Reproducibility (reliability and agreement) of ventilatory threshold and peak responses during cardiopulmonary exercise test in people with stroke

Background A cardiopulmonary exercise test (CPET) is used to determine the ventilatory thresholds and to directly assess cardiorespiratory capacity. However, its reproducibility should be tested in people with stroke as sequelae imposed by the stroke may induce important variations among and within each subject, affecting the reproducibility of the physiological responses to CPET.Purpose This cross-sectional repeated measures study design aims to determine the reproducibility of anaerobic threshold (AT), respiratory compensation point (RCP), and maximal cardiorespiratory capacity assessed during a CPET in people with stroke.Methods Twenty-eight subjects with hemiparesis after stroke aging 60 ± 13 years were submitted to two treadmill CPETs with identical protocols.Data Analysis The reproducibility of heart rate (HR) and oxygen consumption (VO2) obtained at AT, RCP, and peak effort was evaluated by systematic error (paired t-test); reliability (ICC and 95% confidence interval); and agreement (typical error and coefficient of variation).Results There were no systematic errors for HR and VO2assessed at AT, RCP, and peak effort (p > 0,05). Reliability was high for these variables during CPET (ICCs > 0.93). Agreement was good for all variables. Typical errors for HR and VO2 assessed at AT, RCP, and peak effort were, respectively, 7, 7, and 8 bpm, and 1.51, 1.44, and 1.57 ml.kg−1.min−1. Coefficients of variation assessed at AT, RCP, and peak effort were, respectively, 5.7, 5.1, and 6.0% for HR and 8.7, 7.3, and 7.5% for VO2.Conclusions HR and VO2 measured at AT, RCP, and peak effort during a treadmill CPET present good reproducibility in people with stroke, showing high reliability and good agreement.