Non-invasive estimate of cardiac output during exercise based on impedance cardiography and oxygen uptake in the elderly.

OBJECTIVE This study assessed the agreement between cardiac output (Q) measurements generated by two indirect methods and their applicability in 54 men and 77 women (aged 60-90). METHODS The subjects performed a cardiopulmonary test with incremental to maximal exercise on a cycle-ergometer. "Q" was estimated based on oxygen uptake and on impedance cardiography at 50% and 100% of the maximal workload (Wmax) reached during the test. The limits of agreement and mean differences (biases) were estimated using the methodology proposed by Bland-Altman, whereas the relationship between the methods was analyzed by linear regression and intra-class correlation (ICC). RESULTS For the load equivalent do 50% of Wmax, the bias between the means of the two methods was -0.58 L x min-1 (6.3% smaller for the impedance measurement) and the limit of agreement was +/- 2.54 L x min-1 (24.7%). At 100% of Wmax, the bias was 1.03 L x min-1 (8.5% greater for the impedance measurement) with a limit of agreement of +/- 3.35 L x min-1 (27.6%). ICC values and the regression curves obtained between the methods were: Qimp=1.82+0.75Q VO2 (r(2)=0.765, p<0.001; ICC=0.727, p<0.01) at 50% Wmax; and Qimp=0.93+1.00Q VO2 (r(2)=0.755, p<0.001; ICC=0.796, p<0.01) at 100% Wmax. CONCLUSION These results suggest that there is a good agreement between the two methods for the measurement of Q during exercise in the elderly, which is compatible with the findings of similar studies with young subjects.