Phillips · Brain : a journal of neurology 2025 · Multi-cohort cross-ancestry genome-wide association study · n=20513

Novel modelling approaches to elucidate the genetic architecture of resilience to Alzheimer's disease.

Cited 3 times in the scientific literature.

Level 3 - non-randomized controlled study

Multi-cohort observational genetic association study

PubMed 40111762 · doi:10.1093/brain/awaf106 · record verified 2026-08-26

What was done

The authors evaluated whether proxy models of neuropathology using APOE status, age, and demographic variables ('silver' models, n = 17,241) could substitute for post-mortem neuropathology ('gold' models, n = 3,272) to quantify cognitive resilience in Alzheimer's disease. Using mixed-effects regressions across 20,513 participants from eight aging cohorts (18,269 European ancestry, 2,244 African ancestry), they derived cognitive resilience residuals across memory, executive function, and language domains. They performed cross-ancestry genome-wide association studies (GWAS), gene- and pathway-based tests, and genetic correlation analyses in the full cohort and in participants with unimpaired baseline cognition.

What was found

Silver and gold resilience phenotypes were strongly correlated (R = 0.77-0.88). Genetic correlation analyses demonstrated significant associations between resilience and multiple neuropsychiatric and cardiovascular traits (PFDR < 0.05). Pathway analyses identified three significant pathways: metabolism of amino acids and derivatives (PFDR = 0.041), negative regulation of TGF-β production (PFDR = 0.019), and severe acute respiratory syndrome (PFDR = 0.00039). Single-variant analysis in cognitively unimpaired participants identified a chromosome 17 locus approaching genome-wide significance (index SNP rs757022, MAF = 0.18, β = 0.08, P = 1.1 × 10^-7), which correlated with altered brain expression of ATP-binding cassette genes.

Why it matters

Pathology-free proxy modeling enables massive sample size expansions for Alzheimer's resilience genetics, overcoming the primary bottleneck of autopsy availability while maintaining high phenotype fidelity.

Limits

The proxy approach lacks direct measures of in vivo or post-mortem amyloid, tau, and co-pathologies for the majority of participants. The top single-nucleotide polymorphism approached but did not achieve canonical genome-wide significance (P < 5 × 10^-8). Non-European representation was limited, with African ancestry participants comprising approximately 11% of the overall cohort.

Cited by