Comparison of two techniques of robot-aided upper limb exercise training after stroke

Objective: This study examined whether incorporating progressive resistive training into robot-aided exercise training provides incremental benefits over active-assisted robot-aided exercise for the upper limb after stroke. Design: A total of 47 individuals at least 1 yr poststroke were enrolled in this 6-wk training protocol. Paretic upper limb motor abilities were evaluated using clinical measures and a robot-based assessment to determine eligibility for robot-aided progressive resistive training at study entry. Subjects capable of participating in resistance training were randomized to receive either active-assisted robot-aided exercises or robot-aided progressive resistance training. Subjects who were incapable of participating in resistance training underwent active-assisted robotic therapy and were again screened for eligibility after 3 wks of robotic therapy. Those subjects capable of participating in resistance training at 3 wks were then randomized to receive either robot-aided resistance training or to continue with robot-aided active-assisted training. Results: One subject withdrew due to unrelated medical issues, and data for the remaining 46 subjects were analyzed. Subjects in all groups showed improvement in measures of motor control (mean increase in Fugl-Meyer of 3.3; 95% confidence interval, 2.2-4.4) and maximal force (mean increase in maximal force of 3.5 N, P = 0.027) over the course of robot-aided exercise training. No differences in outcome measures were observed between the resistance training groups and the matched active-assisted training groups. Subjects' ability to perform the robotic task at the time, of group assignment predicted the magnitude of the gain in motor control. Conclusion: The incorporation of robot-aided progressive resistance exercises into a program of robot-aided exercise did not favorably or negatively affect the gains in motor control or strength associated with this training, though interpretation of these results is limited by sample size. Individuals with better motor control at baseline experienced greater increases in motor control with robotic training.