Trajectories of Cognition in Middle Age: Implications for Alzheimer's Disease and Related Dementias in the U.S.

Polygenic risk scores (PRS) for dementia and aging-related conditions are known to be associated with cognitive outcomes in older age, but little is known about their relationship to mid-life cognitive decline. We plan to use raw genetic data to derive novel PRS from GWAS sources (including Lambert Alzheimer’s disease PRS, with and without APOE; aPRS for coronary artery disease; a longevity PRS) and evaluate their predictive accuracy for cognitive outcomes in middle age relative to existing PRS. Specifically, we want to create a measure of genetic risk associated with three outcomes: age-related cognition; telomere shortening; and methylation/epigenetic clocks. To achieve this, we will combine the HRS Genotype data with other HRS datasets (Harmonized Cognitive Assessment Protocol (HCAP) (2016 Early V1.0); 2008 Telomere Data; Epigenetic Clocks; 2016 Venous Blood Study (VBS)) to which we already have access. Once we have approved NIAGADS genomics data access, we will additionally request access to the HRS-NIAGADS Cross-Reference File (Genotype Data v3,2006-2012) to link the genomics and HRS datasets. In our ongoing analyses, we would like to update our PRS models by incorporating the most recent GWAS summary statistics. For Alzheimer's disease, this requires access to the full summary statistics from the Kunkle et al., 2019 GWAS. We also would like access to the full summary statistics from the Farrell et al., 2024 GWAS and the Rajabli et al., 2025 GWAS to identify genetic modifiers of tauopathy by comparing progressive supranuclear palsy GWAS results with cross-ancestry Alzheimer’s disease GWAS results.

There is evidence to suggest that differences in people’s genetic code might contribute to differences in age-associated cognitive changes. For example, some people develop memory problems in middle age, and other people experience no changes in memory. Researchers think this may be partially explained by differences in people’s genetic code. We might be able to predict who could experience age-related cognitive changes based on their DNA sequence. If we know which people have experienced memory problems, we can see what their DNA has in common compared to the DNA of people who don’t have any memory problems. Then, we can test this by looking at the DNA of a different group of people; evaluating if their DNA has the same things in common as the group of people with memory problems (vs. no memory problems); and predicting whether they will develop memory problems. The long-term goal of this work is to help identify people who might be at risk for developing memory problems and help them access preventative care or interventions to minimize future cognitive impairment.