Extreme polygenic risk scores track high Alzheimer's risk
In an analysis of 30,521 individuals, no controls were observed among 54 participants with extreme polygenic risk scores across European and African ancestries.
medRxiv
In a study of 11,200 Alzheimer's disease cases and 19,321 controls aged 65 and older, researchers evaluated whether extreme polygenic risk scores identify individuals at high risk. Using data from the Alzheimer's Disease Sequencing Project, the authors evaluated the extreme tails of 1,752 scoring methods across four genetic ancestries: Admixed American, African, East Asian, and European.
Individuals in the extreme upper tail showed significant disease enrichment across all ancestries. Notably, researchers observed no controls among 42 European-ancestry and 12 African-ancestry individuals with extreme scores. These extreme scores spanned APOE diplotypes, and 38% occurred in non-APOE ε4 carriers.
Why it matters
Alzheimer's disease pathology accumulates years before clinical cognitive decline appears with age. Identifying individuals at extreme polygenic risk could improve presymptomatic detection, especially for people who lack common APOE ε4 risk variants.
Caveats
This work is a preprint and has not yet been peer-reviewed. In addition, extreme polygenic risk profiles represent rare subsets of individuals, and observational findings require prospective validation.
The paper
Rare extreme polygenic risk scores strongly indicate Alzheimer's disease risk
University of Kentucky
medRxiv · 9 Sep 2026 · Preprint, not peer-reviewed