Abstract
BACKGROUND: Robot-assisted gait training (RAGT) and transcutaneous spinal cord stimulation (tSCS) are emerging interventions for incomplete spinal cord injury (iSCI). While tSCS modulates spinal excitability, it remains unclear whether its efficacy is enhanced when paired with high-dose robotic stepping compared to conventional physical therapy (CPT). METHODS: This prospective, double-blind, randomized controlled trial evaluated 20 adults with iSCI (AIS B-D). Participants were randomized to RAGT + tSCS (n = 13) or CPT + tSCS (n = 7). Both groups received 40 sessions over 8 weeks (5 sessions/week), including 20 min of tSCS using identical parameters. Primary outcomes were the Berg Balance Scale (BBS) and Timed Up and Go (TUG). Secondary outcomes included 10-m walk test (10MWT) speed, Walking Index for SCI II (WISCI-II), ASIA motor score, and intramuscular electromyography (iEMG) of lower-limb musculature. Data were analyzed using repeated-measures MANOVA and univariate general linear models (GLM) to test group × time effects across outcomes, with ANCOVA adjustment for baseline BBS to refine precision. RESULTS: A significant multivariate group × time interaction was observed (Pillai's trace = 0.982, p < 0.001). Univariate models demonstrated significantly greater gains in the RAGT + tSCS group for BBS (F (1,18) = 42.465, p < 0.001), TUG (F (1,18) = 97.500, p < 0.001), WISCI-II (p = 0.009), and ASIA motor score (p < 0.001). At 8 weeks, the RAGT + tSCS group outperformed the CPT + tSCS group in adjusted BBS (mean difference + 7.67; 95% CI, 5.14-10.19; p < 0.001), TUG (- 10.91 s, p < 0.001), WISCI-II + 1.26 points (p = 0.021), and ASIA score (+ 7.93 points, p < 0.001). iEMG revealed significantly higher activation in the tibialis anterior, vastus lateralis, and gastrocnemius medialis for the RAGT group (all p < 0.001). Conversely, 10MWT speed improved similarly across both arms (p = 0.708). CONCLUSIONS: Pairing tSCS with task-specific robot-assisted stepping yields superior improvements in balance, functional mobility, and neuromuscular activation compared to dose-matched CPT. These results suggest that state-dependent neuromodulation during guided stepping effectively amplifies motor recovery, though walking speed may require additional overground progression to differentiate between modalities. TRIAL REGISTRATION: The trial is registered in the Chinese Clinical Trial Registry. REGISTRATION NUMBER: ChiCTR2300074090. Registration date: 2023 July 30. Chinese Clinical Trial Registry reference: www.chictr.org.cn.