SUMMARY: Alzheimer's disease (AD) and related dementias (ADRD) disease burden is anticipated to affect diverse ethnic and racial groups disproportionately with the most staggering increase of nearly 6-fold impacting Hispanics. Nonetheless, there is a lack of scientific research on Hispanic adults, particularly among the most populous Mexican American (MA) demographic, fostering a poor understanding of the mechanisms contributing to the observed disparities. Tau deposition is strongly linked to cognitive decline in AD and can be visualized in vivo with sensitive and specific positron emission tomography (PET) tracers such as 18F-MK-6240. Data from our team and others indicates that artificial intelligence (AI) methods that quantify tau deposition patterns specific to AD may optimize detection of early disease risk and better predict cognitive decline. Unfortunately, few studies have been conducted in diverse cohorts, which limits generalizability and may precipitate further inequities in AD diagnosis and treatment. In addition, gaps remain in our understanding of the impact of mixed dementia pathologies on cognition. AD and cerebrovascular (CVD) commonly co-occur, and may have a larger impact on disease presentation in MA adults relative to non-Hispanic white (NHW) adults due to socio-economic disparities and broader social determinants of health (SDoH) that increase cardiovascular risk factors (CVRFs). We now propose to develop and validate traditional and novel tau PET imaging and CVD measures in longitudinal cohort of 500 MA older adults (N=150 cognitively unimpaired (CU), N=150 mild cognitive impairment (MCI), N=200 AD dementia) from the South Texas Alzheimer's Disease Research Center and the Nantz National Alzheimer's Center. We will assess the diagnostic accuracy of traditional and AI-derived tau PET indices and examine their associations with longitudinal cognitive decline (Aim 1). We also evaluate associations between the tau PET measures and ADRD plasma biomarkers in order to facilitate efforts to utilize blood-based biomarkers for initial diagnostic screening and clinical trial stratification. Next, we will evaluate the interactive effects of tau and CVD burden on cognition and will employ causal interference modeling to elucidate the pathways linking SDoH and modifiable CVRFs with ADRD (Aim 2). Leveraging the resource of harmonized clinical and neuroimaging data on 200 NHW adults from the Nantz National Alzheimer's Center, we will examine the hypothesis that ethnic disparities in ADRD will be mediated through these pathways. Finally, we will validate our imaging and plasma biomarker findings in an independent sample of 1,000 Hispanic and 1,000 NHW older adults participating in the community-based HABS-HD study (Aim 3). Our proposal will provide important insights into the diagnostic accuracy of traditional and novel tau PET measures. We will further examine the overlay of tau pathology with broader ADRD neuroimaging and blood biomarkers...