Safety and effectiveness of non-invasive brain stimulation on mobility and balance function in children with cerebral palsy: a systematic review and meta-analysis.

Almost all the trials gave the stimulation on top of a training programme, so the size of the effect that comes from the stimulation itself cannot be separated from the training. Most children had mild to moderate cerebral palsy (98% were GMFCS level I to III, with only a few at level IV), so the results say very little about children with severe impairment. The trials were small and their stimulation protocols differed widely in site, dose, session number and outcome measures, so the pooled numbers mix treatments that are not really the same treatment. Follow-up is short. Only one month could be pooled, and fewer than two studies reported anything longer, with reported follow-ups ranging from five days to three months, so nothing can be said about whether any gain lasts. Several trials excluded children with epilepsy, so the safety findings do not cover them. Six of the 16 trials were rated only fair on the PEDro quality scale, and two tDCS trials gave a single stimulation session only. One trial's control group received virtual reality rather than sham stimulation, so that comparison is not stimulation against a placebo.

The paper names four funders: the Featured Clinical Technique of Guangzhou, the Guangzhou Municipal Science and Technology Project, a plan to enhance scientific research at Guangzhou Medical University, and the National Natural Science Foundation of China. The text supplied does not include a competing-interests statement, so it is not known whether the authors declared any.

Do not read this as proof that brain stimulation on its own makes children with cerebral palsy walk better. Nearly every trial added it to a training programme that both groups received, and the authors themselves say the training probably contributes to the result. Nor should you expect balance to improve. Balance measures did not change and Timed Up and Go did not improve, and most children in these trials had mild to moderate cerebral palsy, so the findings do not extend to children with more severe involvement.