Basic Science and Pathogenesis.

Background: Published data have highlighted associations between Alzheimer's disease (AD) susceptibility loci and AD-related brain changes. The amyloid imaging to prevent AD (AMYPAD) consortium is a European collaboration consisting of several parent cohorts, four of which had raw genotype array data available. We sought to integrate and harmonise the genetic data, calculate AD polygenic risk scores (PRS), and investigate their association with global amyloid deposition.

Method: Raw genetic data (GRCh37) was available for 753 non-demented participants, which underwent standard pre-imputation quality control (QC) using PLINK. Imputation was performed using the Michigan Imputation server and HRC reference panel, and standard post-imputation QC was done prior to the final merging of cohorts. PRSice was used for PRS calculations with Stage 1 summary statistics from Kunkle et al. (2019) as base file and European individuals from 1000 Genomes as reference for clumping. We calculated several builds of PRS at three p-value thresholds for SNP inclusion: 5 × 10, 1 × 10, and 0.1. These included PRS, PRS (Chr19:45-48.8Mb), APOE (the weighted sum of the two major APOE SNPs rs429358 and rs7412), PRS (all SNPs excluding Chr19:45-48.8Mb), and a composite PRS+APOE . All were corrected for genetic principal components 1-5 and standardised against 1000 Genomes. Cross-sectional global amyloid burden (expressed in Centiloids) was available for everyone. Linear regression models determined if PRS were associated with amyloid deposition (age, sex, and education as covariates). Significance was based on an uncorrected p<0.05 divided by the number of SNP inclusion thresholds (N = 3).

Result: Participants were 66.9±7.6 years, 58.6% female, and 207 (27%) considered amyloid positive (CL>21). In this preliminary analysis, we found that most PRS builds were significantly associated with amyloid at all SNP inclusion thresholds assessed (Table 1, representative plots Fig. 1A,1B), except for PRS (Table 1, representative plot Fig. 1C).

Conclusion: We highlight the significant influence of APOE in determining AD genetic predisposition, and that AD PRS are associated with amyloid deposition. Our results show that AMYPAD is a suitable cohort for studying genetic factors driving amyloid deposition before clinical onset, which we aim to extend with the inclusion of more parent cohorts, and regional and longitudinal amyloid data.