Metabolomic correlates of improved Alzheimer’s disease polygenic risk score for diverse populations

Alzheimer’s disease (AD), hypothesized to begin in midlife before symptoms appear, is the leading cause of dementia in the United States resulting in loss of cognitive function and eventual death. AD has a higher prevalence in non-Hispanic Black and Hispanic adults compared to non-Hispanic White adults and is expected to increase at a greater rate in non-Hispanic Black and Hispanic populations than non-Hispanic Whites. Although these health disparities for AD exist, little research has been done into the genetic epidemiology behind AD, particularly polygenic risk scores (PRS), for these populations. PRS has poorer performance in populations different than the populations included in large-scale genetic analyses. Since the majority of genetic summary statistics used to construct PRS come from European populations, PRS tends to perform worse in non-European populations. Along with other scientific and clinical uses for PRS, an improved PRS instrument could be used to provide a better understanding of metabolomic changes in midlife associated with high AD genetic risk. Metabolomic changes may provide insight into AD mechanisms as either part of the AD pathogenesis pathway or as biomarkers indicating early AD development. Here, we propose to evaluate multiple methods for AD PRS construction in diverse populations and use the best-performing PRS to examine the relationship between AD genetic risk and metabolomics in midlife. Aim 1 will construct AD PRS for diverse populations by first using summary statistics available in published GWAS with a focus on newly available GWAS with predominantly African ancestry populations. Various PRS methods will be assessed to construct the best possible PRS in diverse testing cohorts, with a focus on participants with significant genetic similarity to African reference populations. Then, aim 2 will use the AD PRS as the exposure for the metabolite analysis in >25,000 participants from NHLBI Trans-Omics for Precision Medicine (TOPMed) with untargeted metabolomics. The use of the PRS instead of cases and controls will provide advantages such as increased power due to increased effective sample size since, in middle-aged TOPMed cohorts, most individuals have not developed AD at the time of metabolite measurement. This method also has the ability to identify metabolites associated with a higher genetic risk burden in midlife before AD develops. Genetic causal inference methods will be utilized to assess metabolites associated with AD incidence. The results of this study will provide a more predictive AD PRS for diverse populations and a better understanding of midlife metabolite changes associated with AD genetic risk. Ms. Drzymalla’s mentoring team, who has experience in AD, genetics, and metabolomic epidemiology as well as PRS and causal inference methods, will provide guidance for the completion of the research and training plans. Ms. Drzymalla will receive training on AD pathogenesis, genetic computation methods, metabolomic epidemiology, and causal inference methods as well as take part in professional development activities to develop the skills to become an independent researcher at the intersection of genetics, epidemiology, and AD. Project Number: 1F31AG094260-01A1 | Fiscal Year: 2026 | NIH Institute/Center: National Institute on Aging (NIA) | Principal Investigator: Emily Drzymalla | Institution: UNIV OF NORTH CAROLINA CHAPEL HILL, CHAPEL HILL, NC | Award Amount: $41,757 | Activity Code: F31 | Study Section: Special Emphasis Panel[ZRG1 F18-E (20)] View on NIH RePORTER: https://reporter.nih.gov/project-details/11318067