QT Dispersion and Mortality in the Elderly

BACKGROUND: The prognostic value of QT interval dispersion measured from a standard 12-lead electrocardiogram (ECG) in the general population is not well established. The purpose of the present study was primarily to assess the value of QT interval dispersion obtained from 12-lead ECG in the prediction of total, cardiac, stroke, and cancer mortality in the elderly. METHODS: A random population sample of community-living elderly people (n = 330, age > or = 65 years, mean 74 +/- 6 years) underwent a comprehensive clinical evaluation, laboratory tests, and 12-lead ECG recordings. RESULTS: By the end of the 10-year follow-up, 180 subjects (55%) had died and 150 (45%) were still alive. Heart rate corrected QT (QTc) dispersion had been longer in those who had died than in the survivors (75 +/- 32 ms vs 63 +/- 35 ms, P = 0.01). After adjustment for age and sex in the Cox proportional hazards model, prolonged QTc dispersion (> or = 70 msec) predicted all-cause mortality (relative risk RR 1.38, 95% confidence interval CI 1.02-1.86) and particularly stroke mortality (RR 2.7, 95% CI 1.29-5.73), but not cardiac (RR 1.38, 95% CI 0.87-2.18) or cancer (RR 1.51, 95% CI 0.91-2.50) mortality. After adjustment for age, sex, body mass index, blood pressure, blood glucose and cholesterol concentrations, functional class, history of cerebrovascular disease, diabetes, smoking, previous myocardial infarction, angina pectoris, congestive heart failure, medication, left ventricular hypertrophy on ECG, presence of atrial fibrillation and R-R interval, increased QTc dispersion still predicted stroke mortality (RR 3.21, 95% CI 1.09-9.47), but not total mortality or mortality from other causes. The combination of increased QTc dispersion and left ventricular hypertrophy on ECG was a powerful independent predictor of stroke mortality in the present elderly population (RR 16.52, 95% CI 3.37-80.89). QTcmin (the shortest QTc interval among the 12 leads of ECG) independently predicted total mortality (RR 1.0082, 95% CI 1.0028-1.0136, P = 0.003), cardiac mortality (RR 1.0191, 95% CI 1.0102-1.0281, P < 0.0001) and cancer mortality (RR 1.0162, 95% CI 1.0049-1.0277, P = 0.005). CONCLUSIONS: Increased QTc dispersion yields independent information on the risk of dying from stroke among the elderly and its component, QTcmin, from the other causes of death.