Basic Science and Pathogenesis.

Background: Genome-wide association studies (GWAS) have identified numerous genetic variants associated with Alzheimer's disease (AD) risk, but genetic variation in the onset and progression of AD pathology is less understood. Accumulation of amyloid-β (Aβ) in the brain is a key pathological hallmark of AD beginning 10 - 20 years prior to cognitive symptoms. We investigated the genetic basis of variation in age at onset (AAO) of brain Aβ by comparing the performance of polygenic scores (PGSs) based on AD risk and resilience with a Aβ-AAO trait-specific PGS.

Method: 1122 participants from the Alzheimer's Dementia Onset and Progression in International Cohorts (ADOPIC) study underwent genome-wide SNP genotyping and assessment of brain Aβ using positron emission tomography (PET) imaging at two or more timepoints. AAO was the age at which participants were estimated to have crossed the 20 centiloid (CL) threshold for high Aβ. We utilised AD risk and resilience GWAS summary statistics and conducted a GWAS for AAO using a cross-validation approach (10 test-validation folds). We used PRSice to identify optimal PGSs for Aβ-AAO for risk (PGS), resilience (PGS) and Aβ-AAO (PGS).

Result: PGS and PGS were both significantly associated with Aβ-AAO, such that higher PGS and lower PGS were associated with an earlier Aβ-AAO. PGS showed the strongest association and explained more variance in Aβ-AAO than did PGS. When stratified by APOE ε4 carriage, the strongest genetic risk factor for AD, the association of PGS with Aβ-AAO was stronger among ε4 non-carriers, whilst PGS, was more strongly associated with Aβ-AAO in ε4 carriers.

Conclusion: PGS based on genetic risk and resilience for AD are both significant predictors of the age at which people are estimated to cross the threshold for high brain Aβ burden. Predicting the age at which a person will pass this threshold would enable treatment at an earlier stage, when it may more effectively delay or prevent symptom onset.