Meta-analysis of the efficacy of different training strategies in animal models of ischemic stroke.

BACKGROUND AND PURPOSE Although several studies have shown beneficial effects of training in animal stroke models, the most effective training strategy and the optimal time to initiate training have not been identified. The present meta-analysis was performed to compare the efficacy of different training strategies and to determine the optimal time window for training in animal stroke models. METHODS We searched the literature for studies analyzing the efficacy of training in animal models of ischemic stroke. Training was categorized into forced physical training, voluntary physical training, constraint-induced movement therapy, and skilled reaching training. Two reviewers independently extracted data on study quality, infarct size, and neurological outcome. Data were pooled by means of a meta-analysis. RESULTS Thirty-five studies with >880 animals were included. A meta-analysis of all treatments showed that training reduced the infarct volume by 14% (95% confidence interval, 2%-25%) and improved the cognitive function by 33% (95% confidence interval, 8%-50%), the neuroscore by 13.4% (95% confidence interval, 1.5%-25.3%), and the running function by 6.6% (95% confidence interval, 1.4%-11.9%). Forced physical training reduced the infarct volume and enhanced the running function most effectively, whereas skilled reaching training improved the limb function most effectively. A meta-regression illustrated that training was particularly efficacious when initiated between 1 and 5 days after stroke onset. CONCLUSIONS Our meta-analysis confirms that training reduces the infarct volume and improves the functional recovery in animal stroke models. Forced physical training and skilled reaching training were identified as particularly effective training strategies. The efficacy of training is time dependent.