Split‐Belt Treadmill for Falls, Gait Asymmetry, and Freezing in Parkinson's Disease

Background Individuals with Parkinson's disease (PD) can adapt to new gait patterns on a split‐belt treadmill (SBTM), a tool previously used to rehabilitate gait asymmetry because of stroke. Objectives The goal was to determine whether SBTM rehabilitation is superior to conventional “tied” treadmill (TM) in reducing falls, freezing of gait (FOG), and improving gait asymmetry. Methods Twenty‐eight participants with idiopathic PD and treatment‐resistant FOG underwent 18 sessions of SBTM and TM training. Primary outcome was the incidence of falls 3 months after training, and secondary measures included spatiotemporal gait variables (treadmill and over‐ground), motor scores, and performance on FOG and balance questionnaires. Results Treadmill training improved falls regardless of intervention (SBTM: P < 0.05; TM: P < 0.01), although benefits were not sustained at 3 months. SBTM did not meet the a priori 20% superiority threshold for falls prevention. Asymmetry index, temporal asymmetry, and stance time ( P ≤ 0.01 in both groups) reduced after training. Swing time increased after training with both interventions ( P ≤ 0.01). FOG and ABC scores also improved ( P ≤ 0.01 in both groups), however, these improvements were not sustained at 3 months. Cadence and over‐ground asymmetry index improved with both interventions ( P ≤ 0.01), although only reduced cadence was sustained at 3 months ( P = 0.05). Other over‐ground parameters including gait velocity, stride length, double‐support time improved with TM training only. Conclusions SBTM was not superior to TM training in improving incidence of falls. Both interventions improved FOG and spatiotemporal gait parameters, although TM training translated better to over‐ground walking.