Thirteen RCTs (n=1,366) were included in the review.
Only 2 studies reported using a method of allocation concealment. Attrition rates ranged from 0 to 30%. No evidence of significant publication bias was found.
Fractures (7 RCTs).
Vitamin K supplementation significantly reduced hip (OR 0.23, 95% CI: 0.12, 0.47), vertebral (OR 0.40, 95% CI: 0.25, 0.65) and all nonvertebral (OR 0.19, 95% CI: 0.11, 0.35) fractures. Pooled absolute risk differences also showed a significant positive effect of supplementation. When the data from the high-risk populations were removed, the effects on hip fracture were not statistically significant, but those on vertebral (4 RCTs; OR 0.40, 95% CI: 0.25, 0.65) and all nonvertebral fractures (2 RCTs; OR 0.24, 95% CI: 0.07, 0.84) were significant. There was no evidence of heterogeneity among studies for the OR, but there was heterogeneity for risk differences in nonvertebral fractures.
BMD.
All except one of the included studies showed an advantage of vitamin K supplementation in terms of BMD. Pooled data from 3 studies of BMD measured at the metacarpals showed a significant positive effect of supplementation (SMD 0.27, 95% CI: 0.03, 0.50).
Adverse events.
No study reported any serious adverse events associated with vitamin K supplementation.