Reliability of hemodynamic parameters measured by bioimpedance cardiography at different intensities during incremental exercise testing

Background Bioimpedance cardiography offers a non-invasive and time-efficient method to measure hemodynamic parameters. Previous studies only investigated its reliability under steady-state conditions and at maximum load but not at ventilatory thresholds (VTs). This is the first study that assesses the reliability of measured hemodynamic parameters at different exercise stages during cardiopulmonary exercise testing (CPET) using prespecified strict criteria to assess reliability. Methods Data from 31 healthy, well-trained adults were analyzed. Each participant completed two CPETs, both following the same ramp protocol, with a 7-day interval between them. Hemodynamic parameters were measured with the PhysioFlow® (Manatec Biomedical, Poissy, France) at characteristic phases and thresholds [VT1, VT2, and peak oxygen uptake (V̇O 2peak )]. To ensure comparability, the wattage (power) corresponding to the thresholds in Test 1 (P VT1 , P VT2 , and P V̇O2peak ) was used for Test 2. Results Heart rate, stroke volume, and cardiac output demonstrated good reliability on a group level (mean intraclass correlation >0.75) at both thresholds (0.91, 0.80, and 0.77 at P VT1 ; 0.92, 0.80, and 0.77 at P VT2 ) and at P V̇O2peak (0.93, 0.82, and 0.80). For stroke volume at P V̇O2peak , both individual differences (−39.0 to 36.9 mL for the women and −39.9 to 45.2 mL for the men) and mean detectable change (17.5 mL) were larger than the a priori defined acceptable ranges of agreement (−3.6 to 3.8 mL for the women and −4.5 to 3.3 mL for the men). Conclusion The PhysioFlow® reliably measures heart rate, stroke volume, and cardiac output during CPET on a group level. However, as shown by the Bland–Altman plots, the reliability is too low to be used for individual comparisons.