Two BMJ meta-analyses on vitamin D have found that evidence for 
benefit is narrow and that if there is a benefit, the form of 
supplementation has a bearing on its magnitude.

One, an “umbrella review” of meta-analyses and systematic reviews, 
found no “highly convincing” evidence linking circulating vitamin D 
levels with any of 137 outcomes, such as colorectal cancer or 
hypertension. It found a “probable” link for only three outcomes: 
birth weight, parathyroid hormone levels in dialysis patients, and 
dental caries in children.

The other meta-analysis found, in observational data, a salutary 
effect of increasing vitamin D levels on mortality. In randomized, 
controlled trials, the analysis found a significant mortality benefit 
from vitamin D3 supplements, but not from the D2 form.

Editorialists champion new trials focusing on risks as well as 
benefits of supplementation, and advise against measuring circulating 
vitamin D beyond bone-disease-related conditions.

BMJ article 1