ALTERATIONS IN LOCOMOTOR PERFORMANCE IN INDIVIDUALS WITH HEMIPLEGIA POST-STROKE FOLLOWING ROBOTIC- OR THERAPIST-ASSISTED LOCOMOTOR TRAINING

摘要

Purpose/Hypothesis: Previous evidence has suggested that intensive locomotor training using a treadmill and body weight support can improve walking ability in individuals with hemiplegia post-stroke. Unfortunately, delivery of intensive locomotor training may be limited by the labor intensive demands on the therapists. Robotic locomotor devices have been developed to assist in locomotor training, although specific limitations of many devices are reduced muscle activity and metabolic costs during robotic- vs. therapist-assisted treadmill walking, and may minimize locomotor adaptations. The purpose of this study is to investigate changes in locomotor performance in ambulatory subjects with post-stroke hemiplegia following robotic- vs. therapist-assisted locomotor training. We hypothesized that locomotor adaptations would be improved in both groups, although greater changes would be observed in the therapist-assisted group. Number of Subjects: Forty-eight ambulatory subjects (<.8m/s with chronic (>6 months) hemiplegia following unilateral stroke. Materials/Methods: Subjects were blocked by ambulation speed (< or > 0.5 m/s) and randomized to therapist- or robotic-assisted locomotor training groups. Each group participated in 30 minutes of therapist- or robotic-assisted walking 3x/week for 4 weeks at up to 3.0km/hr walking speed. Robotic- and therapist-assisted training was performed with the assistance of a single therapist. Body weight support was initiated at 30–40% of body weight and decreased as tolerated. Primary outcome measures included spatio-temporal gait characteristics during overground walking, and walking performance during the 10m and 6 minute walk. Secondary measures included Berg Balance Scale, manual muscle test scores, Modified Ashworth Scale, and Emory Functional Ambulation Profile. Results: Therapist- and robotic-assisted locomotor training resulted in significant increases in gait speed. However, those in the therapist assisted group improved to a greater extent (0.13 vs. 0.07 m/s). Preliminary analyses suggest that improvements in both cadence and stride length accounted for the differences between groups, with no differences in step length or timing asymmetry. Conclusions: Intensive walking training using robotic- or therapist-assistance yields significant improvements in walking in ambulatory subjects with chronic hemiplegia, with greater improvements observed following therapist-assisted walking. While robotic devices may be helpful in providing assistance in subjects who are not ambulatory, the use of robotic locomotor devices in individuals who ambulate does not maximize improvements in locomotor performance. Clinical Relevance: Evidence of changes in locomotor performance following specific gait training paradigms assists therapists in decision making to provide the best available treatment for recovery of locomotion following stroke.