The aim of this study is to investigate the effects of constraint-induced movement therapy on stroke patients who had intact cognition and some voluntary finger extension and to identify optimal protocols to apply this therapy method. We searched PubMed, Cochrane Library, and Embase for randomized controlled trials conducted prior to January 2022. The outcomes included the Motor Activity Log, Fugl-Meyer Assessment, and Wolf Motor Function Test. The inverse variance method fixed-effect model as well as the DerSimonian and Laird estimator random-effects model were applied, and the mean difference was calculated with 95% confidence interval to measure continuous outcomes. Six randomized controlled trials involving a total of 169 patients with stroke were enrolled. Compared with conventional rehabilitation methods, there was no significant effect of constraint-induced movement therapy when evaluated by the Motor Activity Log, including the amount of use (random-effect, standardized mean difference 0.65; 95%, confidence interval: -0.23-1.52) and quality of movement (random-effect, standardized mean difference 0.60; 95% confidence interval: -0.19-1.39). However, among patients with chronic stroke symptoms, meta-regression analyses showed better performance with a constraint time of at least 6 h per day and 6 h training per week when assessing the amount of use (p = 0.0035) and quality of movement (p = 0.0031). Daily intervention time did not lead to a significant difference in functional upper limb performance. An efficient protocol of constraint-induced movement therapy designed as 6 h of training per week with 6 h constraint per day could bring significant stroke symptom improvement to patients with chronic stroke.
Keywords: Fugl-Meyer assessment; cognitive function; constraint-induced movement therapy (CIMT); moto activity log; stroke; wolf motor function test.