Utility of graded exercise tolerance tests for prediction of cardiovascular mortality in old age: The Rancho Bernardo Study

•Integrated analyses of GXT predicted long-term all-cause and CHD mortalities.•937 community-dwelling adults underwent GXT and were followed up to 36 years.•Participants with ≥2 positive findings had 2.18-fold higher risk for CHD mortality.•Participants with ≥ 3 positive findings had 6.13-fold higher risk for CHD mortality.

BackgroundExercise electrocardiography in asymptomatic adults has been criticized because of relatively poor accuracy predicting future heart disease risk, but studies may have been too short. We investigated if integrated analysis of graded exercise tolerance tests (GXT) predicted long-term coronary heart disease (CHD) and all-cause mortalities among community-dwelling older adults.Methods and resultsFrom 1972 to 1974, 1789 adult residents of a predominantly Caucasian, middle- to upper-middle-class southern California community participated in a clinical evaluation that included a GXT; 52.4% (N = 939) of those who had baseline GXT were followed up to 2010-up to 36 years-for vital status, CHD and all-cause mortality. Multiply adjusted hazard ratios of an abnormal graded GXT were 1.65 (95% CI 0.78-3.49) and 1.56 (95% CI 1.15-2.11) for CHD and all-cause mortality, respectively. An integrated analysis hazard ratio was calculated based on the following GXT findings: significant ST change, inability to achieve target heart rate [THR], abnormal heart rate recovery [HRR], and chronotropic incompetency [ChI]. Compared to those with 0 or 1 abnormality, participants with 2 or more positive findings had significantly higher CHD (HR 2.18) and all-cause (HR 1.92) mortalities. Participants with 3 or more positive findings showed even higher hazard ratios-CHD (HR 6.16) and all-cause (HR 2.49) mortalities. When adjusted for any of 3 Framingham risk models, the integrated electrocardiographic model correlated well with CHD and all-cause mortalities.ConclusionsAn integrated analysis of electrocardiographic and non-electrocardiographic measures of GXT is useful in predicting long-term CHD and all-cause mortalities in an asymptomatic middle-aged population.