Genetic Basis and Prognostic Value of Exercise QT Dynamics

Abstract

Background: Abnormal QT interval responses to heart rate (QT dynamics) is an independent risk predictor for cardiovascular disease in patients, but its genetic basis and prognostic value in a population-based cohort have not been investigated.

Methods: QT dynamics during exercise and recovery were derived in 56 643 individuals from UK Biobank without a history of cardiovascular events. Genome-wide association studies were conducted to identify genetic variants and bioinformatics analyses were performed to prioritize candidate genes. The prognostic value of QT dynamics was evaluated for cardiovascular events (death or hospitalization) and all-cause mortality.

Results: Heritability of QT dynamics during exercise and recovery were 10.7% and 5.4%, respectively. Genome-wide association studies identified 20 loci, of which 4 loci included genes implicated in mendelian long-QT syndrome. Five loci did not overlap with previously reported resting QT interval loci; candidate genes included KCNQ4 and KIAA1755. Genetic risk scores were not associated with cardiovascular events in 357 882 unrelated individuals from UK Biobank. We also did not observe associations of QT dynamics during exercise and recovery with cardiovascular events. Increased QT dynamics during recovery was significantly associated with all-cause mortality in the univariate Cox regression analysis (hazard ratio, 1.09 [95% CI, 1.05-1.13], P=2.28×10^-5), but the association was not significant after adjusting for clinical risk factors.

Conclusions: QT interval dynamics during exercise and recovery are heritable markers but do not carry independent prognostic information for clinical outcomes in the UK Biobank, a population-based cohort. Their prognostic importance may relate to cardiovascular disease cohorts where structural heart disease or ischemia may influence repolarization dynamics. The strong overlap between QT dynamics and resting QT interval loci suggests common biological pathways; however, nonoverlapping loci suggests alternative mechanisms may exist that underlie QT interval dynamics.

Keywords: cardiovascular disease; exercise; genome-wide association study; mortality; risk factors.

Figure 1.

Flowchart on selection of individuals for prognostic and genetic analyses of QT dynamics. QT dynamics were derived from the exercise cohort. Yellow presents the data selection for the prognostic analysis, blue presents the selection for the genome-wide association studies (GWAS) and brown represents the individuals that were unrelated to individuals included in the GWAS to allow unbiased genetic risk analyses using the variants discovered in the GWAS. CV indicates cardiovascular; and QC, quality control.

Figure 2.

Association results of the QT dynamics genome-wide association studies (GWAS) in the full data. Note: To ensure there was no overlap between the discovery and the replication cohorts, we removed first and second-degree related individuals (kinship coefficient >0.88). SNV indicates a single-nucleotide variant.

Figure 3.

Overlap between loci for QT dynamics during exercise (blue box) and recovery (yellow box) and reported loci for resting QT interval (black box). A substantial proportion of the loci for QT dynamics overlapped loci previously reported for resting QT interval. Five loci (bold) for QT dynamics during exercise did not overlap with previously reported loci for resting QT interval. All loci for QT dynamics during recovery overlapped with resting QT interval loci. *Sex-specific (males) locus for QT dynamics during exercise.