Stevenson · Health technology assessment (Winchester, England) 2009 · systematic review and economic evaluation · n=5 studies

Vitamin K to prevent fractures in older women: systematic review and economic evaluation.

Cited 91 times in the scientific literature.

Level 1 - systematic review of randomized trials

Systematic review of randomized controlled trials

PubMed 19818211 · doi:10.3310/hta13450 · record verified 2026-08-29

What was done

A systematic review evaluated randomized trials assessing vitamin K (phylloquinone/K1 or menatetrenone/K2) in postmenopausal women with osteoporosis or osteopenia, searched across multiple medical databases up to March 2009. Methodological quality was appraised using standard methods, meta-analysis was performed where appropriate, and a mathematical decision-analytic model evaluated the cost-effectiveness of vitamin K1 against alendronate, risedronate, and strontium ranelate.

What was found

Five trials (14 publications) met inclusion criteria. One good-quality double-blind trial in Canadian women with osteopenia found phylloquinone (5 mg) significantly reduced clinical fractures versus placebo (relative risk 0.46, 95% CI 0.22 to 0.99) with no increase in adverse events. Four open-label Japanese trials tested menatetrenone (45 mg) in osteoporosis; three very small trials found reduced morphometric vertebral fractures, but the larger Osteoporosis Fracture study found no reduction. No significant difference was found for non-vertebral fractures across the small trials. In the economic model, vitamin K1 and alendronate were more cost-effective than risedronate or strontium ranelate, but high uncertainty remained regarding vitamin K1 versus alendronate.

Why it matters

While vitamin K1 demonstrated a reduction in clinical fractures, the overall evidence base is currently too uncertain to change standard first-line prescribing guidelines favoring alendronate.

Limits

Only five trials were identified. All four menatetrenone trials were open-label, three were very small (fewer than 100 participants per group), adverse event reporting was generally poor, and results conflicted between small and larger trials. The economic model depended heavily on unconfirmed efficacy assumptions for hip and vertebral fracture prevention.

Cited by