Rehabilitation of Reaching Poststroke

Abstract

For individuals poststroke, reaching with the paretic limb is often associated with compensatory trunk movement due to limited active arm movement. We conducted a pilot feasibility study to compare the effects of task-related training on reaching with the paretic limb using each of 2 different techniques for providing feedback about trunk position. We hypothesized that the use of an auditory feedback signal in response to pressure (Sensor group) would be more effective than feedback arising from an external device used to restrain the trunk (Stabilizer group). Sixteen individuals with chronic stroke were enrolled. Participants had scores of 20 to 44 on the Upper-Arm subsection of the Fugl-Meyer Scale and demonstrated some trunk movement during the pretest assessment of the Reaching Performance Scale (RPS). Participants were randomly assigned to either the Sensor or the Stabilizer group. Both groups participated in 12 structured rehabilitation sessions with equal time duration, number of repetitions, and task-related training activities. Feedback was systematically and equally faded for all training. Clinical outcome measures were assessed prior to and following training. Both forms of feedback were associated with changes in active shoulder motion, Wolf Motor Function Test (WMFT), Fugl-Meyer Scale and RPS (near and far) as determined by a 2 x 2 (Group x Time) analysis of variance. An interaction of RPS-near revealed that the Sensor group improved more than the Stabilizer group for this measure. However, sample size may not have been sufficient to identify differences in other measures. Although both forms of feedback led to improvements in the majority of outcome measures, the auditory feedback signal was associated with greater improvement in RPS-near scores. In the rehabilitation of reaching function in individuals with stroke, the use of an auditory signaling device appears to be a feasible alternative to imposed trunk stabilization. Additional studies with larger sample sizes are needed to determine whether one approach is superior to the other.