Evidence of heterogeneity in genetic determinants of qt interval across global populations: The population architecture using genomics and epidemiology (PAGE) study

Abstract

QT interval (QT) prolongation is an established risk factor for ventricular tachyarrhythmia and sudden cardiac death and is highly heritable (~40%). Previous genome-wide association studies in European ancestral populations identified multiple genetic loci that influence QT. Yet, few studies have examined the role of these loci in ethnically diverse populations. Here, for 21 SNPs from 12 loci previously associated with QT, we examined direction, magnitude, and precision of effect sizes in European Americans (n=16,398), African Americans (n=5,437), American Indians (n=5,032), Hispanics/Latinos (n=1,143), and Asians (n=932) from the participating studies of the Population Architecture Using Genomics and Epidemiology study. Tests of association using linear regression models and assuming an additive genetic model were performed separately by each study and stratified by race/ethnicity. Summary results were combined using inverse-variance weighted meta-analysis and heterogeneity across race/ethnicity was evaluated using Cochran's Q test. Six of the 21 previously identified QT SNPs located in the NOS1AP, SCN5A, KCNQ1, NDRG4, and KCNE1 genes showed a consistent direction of effect across all five populations, although significant heterogeneity (P < 0.05) was noted for two SNPs. For example, effect sizes for the NOS1AP SNP rs12143842 ranged from 0.87 ms (95% CI: -0.09, 1.8, minor allele frequency [maf] = 0.20) in American Indians to 3.1 ms (95% CI: 2.6, 3.5, maf = 0.25) in European Americans (Pheterogenetiy = 0.001). A further 10 SNPs from seven genes (NOS1AP, ATP1B1, SCN5A, PLN, KCNH2, KCNQ1, and KCNJ2) showed a consistent direction of effect in four of five populations; significant heterogeneity (P < 0.05) was detected across race/ethnicity for five of the 10 SNPs. For the remaining five SNPs, estimated effects were highly variable across race/ethnicity; we also note that the KCNQ1 SNP rs12576239 was the only SNP that failed to show a significant association in European Americans. These results provide evidence that loci previously associated with QT in European Americans have consistent direction of effect across racial groups, although substantial heterogeneity detected. Such heterogeneity among race/ethnicity suggests either varied linkage disequilibrium (LD) patterns between the tag SNP and underlying causal variant or allelic heterogeneity. Fine-mapping studies evaluating differences in LD structure across diverse populations are needed to pinpoint potential causal variants and aid functional characterization of loci underlying QT.