivity of Restriction Spectrum Imaging (RSI) to microstructural brain changes in Mild Cognitive Impairment (MCI) and early AD. RSI, cognitive function and CSF amyloid-β were measured in 31 healthy controls (HC), 12 individuals with MCI and 13 with mild AD. Neurite density (ND), a measure of restricted water diffusion that accounts for crossing fibers, was lower in several white matter tracts, and gray matter isotropic free water diffusion (IF) was higher for MCI/AD than HC. ND and IF correlated with episodic memory and with amyloid-β burden. Proposed Studies: Dynamic contrast-enhanced MRI, RSI, structural MRI, CSF amyloid-β and tau, and neuropsychological data will be acquired on 15 HC with low genetic risk for AD (APOE4 non-carriers) and 15 individuals with MCI. BBB permeability will be compared between groups and correlated with memory performance, RSI measures, morphometry and amyloid/tau to test whether BBB breakdown is associated with microstructural changes, cognitive impairment and established AD biomarkers during prodromal AD (Aim 1). During the R00 stage, neuroimaging, CSF and cognitive data will be collected on 40 HC with low genetic risk for AD, 70 HC with elevated genetic risk for AD (APOE4 carriers) and 70 individuals with MCI. All participants will be clinically and cognitively evaluated 2-3 years after baseline to test the hypothesis that BBB breakdown and RSI metrics predict cognitive decline and are sensitive to preclinical AD (Aim 2A). Data from all stages will be combined to model patterns of change in BBB permeability, brain microstructure and molecular pathology from normal to prodromal AD (Aim 2B). Amyloid PET will be performed on a subset of participants (31 amyloid-negative HC, 37 amyloid-positive HC, 31 MCI) to test the hypotheses that BBB breakdown correlates and colocalizes with amyloid deposition, and that the association between increased BBB permeability and amyloid is greater in APOE4 carriers than non- carriers (Aim 3).