Time-variability of muscle oxygen saturation during graded maximal exercise

The time-variability of physiological and kinematic variables, extracted at mesoscopic and macroscopic levels, respectively, has shown potential in detecting changes in exercise workload and associated fatigue effects. However, the sensitivity of microscopic variables -such as muscle oxygen saturation, which reflect the dynamics of muscle metabolism-remains unexplored. This study aimed to compare the time-variability structure of the tissular saturation index (TSI) during a graded maximal exercise performed until exhaustion. Nineteen participants started running at 8 km/h with the speed increasing by 1 km/h every 100 s until they could not keep the prescribed velocity. The time-variability of TSI, recorded from the quadriceps, was analyzed using Detrended fluctuation analysis (DFA) and Sample entropy (SampEn) over the first and last 2048 recorded data points (corresponding to 204 s each). Wilcoxon test and Cohen's d were used to compare the initial and final parts of the test. Results revealed a significant decrease in the Hurst (H) exponent (from H = 0.84 ± 0.21 to H = 0.49 ± 0.10; p < 0.01; d = -1.57) and a corresponding increase in SampEn (from 1.12 ± 0.20 to 1.40 ± 0.13; p < 0.01; d = 1.17). These findings indicate a shift towards uncorrelated white-noise as exhaustion approached, suggesting reduced efficacy of oxygen transportation with increasing workloads. The time-variability of muscle oxygen saturation appears to be a) a promising measure for assessing exercise intensity, and b) allow the study of physiological network interactions extracted from different levels (from microscopic to macroscopic).