Introduction: Objective of this study was to determine the optimal (most heritable) phenotype for gene finding studies of QT interval in the general population. We also studied the extent to which heritability of QT interval can be explained by genes that also influence resting heart rate.
Methods and results: Subjects in this classic twin study were 105 monozygotic and 256 dizygotic female twin pairs (mean age: 49.9 +/- 11.5). ECG parameters were measured electronically using the Cardiofax ECG-9020. Quantitative genetic modeling was performed with Mx software. Best-fitting univariate models showed significant heritabilities for resting heart rate (0.55, 95% CI: 0.44-0.65), uncorrected QT interval (0.60, 95% CI: 0.49-0.69), and the Framingham QTc interval (0.50, 95% CI: 0.39-0.60). Familial resemblance of Bazett's QTc was best explained by shared environmental factors (0.34, 95% CI: 0.24-0.43) rather than genes. Simultaneously modeling heart rate and the uncorrected QT interval confirmed considerable heritabilities of 56% and 60%, respectively. Forty-four percent of the variance in QT interval was due to genes in common with heart rate, whereas 16% was due to genes specific to QT interval. The heritability of QT interval after the removal of effects shared with heart rate within the bivariate model (cf. QTc) was 51%.
Conclusion: About a quarter of the QT interval heritability is due to genes specific for QT interval, while the majority is shared with genes for heart rate. Differences in QTc heritability estimates indicate that use of correction formulae is best avoided in gene finding studies to avoid erroneous results.