Robot-assisted gait training (RAGT) may have beneficial effects on Parkinson's disease (PD); however, the evidence to date is inconsistent.

This study compared the effects of RAGT and treadmill training (TT) on gait speed, dual-task gait performance, and changes in resting-state brain functional connectivity in individuals with PD.

In this prospective, single-center, randomized controlled trial with a parallel two-group design, 44 participants were randomly allocated to undergo 12 sessions (3 times per week for 4 weeks) of RAGT or TT. The primary outcome was gait speed on the 10-m walk test (10mWT) under comfortable walking conditions. Secondary outcomes included dual-task interference on gait speed, balance, disability scores, fear of falling, freezing of gait, and brain functional connectivity changes. All clinical outcomes were measured before (T0), immediately after (T1), and 1 month after treatment (T2).

The mean (SD) age of the participants was 68.1 (8.1) years, and mean disease duration 108.0 (61.5) months. The groups did not significantly differ on the 10mWT (T0-T1, p = 0.726, Cohen's d = 0.133; T0-T2, p = 0.778, Cohen's d = 0.121). We observed a significant time-by-group interaction (F = 3.236, p = 0.045) for cognitive dual-task interference, controlling for confounders. After treatment, coupling was decreased to a greater extent with RAGT than TT between the visual and dorsal attention networks (p = 0.015), between bilateral fronto-parietal networks (p = 0.043), and between auditory and medial temporal networks (p = 0.018). Improvement in cognitive dual-task interference was positively correlated with enhanced visual and medial temporal network coupling overall (r = 0.386, p = 0.029) and with TT (r = 0.545, p = 0.024) but not RAGT (r = 0.151, p = 0.590).

RAGT was not superior to intensity-matched TT on improving gait functions in individuals with PD but may be beneficial in improving gait ability under cognitive dual-task conditions. The therapeutic mechanism and key functional connectivity changes associated with improvement may differ between treatment strategies.

#NCT03490578