Based on the linked source; automatically prepared and checked against the original report.
Study design
Ten long-term manual wheelchair users aged 29 to 38 completed three-minute trials at three progressively loaded conditions on a wheelchair ergometer. Researchers measured wheel velocity and braking torque, calculated mechanical power, and assessed metabolic variables with a wearable respiratory gas analyzer. The study addressed concerns that wearable devices may not fully capture the external mechanical demand of wheelchair propulsion.
Model findings
Oxygen uptake per kilogram increased across the loading conditions, even as velocity fell under the highest load. The final linear model combined mechanical power and heart rate: V̇O2/kg = 5.5738 + 0.2444P + 0.03587HR. Both predictors were significant positive contributors, with reported 95% confidence intervals of 0.1330 to 0.3558 for power and 0.01342 to 0.05833 for heart rate.
Internal evaluation and limits
The model explained 72.2% of variance, while leave-one-participant-out cross-validation produced an R2 of 0.575, RMSE of 1.128, and MAE of 0.917 mL O2·kg−1·min−1. Cross-validated predictions had a mean bias of −0.156, with limits of agreement from −2.389 to 2.077. The authors characterized power and heart rate as complementary measures, but the exploratory model had only internal evaluation and requires external validation before practical application.
