Abstract
Background: Cerebral palsy is a complex lifespan disability caused by a lesion to the immature brain. Evaluation of interventions for children with cerebral palsy requires valid and reliable outcome measures. Motor development curves and reference percentiles for The Gross Motor Function Measure (GMFM-66) are valuable tools for following, predicting, comparing, and evaluating changes in gross motor skills. The aims of this study were to create motor development curves with reference percentiles based on Norwegian data and compare them with published counterparts for Canadian children aged 2-21 years. Method: Prospective population-based cohort data from the Norwegian Quality and Surveillance Registry for Cerebral Palsy (NorCP) for 1206 children with 3612 GMFM-66 tests between 0.5 and 17.3 years of age. Median development by Gross Motor Function Classification System (GMFCS) levels was estimated using a generalized additive regression model with smoothed parameters for location, scale, and shape (based on the R GAMLSS library). To adjust for repeated individual measurements, we report the median curve of 100 random samples with only one observation per observed child. Results: The Norwegian motor development curves for GMFCS levels I-IV increase up to 7 years of age before flattening off, while GMFCS level V curves are relatively flat. Overall, both motor development curves and GMFM-66 percentiles are very similar to Canadian counterparts. Conclusions: The existing Canadian reference curves are valid also for Norway, working well for both clinical and research applications. However, Norwegian percentiles can be used from an earlier age.