Vitamin K2 is a fat-soluble vitamin that activates two proteins controlling where calcium goes: osteocalcin, which binds it into bone, and matrix Gla protein, which keeps it out of artery walls. The observational evidence is strong, the randomised evidence is mixed, and doses above 50 µg of MK-7 can interfere with warfarin.
Vitamin K2 has one of the widest gaps in nutrition between what the mechanism promises and what the trials deliver. The mechanism is genuinely elegant, and it is real: two vitamin K-dependent proteins that decide whether calcium ends up in your skeleton or in your arteries. The trials are a more mixed story than any page selling K2 will tell you.
This article gives both, with numbers, and covers the one thing a British reader needs before buying a bottle: how a 180 µg MK-7 capsule interacts with warfarin.
What is vitamin K2, and how is it different from K1?
Vitamin K is not one molecule. It is a group of vitamins with a shared role. Vitamin K1, phylloquinone, comes from green leafy plants and is most of the vitamin K in a British diet. Vitamin K2, menaquinone, comes from fermentation and animal tissue, and is a family within the family, its members numbered by side-chain length: MK-4, MK-7 and so on.
The distinction matters because the two behave differently in the body. K1 is taken up largely by the liver, where it supplies the clotting factors. K2 circulates for longer and reaches tissues outside the liver, including bone and the artery wall. That is the whole basis of the K2 case, and it is why the Rotterdam Study result below found an association for menaquinone intake and none at all for phylloquinone.
The NHS treats vitamin K as one group with one recognised function, blood clotting, and notes some evidence for bone health. It sets no separate recommendation for K2, which is a real difference from the American pages that dominate these results.
What is vitamin K2 taken for?
Vitamin K works as a cofactor for an enzyme that carboxylates specific proteins. Carboxylation is what switches those proteins on, and until it happens they circulate in an inactive form. Three of them matter here.
What vitamin K switches on
CLOTTING FACTORS
Made in the liver. This is the function the NHS recognises, and the reason warfarin works by blocking vitamin K.
OSTEOCALCIN
Made in bone. Once carboxylated it binds calcium into the bone matrix. Uncarboxylated osteocalcin is the marker used to measure vitamin K status in bone.
MATRIX GLA PROTEIN
Made in cartilage and the vessel wall. It inhibits calcification. Its uncarboxylated form, dp-ucMGP, tracks with arterial stiffness.
Within K2, the form on the label changes what you get. Schurgers and colleagues compared synthetic K1 with natto-derived MK-7 in healthy volunteers, publishing in Blood in 2006. Both were absorbed well, peaking in serum 4 hours after intake. The difference was the half-life: MK-7 persisted far longer, accumulating to serum levels 7 to 8 times higher during prolonged intake, and produced more complete carboxylation of osteocalcin.
| Form | Where it comes from | Behaviour in the body |
|---|---|---|
| K1, phylloquinone | Green leafy vegetables, vegetable oils, cereal grains | Taken up mainly by the liver, short half-life, supplies clotting factors |
| K2 as MK-4 | Animal tissue: liver, egg yolk, butter, some meat | Very short half-life, needs frequent or high dosing to raise blood levels |
| K2 as MK-7 | Fermented foods, principally natto. The usual supplement form | Long half-life, accumulates 7 to 8 times higher with continued intake, carboxylates osteocalcin more completely |
Bone: what the randomised trials found
The single best trial is Dutch. Knapen and colleagues, in Osteoporosis International in 2013, randomised 244 healthy postmenopausal women to 180 µg of MK-7 daily or placebo for 3 years, measuring bone mineral density of the lumbar spine, total hip and femoral neck by DXA. MK-7 significantly reduced the age-related decline in bone mineral density and in calculated bone strength indices at the femoral neck. The same postmenopausal bone endpoint has been moved by a very different intervention: a 12-month trial of collagen peptides raised spine and femoral neck density, with the formation and resorption markers shifting in the same direction as the scan.
