Using the conventional genome-wide significance (GWS) criterion for GWAS (P ≤ 5 × 10−8), few studies have reported AD-relevant GWS. In European-ancestry populations, Treutlein et al.9 reported association with SNPs mapped to chromosome 2q35, and Schumann et al.10 reported association at the AUTS2 locus. Frank et al.11 reported association with rs1789891, which maps between ADH1B and ADH1C. Zuo et al.,12 in a re-analysis of publicly available data, identified association near SERINC2. In Asian populations, Park et al.13 reported association with multiple markers mapped to the chromosome 4 ADH cluster, in a Korean sample, and we14 identified GWS association to ALDH2 in a Chinese population. Both of the Asian samples were small for GWAS, but the AD risk loci had relatively large effects on phenotype. Several groups have used quantitative traits to increase power. Using a symptom count phenotype, Wang et al.15 identified a significant association of AD with C15orf53.
