Vitamin K2 for bones and arteries: strong marketing, thin trial evidence

Vitamin K2, or menaquinone, is sold as the nutrient that steers calcium into bone and away from arteries, and there is a genuine observational signal behind that pitch — but when the idea has been put to randomised trials with hard endpoints, it has mostly not held up [s1][s2]. The strongest population data link high dietary K2 to less heart disease and less arterial calcification, yet the one randomised trial to measure calcification directly found no effect, and vitamin K supplements have not been shown to preserve bone density [s2][s4].

Where the story starts: the observational signal

The marketing narrative traces to the Rotterdam Study, a population cohort of 4,807 adults with no history of heart attack, followed from the early 1990s to 2000 [s1]. Comparing the highest third of dietary menaquinone intake with the lowest, the researchers found a relative risk of coronary heart disease mortality of 0.43 (95% confidence interval 0.24 to 0.77), a lower risk of all-cause mortality (0.74; 0.59 to 0.92) and about half the odds of severe aortic calcification (odds ratio 0.48; 0.32 to 0.71) [s1]. Notably, phylloquinone — vitamin K1, the form in leafy greens — was not related to any of these outcomes, which is what launched the idea that K2 specifically matters [s1]. But this is an association: people who eat more menaquinone, found largely in cheese, other fermented foods and animal products, differ in many ways from those who eat less, and a cohort cannot prove the vitamin is doing the work.

The randomised test of the artery claim

The cleanest test of the calcification idea is a 2022 trial that gave the vitamin and then imaged the arteries. Researchers randomised 365 older men with an aortic valve calcification score above 300 arbitrary units to a daily tablet of 720 micrograms of menaquinone-7 plus 25 micrograms of vitamin D, or to placebo, for 24 months [s2]. The valve calcification score rose in both groups, and the mean difference in progression between them was just 17 arbitrary units, which was not statistically significant (P=0.64) [s2]. There was no difference in aortic valve area (0.02 cm², P=0.78), peak jet velocity, heart-valve surgery, death or cardiovascular events [s2]. The supplement did what it is biochemically supposed to do — it sharply lowered dephosphorylated-uncarboxylated matrix Gla-protein, a marker of vitamin K status (-212 versus 45 pmol/L; P<0.001) — and yet the calcification it was meant to slow progressed just the same [s2]. That gap between moving a biomarker and moving the disease is the recurring trap in supplement science.

One softer signal exists. A three-year trial gave 244 healthy postmenopausal women either 180 micrograms of menaquinone-7 a day or placebo and found that carotid-femoral pulse-wave velocity and the arterial Stiffness Index improved in the supplement group [s3]. That is a surrogate measure of arterial stiffness, not a count of heart attacks or strokes, and it sits against the null calcification result rather than overturning it [s2][s3].

The bone claim

For bone, the best long-term randomised evidence used vitamin K1 rather than K2, but it is the most rigorous test available and it was negative on the primary outcome. The ECKO trial randomised 440 postmenopausal women with osteopenia to 5 milligrams of vitamin K1 daily or placebo for two to four years [s4]. Over two years, bone mineral density fell almost identically in both groups: the between-group difference was -0.06% at the lumbar spine (95% confidence interval -0.67 to 0.54) and 0.19% at the total hip (-0.37 to 0.75) — no protection against age-related bone loss [s4]. The trial did record fewer clinical fractures in the vitamin K group (nine versus 20; P=0.04), but it was not designed or powered to measure fractures, the numbers were small, and the authors flagged this as a hypothesis to test rather than a finding to act on [s4].

How to read this

The honest verdict is a wide gap between a real observational signal and thin randomised confirmation. Eating menaquinone-rich foods tracks with less heart disease and less calcification, but the trial that gave the supplement and imaged the arteries found no effect, one trial improved a stiffness surrogate, and vitamin K did not preserve bone density [s1][s2][s3][s4]. For a well-nourished person, the case for K2 pills rests mostly on associations and biomarkers, not on the outcomes the marketing implies. This connects to the calcium-and-heart-disease question the K2 pitch often piggybacks on, to vitamin D in healthy adults, to what the multivitamin trials show, and — for the one setting where vitamin K clearly matters — to newborn vitamin K prophylaxis.

This article is informational and not dietary or medical advice; decisions about supplements should be made with a qualified clinician, particularly for anyone taking a vitamin K-dependent anticoagulant such as warfarin.