No added value of Sex and <i>APOE</i>*4 stratification for Alzheimer’s Disease Genetic Risk Scores

Background Genetic risk scores (GRS), be it polygenic (PRS) or hazard (PHS), have shown promise to support genetic prediction and clinical trial recruitment for Alzheimer’s disease (AD). A recent study further suggested that sex-matched AD PHS (e.g., male-to-male) outperform sex-mismatched male-to-female), while no benefit was observed PRS. Similar benefits may expected matching by APOE*4 positivity status, but this question has remained unexplored. The prior sex-stratified work neither evaluated whether full sample GRS (males+females) GRS, nor compared (including many variants regardless of significance) oligogenic fewer more significant variants) effects. Here, we performed the largest-to-date sex-stratified, first APOE*4-stratified, evaluation GRS. Methods design is detailed in Figure-1: genome-wide association studies (GWAS) were a Discovery cohort (LMM-BOLT v.2.3.5) constructed an independent Replication (PRSice-2 & R v3.6). All analyses multiple linear regression on AD-age score models resilience age-related (Figure-2). Models adjusted sex, APOE*4/APOE*2 dosage (rendering estimates/effects APOE), five principal components (PC-AiR; GENESIS; v3.6), array/sequencing center. at different P-value thresholds elucidate versus effects, all with other covariates combined evaluate predictive performance Replication. Results Stratified discovery GWAS are Figure-3. tended perform best stringent (i.e., oligogenic) derived from full-sample virtually always outperformed stratified (Figure-4-5). Predictive similarly highest (Table-1). Conclusions Our results consistent emerging literature supporting showed sex- APOE*4-stratificatied outweighed using While not here, made observations when case-control model (thus equivalent PRS) without age adjustment. These findings important implications guide efforts translating into practice. aging after 90.