Perturbation recovery time as a valid measure of balance control in human running

Abstract Balance control during running may be critical for performance and injury prevention, but the lack of valid and interpretable metrics limits research. This study introduces perturbation recovery time (PRT) as a novel measure of running-specific balance control, defined as the time required for whole-body angular momentum (WBAM) to return to statistically defined steady-state dynamics following a perturbation. Thirty recreational runners completed treadmill trials under normal running and two artificial balance impairment conditions (restricted arm swing and weighted vest loading). Participants were exposed to pseudorandomized anterior–posterior and mediolateral perturbations of varying magnitudes. PRT was computed using a functional data analysis approach based on 95% prediction bands derived from baseline WBAM dynamics. Peak deviation (PD), peak-to-peak range (PPR) and maximum Lyapunov exponent (MLE) were calculated for comparison. Detection accuracy was defined as the proportion of runners showing larger values in impaired vs. normal conditions and in higher vs. lower perturbation intensities. PRT demonstrated the highest overall detection accuracy (80.8 ± 10.5%) compared to PPR (75.0 ± 6.0%), PD (73.0 ± 11.8%) and MLE (34.0 ± 28.1%). PRT detected balance impairment induced by restricted arm swing in 82.8% and by weighted vest in 66.7% of runners, respectively. It also showed high sensitivity to perturbation intensity (86.8%). In contrast, deviation-based WBAM metrics failed to detect impairment in the weighted vest condition and MLE showed poor discriminatory ability in all comparisons. Notably, MLE was negatively associated with PRT, suggesting that high local dynamic stability predicts reduced ability to recover from large perturbations. These findings support PRT as a valid and interpretable measure of balance control in running, potentially capturing aspects of recovery dynamics not captured by existing metrics. PRT may provide a useful tool for assessing balance control in athletic and clinical populations, although further work is needed to establish its relationship with performance and injury risk.