Prognostic value of ambulatory heart rate revisited in 6928 subjects from 6 populations

The evidence relating mortality and morbidity to heart rate remains inconsistent. We performed 24-hour ambulatory blood pressure monitoring in 6928 subjects (not on beta-blockers; mean age: 56.2 years; 46.5% women) enrolled in prospective population studies in Denmark, Belgium, Japan, Sweden, Uruguay, and China. We computed standardized hazard ratios for heart rate, while stratifying for cohort, and adjusting for blood pressure and other cardiovascular risk factors. Over 9.6 years (median), 850, 325, and 493 deaths accrued for total, cardiovascular, and noncardiovascular mortality, respectively. The incidence of fatal combined with nonfatal end points was 805, 363, 439, and 324 for cardiovascular, stroke, cardiac, and coronary events, respectively. Twenty-four-hour heart rate predicted total (hazard ratio: 1.15) and noncardiovascular (hazard ratio: 1.18) mortality but not cardiovascular mortality (hazard ratio: 1.11) or any of the fatal combined with nonfatal events (hazard ratio: < or =1.02). Daytime heart rate did not predict mortality (hazard ratio: < or =1.11) or any fatal combined with nonfatal event (hazard ratio: < or =0.96). Nighttime heart rate predicted all of the mortality outcomes (hazard ratio: > or =1.15) but none of the fatal combined with nonfatal events (hazard ratio: < or =1.11). The night:day heart rate ratio predicted total (hazard ratio: 1.14) and noncardiovascular mortality (hazard ratio: 1.12) and all of the fatal combined with nonfatal events (hazard ratio: > or =1.15) with the exception of stroke (hazard ratio: 1.06). Sensitivity analyses, in which we stratified by risk factors or from which we excluded 1 cohort at a time or the events occurring within 2 years of enrollment, showed consistent results. In the general population, heart rate predicts total and noncardiovascular mortality. With the exception of the night:day heart rate ratio, heart rate did not add to the risk stratification for fatal combined with nonfatal cardiovascular events. Thus, heart rate adds little to the prediction of cardiovascular risk.