Perales-Puchalt · medRxiv : the preprint server for health sciences 2025 · cross-sectional study · n=77,676

Variation in APOE ε4 Prevalence and Brain Health Associations by European Descent in White All of Us Participants.

Cited 1 times in the scientific literature.

Level 4 - case-series / case-control

Cross-sectional observational analysis of cohort survey and genetic data

PubMed 40791670 · doi:10.1101/2025.07.16.25331679 · record verified 2026-08-28

What was done

Cross-sectional analysis of 77,676 adult White participants in the United States All of Us cohort study. Participants were categorized into Northern, Central, or Southern European descent without overlap based on the Basics Survey. APOE ε4 status (zero, one, or two copies) was determined using rs429358 and rs7412 variants. Logistic regression models evaluated regional differences in allele prevalence, and interaction tests assessed whether descent moderated associations between APOE ε4 and brain health outcomes.

What was found

Among participants, 7.7% were of Southern European descent and 64.7% were of Northern European descent. One copy of APOE ε4 was present in 17.4% of Southern versus 24.9% of Northern European descent participants (OR 1.60, 95% CI 1.49-1.71). Two copies were present in 0.1% of Southern versus 2.0% of Northern European descent participants (OR 2.21, 95% CI 1.71-2.88). European descent moderated associations between APOE ε4 frequency and two brain health outcomes (interaction p <= 0.10), though specific outcomes and effect sizes were not reported in the abstract.

Why it matters

Treating White individuals as a monolithic group masks significant intra-European variations in APOE ε4 prevalence. Accounting for regional European ancestry is necessary to avoid misestimating Alzheimer's disease risk profiles and skewing clinical trial recruitment.

Limits

This is an observational cross-sectional analysis from a preprint. Regional descent was determined by self-report rather than quantitative genetic ancestry. Specific brain health outcomes were not named in the abstract, and moderation used a permissive significance cutoff (p <= 0.10). The cohort is a volunteer sample rather than a representative probabilistic population.

Cited by