This study sought to investigate whether correction of exercise-induced desaturation by oxygen supply (O₂) systematically improves exercise tolerance and cardiorespiratory adaptations in COPD patients.

Twenty-five COPD patients [FEV₁=52±2.5% pred] exhibiting exercise-induced desaturation performed cyclo-ergometer endurance exercise at 60%of their maximal workload in two randomized conditions: air vs. O₂. O₂ was adjusted to ensure 90≤SpO₂≤95%. Endurance time (Tlim), dyspnoea, ventilation ( ), breathing frequency (fb), tidal volume (V_T), cardiac output (CO), heart rate (HR) and arterio-venous difference in oxygen (AVD) were compared between conditions.

The comparison of whole group performance between conditions revealed no differences, but individual analysis showed that O₂ increased Tlim for 14 patients [+68%; p<0.01; (positive responders)], decreased it for seven [−36%; p<0.05; (negative responders)] and induced no change for four (non-responders). For positive responders, improved performance was supported by reduced dyspnoea,  , fb, HR and CO and increased AVD. For negative responders, hyperoxia resulted in increased dyspnoea and fb without change in   or cardiovascular parameters.

For comparable correction of exercise desaturation, O₂ does not induce similar effects on exercise responses in all patients. These results were confirmed in complementary study with 11 consecutives patients at higher exercise intensity. For R+, we recorded the classic and expected O₂ effects on cardiorespiratory adaptations (i.e. reduced ventilatory demand and cardiac output). In the other group, exercise breathing frequency and dyspnoea were paradoxically increased despite desaturation correction. However, this study must be considered as pilot study, which will need to be confirmed in future studies conducted on a larger case series.