Exercise endpoints in patients with chronic heart failure.
BACKGROUND: Peak oxygen consumption (V(O(2))) is a powerful predictor of outcome in patients with chronic heart failure. This is not a test that is readily clinically available. We therefore sought to establish a method of assessing peak V(O(2)) from non-invasively acquired data. METHODS: We analysed the results from incremental treadmill exercise tests in 60 patients [aged 59.0 (S.D. 12.4) years] with chronic heart failure or left ventricular dysfunction [left ventricular ejection fraction (29.6 (15.2)%)] and 52 control subjects [aged 36.7 (12.3)]. Metabolic gas exchange during exercise was measured with a respiratory mass spectrometer. Heart rate and blood pressure were measured. RESULTS: Peak V(O(2)) was lower in patients than controls [19.9 (7.7) ml/kg/min vs. 38.3 (9.0), P<0. 001]. Exercise time (r=0.84, P<0.001), heart rate at peak exercise (r=0.63, P<0.0001), change in heart rate (r=0.72, P<0.0001), rate pressure product at peak exercise (r=0.64, P<0.0001) and change in systolic blood pressure (r=0.31, P=0.002) all correlated with peak V(O(2)). In a stepwise regression model, exercise time was the most powerful predictor of peak V(O(2)) (r(2)=0.79). The only additional independent variable was change in heart rate from rest to peak exercise, which increased r(2) to 0.80. In a survival analysis, measured peak V(O(2)) and the peak V(O(2)) estimated from exercise time and change in heart rate had similar predictive power. CONCLUSIONS: In this preliminary study, peak V(O(2)) can be estimated from non-invasively acquired parameters. Estimated peak V(O(2)) and measured peak V(O(2)) have similar predictive power for outcome. Further work is necessary to see if estimated peak V(O(2)) is widely applicable in a clinical setting.