Complex Skill Training Transfers to Improved Performance and Control of Simpler Tasks After Stroke.

Abstract

Given limited therapy time, it is important to practice tasks that optimize transfer to other tasks that cannot be practiced during therapy. However, little is known about how tasks can be selected for practice to optimize generalization. One dimension of task selection is the complexity of the task. The purpose of the current study was to test if learning of a complex motor skill with the paretic arm would transfer to a simpler unpracticed goal-directed reaching task. This is an observational study, repeated measures design. Fifteen participants with mild-to-moderate stroke practiced a complex motor skill using their paretic arm for 2 consecutive days. Complex skill learning was quantified using change in the speed-accuracy trade-off from baseline to 1 day and 1 month post-practice. Motor transfer was assessed as the change in goal-directed planar reaching performance and kinematics from 2 baselines to 1 day and 1 month post-practice. Nine additional participants with stroke were recruited as the test-alone group who only participated in the transfer tests to rule out the effects of repeated testing. Practice improved the speed-accuracy trade-off for the practiced complex skill that was retained over a period of 1 month. Importantly, complex skill practice, but not repeated testing alone, improved the long-term performance and kinematics of the unpracticed simpler goal-directed planar reaching task. Improvements in the unpracticed transfer task (reaching) strongly correlated with improvements in the practiced complex motor skill. We did not have a comparison stroke group that practiced task-specific reaching movements. Given the limited number of tasks that can be practiced during therapy, training complex tasks may have an added advantage of transfer to improved simpler task performance.