Associations between NOS1AP single nucleotide polymorphisms (SNPs) and QT interval duration in four racial/ethnic groups in the Multi-Ethnic Study of Atherosclerosis (MESA)

Abstract

Backgrounds: QT is a risk factor for sudden cardiac death (SCD). A genome-wide association study identified NOS1AP variants associated with QT, which have been replicated in predominantly Caucasian (CAU) populations. We used the Multi-Ethnic Study of Atherosclerosis to examine association of QT with NOS1AP variants in an ethnically diverse cohort.

Methods: Twenty-eight tagging SNPs spanning NOS1AP were genotyped in 2847 MESA participants (approximately equal numbers of CAU, African Americans (AFA), Hispanics (HIS), and Chinese (CHN)), age 45-84 years, without cardiovascular disease. QT was measured using 12-lead ECG. Associations between QT and NOS1AP variants were evaluated using linear regression, adjusted for heart rate, age, gender, and field center stratified by ancestry, using an additive inheritance model. Ancestry informative markers (AIMs) and principal components using AIMs were used as additional covariates.

Results: More NOS1AP SNPs were associated with QT in CAU than the other races. In CAU, each copy of rs1932933 risk allele was associated with an increase in QT (4.9 msec, P = 7.20 × 10-7). Significant associations in CAU and HIS were located at the 5' end, while associations in CHN were located at the 3' end.

Conclusions: NOS1AP variants were associated with QT in CAU, with weaker evidence for selected variants in HIS and CHN. Location of significant SNPs varied across ancestry. We identified possible novel associations at the 3' end of NOS1AP, where we observed significant association with QT in CHN only. Genotyping within these regions may determine functional variants affecting QT and SCD risk. In addition, investigations are needed across ethnically diverse population cohorts.

Figure 1

Race‐specific associations (‐Log₁₀(P value)) for each SNP across NOS1AP. Red markers (adjusted model) represent the P values for each SNP‐trait association with respect to the nominal and Bonferroni‐adjusted alpha levels (0.05 and 0.002, respectively). LD structure (r ²) is shown below.

Figure 2

QT residuals vs. number of risk alleles summed across all nominally significant SNPs. The y‐axis is the residual QT after adjusting for HR, age, gender, MESA field site, QT altering medications, and population stratification.