A systematic review examining the effect of vitamin D supplementation on functional outcomes post-stroke.

The included studies were small, and five of eight randomized trials had a high risk of bias; the two non-randomized studies had serious risk of bias. The overall certainty of evidence was very low, so the results cannot reliably guide practice. A meta-analysis could not be done because studies used different outcome measures, reported data incompletely, and were clinically heterogeneous. Vitamin D doses and schedules varied widely, with some studies giving vitamin D once and others giving it daily, weekly or monthly, so no optimal treatment can be inferred. Some studies included people regardless of vitamin D status, while others only included people who were deficient or insufficient, so it is unclear who might benefit. Adverse events were not actively collected in any study; only one study reported mortality, and the authors said this may be due to residual confounding or small sample size. Only a few reported improvements exceeded known minimal clinically important differences. The studies came from a limited range of geographic and cultural populations, and sociodemographic and clinical details were often underreported. All included studies came from non-native English-speaking countries, and the search may have missed other non-English trials.

The authors declared no conflicts of interest and received no financial support for the work.

Do not conclude that vitamin D is an effective treatment after stroke. The functional outcomes were mixed and very low certainty, and only one study of 53 patients reported lower mortality, which the authors said may be due to residual confounding.