Griffin et al., 2026. Voluntary engagement predicts recovery across rehabilitation time points after experimental spinal cord injury
Description
This dataset supports the systematic review and multilevel meta-analysis reported in “Voluntary engagement predicts recovery across rehabilitation time points after experimental spinal cord injury.” The dataset contains study characteristics, experimental demographics, rehabilitation parameters, behavioural outcomes, effect-size calculations, and citations for 150 preclinical studies of rehabilitation following spinal cord injury in adult rodent models. Collectively, the included studies comprise more than 3,000 animals and 292 effect sizes. Variables include species and strain, injury level and model, injury severity, rehabilitation modality and subclassification, voluntary or forced participation, task specificity, weight support, rehabilitation onset, daily exposure, total intervention duration, behavioural outcome measure, group sample sizes, means and measures of variability, and calculated standardized mean differences. The repository also includes the analysis files used for the multilevel random-effects meta-analyses and meta-regressions examining rehabilitation type, total duration, initiation phase, daily exposure, delivery mode, and estimated stress burden. These files are provided to support transparency, reproducibility, secondary analysis, and future evidence synthesis in preclinical spinal cord injury rehabilitation research.
Files
Institutions
- University of AucklandAuckland, Auckland
- German Center for Neurodegenerative DiseasesNorth Rhine-Westphalia, Bonn
- ETH ZurichZurich, Zurich
Categories
Funders
- CatWalk Spinal Cord Injury TrustWellington Region, MastertonGrant ID: Postdoctoral Research Fellowship
- International Foundation for Research in ParaplegiaGeneva, Geneva
- Ministerium für Kultur und Wissenschaft des Landes Nordrhein-WestfalenNorth Rhine-Westphalia, DüsseldorfGrant ID: ibehave Network Nordrhein-Westfalen
- European CommissionEuropean UnionBrussels Capital, BrusselsGrant ID: ERANET: Axon repair and Rater SCI