Effects of movement imagery and electromyography-triggered feedback on arm-hand function in stroke patients in the subacute phase

Objective: To investigate the effects of movement imagery-assisted electromyography (EMG)-triggered feedback (focused on paretic wrist dorsiflexors) on the arm-hand function of stroke patients. Design: Single-blinded, longitudinal, multicentre randomized controlled trial. Measurements were performed (on average) 54 days post stroke (baseline), three months later (post training) and at 12 months post baseline. Setting: Two rehabilitation centres. Subjects: Twenty-seven patients with a first-ever, ischaemic, subacute stroke. Interventions: A reference group received conventional electrostimulation, while the experimental group received arm-hand function training based on EMG-triggered feedback combined with movement imagery. Both groups were trained for three months, 5 days/week, 30 minutes/day, in addition to their therapy as usual. Main measures: Arm-hand function was evaluated using the upper extremityrelated part of the Brunnstrom Fugl-Meyer test and the Action Research Arm test. Results: During training, Brunnstronn Fugl-Meyer scores improved 8.7 points and Action Research Arm scores by 19.4 points (P<0.0001) in both groups relative to baseline results, rising to 13.3 and 28.4 points respectively at one year follow-up (P<0.0001). No between-group differences were found at any time. Conclusions: EMG-triggered feedback stimulation did not lead to more arm-hand function improvement relative to conventional electrostimulation. However, in contrast to many clinical reports, a significant improvement was still observed in both groups nine months after treatment ceased.