As over 25 million people living with HIV (PWH) in sub-Saharan Africa (sSA) reach older age, determining their risk for and causes of mild cognitive impairment (MCI) and Alzheimer’s related dementia (ADRD) is emerging as a major public health priority. Although neuropsychological battery data exist, it has been limited to young populations and lacked specialist evaluations, brain imaging and biomarkers to confirm disease phenotypes. Moreover, social and clinical health determinants differ meaningfully in sSA, so risk factors for MCI and its impact on households require further study. This proposal is led by experts in the epidemiology of HIV in sSA, diagnosis and deep-phenotyping of MCI and ADRD with biomarker and brain imaging techniques, and machine learning methods to elucidate mechanisms. Their preliminary data include neuropsychological screening of 300 older PWH in Uganda on ART (mean age >60), and 300 demographically-similar HIV-negative comparators, showing that >30% of PWH have characteristics of MCI and that brain MRI and machine learning techniques add critical phenotyping data to standard batteries. Four specific aims are proposed: Aim 1: Determine the prevalence and classification of MCI and ADRD (1A) and compare trajectories of cognitive decline (1B) between older PWH in Uganda and demographically similar HIV-negative individuals. Comprehensive neuropsychological assessments will be completed among older adults with and without HIV in the Uganda Aging Cohort (n=600) in years 1 and 4, and MCI and ADRD type (ie vascular, Alzheimer’s Disease, HIV-related) will be defined using multi-disciplinary case consensus criteria to provide clinical diagnoses and etiologies. Aim 2: Identify pathophysiologic contributors to MCI and ADRD in older adults in Uganda by performing deep-phenotyping with novel plasma biomarkers and neuroimaging. Assessments will include Aβ42/Aβ40, p-tau217, GFAP, and NfL biomarkers of ADRD and brain MRIs to characterize MCI phenotypes among PWH in Uganda. Aim 3: Estimate the psychosocial and economic impacts of MCI and ADRD on adult household members in Uganda. Adult household members will be surveyed about employment/resource use, caregiving burden, quality of life, stigma, social participation, loneliness, and mental health to compare participants by the presence vs absence of MCI in the household. Aim 4: Discover and validate novel explanatory models of MCI and dementia among older PWH by employing ML methods with the full array of data collected in Aims 1-3. The elements collected in Aims 1-3 will be used to determine the combinations of highly dimensional features that most accurately and reliably classify individuals with MCI and ADRD. Completing these aims will advance our understanding of the epidemiology of MCI and ADRD, their phenotypes, and their societal impacts in Uganda. In doing so, it will lead to diagnostic and intervention approaches to address NIA research priorities.