Trial sequential meta-analysis.
Link: http://www.thelancet.com/journals/landia/article/PIIS2213-8587%2813%2970212-2/abstract
From interpretation:
"Our findings suggest that vitamin D supplementation with or without calcium does not reduce skeletal or non-skeletal outcomes in unselected community-dwelling individuals by more than 15%. Future trials with similar designs are unlikely to alter these conclusions."
Comments
Paul Christiano
A 5% change in mortality / 15% change in other endpoints would be surprisingly large to me. Does this analysis conclude anything stronger? Is it underpowered to notice smaller effects? Is there some reason we should expect effects this large?