A Weighted Genetic Risk Score Based on Four APOE-Independent Alzheimer's Disease Risk Loci May Supplement APOE E4 for Better Disease Prediction.

Abstract

Recent genome-wide association studies (GWAS) in European descent population have identified more than 30 independent single-nucleotide polymorphisms (SNPs) associated with Alzheimer's disease (AD). In the present study, we genotyped 39 AD-risk SNPs in 499 sporadic AD patients and 760 matched healthy controls of Mongolian ethnicity from the Inner Mongolia, China. Further, we investigated whether genetic risk score (GRS) combining multiple AD-risk loci confirmed in our study population could improve AD prediction. Two approaches were used for GRS calculation: a simple risk allele count (cGRS) and a weighted approach (wGRS). The area under the receiver operating characteristic curve (AUC) was used to compare the discriminatory ability of the GRS models. Seven SNPs were confirmed associated with AD and four SNPs were associated with AD risk independent of APOE genotypes in our population. GRS based on either seven SNPs or four APOE-independent SNPs were significantly associated with AD risk (P = 2.3E-17~2.0E-6). The AUC for wGRS was significantly greater than for cGRS (0.6416 versus 0.6339, P = 0.0049 for seven SNPs; 0.5857 versus 0.5765, P = 0.0047 for four APOE-independent SNPs). Furthermore, we found that wGRS combining four APOE-independent SNPs and APOE E4 genotypes reached AUC 0.7023, significantly better than the discriminate ability of APOE E4 genotypes alone (AUC = 0.6699, P = 0.0379). The combined model, with an AUC of 0.6989, significantly higher than that of APOE E4 alone (0.6529) (P= 0.0284), for subjects in a validation cohort comprising 250 cases and 380 controls, randomly selected from our original cohort. In summary, we found that wGRS based on four APOE-independent AD risk SNPs may supplement APOE E4 for better assessing individual risk for AD in Mongolian population in China.