Quantifying gait parameters to determine intensity of gait perturbations prescribed by physical therapists to train subjects with balance deficits

Abstract

Gait perturbation training for fall prevention is gaining recognition over regular strength and balance exercises. Difficulty of perturbations chosen varies between interventions and the most effective intensity is unknown. Some physiotherapists already use perturbations in clinical settings. Our objective was to quantify the changes in balance parameters during gait perturbations selected by an experienced physiotherapist. Seven subjects with moderate to severe traumatic brain injury and deficits in dynamic balance participated. Three-dimensional whole-body motion and kinetics analysis was performed, after one habituation session. Self-imposed (head rotations) or physiotherapist-imposed (repeated or random slips and trips induced by changes in belt speed) perturbations were applied while walking on a split-belt treadmill. Relative anteroposterior position (APP) and velocity of the center of mass (COM) and center of pressure (COP) were compared between perturbations and natural and fast gait speed. No differences were found in relative APP of COM and COP for physiotherapist-imposed slip-like or self-imposed perturbations compared to natural gait ( P > 0.134). Physiotherapist-imposed trip-like perturbations caused significantly more forward relative APP of COM and COP than during fast gait ( P ≤ 0.05). Velocity of COM and COP was significantly greater during physiotherapist-imposed ( P ≤ 0.001), but not self-imposed ( P > 0.106), perturbations than at natural gait speed. COM velocity remained lower than during fast gait for all perturbations ( P ≤ 0.05), but COP velocity in physiotherapist-induced perturbations reached fast gait COP velocity ( P > 0.279). Balance-related gait parameters were not affected during self-imposed perturbations, while physiotherapist-imposed trip-like perturbations altered parameters the most. During physiotherapist-imposed perturbations, balance control (COP-related) parameters reached values equal or greater than at fast gait speed. Meanwhile, the level of balance quantified by COM-related parameters, reached levels similar to fast gait for relative APP but not velocity. Future research should examine the possibility of personalizing perturbations to train specific balance parameters based on patient needs.