Comparison of cardiac output estimates by bioreactance and inert gas rebreathing methods during cardiopulmonary exercise testing

Purpose This study assessed the agreement between cardiac output estimated by inert gas rebreathing and bioreactance methods at rest and during exercise. Methods Haemodynamic measurements were assessed in 20 healthy individuals (11 females, nine males; aged 32 ± 10 years) using inert gas rebreathing and bioreactance methods. Gas exchange and haemodynamic data were measured simultaneously under rest and different stages (i.e. 30, 60, 90, 120, 150 and 180 W) of progressive graded cardiopulmonary exercise stress testing using a bicycle ergometer. Results At rest, bioreactance produced significantly higher cardiac output values than inert gas rebreathing (7·8 ± 1·4 versus 6·5 ± 1·7 l min−1, P = 0·01). At low‐to‐moderate exercise intensities (i.e. 30–90 W), bioreactance produced significantly higher cardiac outputs compared with rebreathing method (P<0·05). At workloads of 120 W and above, there was no significant difference in cardiac outputs between the two methods (P = 0·10). There was a strong relationship between the two methods (r = 0·82, P = 0·01). Bland–Altman analysis including rest and exercise data showed that inert gas rebreathing reported 1·95 l min−1 lower cardiac output than bioreactance, with lower and upper limits of agreement of −3·1–7·07 l min−1. Analysis of peak exercise data showed a mean difference of 0·4 l min−1 (lower and upper limits of agreement of −4·9–5·7 l min−1) between both devices. Conclusion Bioreactance and inert gas rebreathing methods show acceptable levels of agreement for estimating cardiac output at higher levels of metabolic demand. However, they cannot be used interchangeably due to strong disparity in results at rest and low‐to‐moderate exercise intensity.