Jonas Schröder, Steven Truijen, Tamaya Van Criekinge, Wim Saeys
Rehabilitation Sciences & Physiotherapy, University of Antwerp, 2610 Wilrijk, Belgium: E-mail: jonas.schroder@uantwerpen.be
DOI: 10.2340/16501977-2505

Background: Pre-clinical evidence suggests a period
early after stroke during which the brain is most receptive to rehabilitation, if it is provided as high-dose motor training.
Objective: To evaluate the feasibility of repetitive gait training within the first 3 months post-stroke and the effects on gait-specific outcomes.
Methods: PubMed, Web of Science, Cochrane Library, Rehab Data and PEDro databases were searched systematically. Randomized controlled trials were included to descriptively analyse the feasibility and quantitatively investigate the effectiveness of repetitive gait training compared with conventional
therapy.
Results: Fifteen randomized controlled trials were included. Repetitive training can safely be provided through body weight support and locomotor assistance from therapists or a robotic device. No difference in drop-out rates was reported despite the demanding nature of the intervention. The meta-analysis yielded significant, but small, effects on walking independence and endurance. Training with end-effector robots appears most effective.
Conclusion: Robots enable a substantial, yet feasible, increase in the quantity of walking practice early post-stroke, which might enhance functional recovery. However, the mechanisms underlying these effects remain poorly understood.

Animal models suggest that rehabilitation provided as repetitive motor training is most effective early after stroke. To investigate whether such a rehabilitation approach can enhance long-term walking recovery in human patients with stroke, this review gathered clinical studies on the effects of repetitive gait training. We found that robots, in particular, provide a substantial, yet feasible, increase in the amount of walking practice in those stroke patients who are unable to walk. This increase in rehabilitation dose improves walking ability in the long-term. However, these effects are inconsistent, rather small, and in contrast with neutral effects on motor functions of the paretic leg. Therefore, the effects of repetitive training in the context of early stroke rehabilitation remain poorly understood and further research is required.

Figure S1-S7
Table SI-II