A Comparison of Different Models to Compute Cardiac Output by Open-Circuit Inert Gas Wash-In

Purpose: Cardiac output (Q̇) can be measured non-invasively during exercise using the open-circuit acetylene (C 2 H 2 ) uptake method. However, it is unclear whether different calculation techniques for these breathing-based Q̇ measurements yield consistent values and how variations in breathing patterns affect them. This study compared breathing-derived Q̇ estimates from three validated models (Wagner et al.: WAG, Gan et al.: GAN, and Iterative: ITER) and examined the effect of doubling breathing frequency on these estimates during continuous cycling exercise. Methods: Fourteen healthy adults completed two 28-min constant-work-rate cycling stages at moderate (MOD) and heavy (HVY) intensities. At each stage, Q̇ was measured under normal, self-selected breathing (NF) and under controlled high-frequency breathing (HF) at 200% of NF. For every C 2 H 2 uptake trial, Q̇ was calculated using WAG, GAN, and ITER models. Results: Doubling breathing frequency had model-specific effects on Q̇ estimates. WAG produced significantly lower Q̇ during HF breathing compared with NF at both MOD (3.3±1.1 L·min -1 ; p<0.001) and HVY (2.8±1.8 L·min -1 ; p<0.001) intensities. In contrast, ITER yielded significantly higher Q̇ under high-frequency breathing (MOD: -1.7±1.0 L·min -1 and HVY: -3.0±1.6 L·min -1 ; p<0.001 for both). GAN showed only a small significant difference between breathing conditions at MOD (-0.4±0.7 L·min -1 ; p=0.039), but not at HVY (p=0.845). Conclusions: These findings demonstrate that breathing frequency can meaningfully influence C 2 H 2 -derived Q̇ estimates and that computational models differ markedly in their sensitivity to ventilatory pattern. GAN appears to be the most robust to altered breathing frequency compared to WAG and ITER. Consistency in model selection is essential when using inert-gas uptake techniques to assess Q̇ during exercise.