Walkbot-Based Robot-Assisted Gait Training and Phase-Specific Lower-Limb Torque in Parkinson's Disease: A Retrospective Comparative Study.

Gait impairment, postural instability, and freezing of gait are major contributors to mobility limitations in Parkinson's disease (PD). Robot-assisted gait training (RAGT) has emerged as a promising rehabilitation strategy; however, its association with phase-specific lower-limb kinetic outcomes remains insufficiently characterized. This study compared Walkbot-based RAGT and conventional rehabilitation in terms of Walkbot-derived phase-specific lower-limb torque, functional balance, self-reported freezing-related gait impairment, and health-related quality of life in individuals with PD. This retrospective comparative study included 60 individuals with idiopathic PD (Hoehn and Yahr stages II-III) who received either Walkbot-based RAGT (n = 30) or conventional rehabilitation (n = 30). Both groups completed 30 supervised sessions over 6 weeks. Biomechanical outcomes comprised Walkbot-derived lower-limb torque during the stance and swing phases for the clinically more affected and contralateral limbs. Clinical outcomes included the Berg Balance Scale (BBS), Freezing of Gait Questionnaire (FOG-Q), and Parkinson's Disease Questionnaire-39 (PDQ-39). Between-group post-treatment differences were evaluated using analysis of covariance (ANCOVA), adjusting each outcome for its corresponding baseline value. After baseline adjustment, post-treatment phase-specific lower-limb torque was higher in the RAGT group for the more affected limb during swing (adjusted difference, 0.057 Nm/kg; 95% CI, 0.019-0.096; p = 0.004) and stance (0.145 Nm/kg; 95% CI, 0.098-0.193; p < 0.001), and for the contralateral limb during swing (0.051 Nm/kg; 95% CI, 0.006-0.096; p = 0.027) and stance (0.205 Nm/kg; 95% CI, 0.088-0.321; p = 0.001). The RAGT group also had higher baseline-adjusted BBS scores (adjusted difference, 5.78 points; 95% CI, 3.86-7.69; p < 0.001) and lower FOG-Q scores (-0.94 points; 95% CI, -1.27 to -0.61; p < 0.001). No significant between-group difference was observed for PDQ-39 (0.77 points; 95% CI, -0.16 to 1.69; p = 0.102). Walkbot-based RAGT was associated with more favorable baseline-adjusted phase-specific lower-limb torque, functional balance, and self-reported freezing-related gait impairment than conventional rehabilitation, whereas disease-specific quality of life did not differ significantly between groups. These findings highlight the potential value of integrating Walkbot-derived phase-specific kinetic assessment with clinical outcomes to provide a more quantitative characterization of rehabilitation-related locomotor changes in PD. Prospective randomized studies are needed to confirm these associations and establish the clinical utility of Walkbot-derived phase-specific lower-limb torque as an outcome measure.