Associations between Electrocardiographic Interval Durations and Coronary Artery Calcium Scores: The Diabetes Heart Study

Abstract

The mechanisms underlying the associations between QT interval duration and risk of cardiovascular disease (CVD) remain unclear. It has been assumed that these associations are driven by abnormal myocardial repolarization. We examined the relationship between coronary artery disease, measured by coronary artery calcified plaque (CAC), and the duration of QRS, JT, and QT intervals, among predominantly type two diabetic participants. The study sample included 1,123 subjects from the Diabetes Heart Study, of whom 85% had type 2 diabetes. Correlations between electrocardiogram interval durations and log-transformed coronary artery calcified (CAC) were assessed in univariate and sequential multivariable generalized estimating equation models adjusted for familial correlations, heart rate, age, race, gender, diabetes status, hypertension status, Body Mass Index (BMI), smoking status, systolic blood pressure, Low Density Lipoprotein (LDL) cholesterol, QT-prolonging medications, and use of exogenous estrogen. QT interval duration significantly correlated with the extent of CAC in univariate (r = 0.09, P = 0.01) and multivariable models (r = 0.08, P = 0.01). We observed strong correlations between the QRS duration and CAC in univariate (r = 0.23, P < 0.0001) and adjusted models (r = 0.10, P = 0.01). In contrast, the JT interval was not associated with CAC. A strong correlation existed between the QRS interval and CAC in men (QRS: r = 0.24, P < or = 0.0001) and diabetics (QRS: r = 0.25, P < or = 0.0001) but was absent in women and nondiabetics. These relationships were not modified by CVD, race, or presence of bundle branch block. QT duration correlates with the amount of CAC in a predominantly diabetic population. The association between QT duration and CAC is driven by QRS and not JT interval duration.