Then the meta-analysis, and this is the part that is missing from every competing page. Mott and colleagues, in Osteoporosis International in 2019, updated the systematic review of oral vitamin K and bone across 36 studies. In postmenopausal or osteoporotic patients, the odds of any clinical fracture were lower with vitamin K, at an odds ratio of 0.72 with a confidence interval of 0.55 to 0.95. Then they restricted the analysis to trials at low risk of bias. The odds ratio moved to 0.76 with a confidence interval of 0.58 to 1.01, which crosses 1 and is no longer statistically significant. There was no difference at all in vertebral fractures.
Read that as it is meant. The signal points the right way and survives in the full dataset. It does not survive contact with the highest-quality subset. The review itself notes it was needed partly because the integrity of some earlier evidence had been questioned.
Heart and arteries: strong observation, weaker trials
The observational case is genuinely striking. The Rotterdam Study followed 4 807 people with dietary data and no history of myocardial infarction from the early 1990s to 2000. Geleijnse and colleagues reported in the Journal of Nutrition in 2004 that, comparing the highest tertile of dietary menaquinone intake with the lowest, the relative risk of coronary heart disease mortality was 0.43 with a confidence interval of 0.24 to 0.77. All-cause mortality was 0.74, and severe aortic calcification had an odds ratio of 0.48. Phylloquinone intake was related to none of these outcomes.
The intervention data is thinner. Knapen and colleagues followed the same 244 women from their bone trial and reported in Thrombosis and Haemostasis in 2015 that after 3 years of 180 µg MK-7 daily, carotid-femoral pulse wave velocity and a stiffness index both improved against placebo, alongside a fall in dephosphorylated uncarboxylated matrix Gla protein.
But when Vlasschaert and colleagues gathered every controlled trial of vitamin K supplementation with a cardiovascular surrogate endpoint, publishing in Nutrients in 2020, they found nine randomised trials and concluded that vitamin K does not consistently prevent the progression of calcification, atherosclerosis or arterial stiffening. They noted there may be some benefit in people who already had calcification when they entered a study, and that follow-up across the field was relatively short.
How much confidence each vitamin K2 claim deserves
Which foods have vitamin K2?
K2 comes from two places: bacterial fermentation, and the tissue of animals that have converted K1 into it. The list is short and it is not a list of health foods.
- Natto, fermented soybeans, by an enormous margin the richest source and the only food that supplies MK-7 in quantity. It is also, to most British palates, an acquired taste.
- Hard and aged cheeses, particularly Dutch styles such as Gouda and Edam, where the bacterial cultures produce menaquinones during ripening.
- Egg yolk, especially from hens fed on pasture.
- Liver and organ meats, and to a lesser extent chicken and other meat.
- Butter and full-fat dairy, which is why K2 intake fell in Britain as low-fat products replaced them.
- Sauerkraut and other fermented vegetables, in modest amounts.
Bananas contain essentially no vitamin K2, and very little vitamin K of any kind. The question comes up often enough to be worth answering plainly: fruit is not a source. Green leafy vegetables are an excellent source of K1 and a poor source of K2, which are not interchangeable.
How much do you need, and should you take K2 with D3?
The NHS figure is for vitamin K as a whole, not for K2. Adults need approximately 1 microgram a day for each kilogram of body weight, which puts a 70 kg adult at about 70 µg of vitamin K daily. The NHS adds that a varied diet should supply all you need, that there is not enough evidence to know the effects of taking high doses long term, and that taking 1 mg or less a day of supplements is unlikely to cause harm.
That sits well below the 90 to 120 µg quoted by the American pages that dominate searches for vitamin K2 benefits. Neither figure is wrong. They are different reference systems, and the British one is what your GP works from.
On the D3 question. Vitamin D increases calcium absorption from the gut. Vitamin K2 influences where that calcium is deposited. The pairing is therefore mechanistically sensible, and it is the entire reason combined D3 and K2 products exist. What does not exist is a large randomised trial showing that adding K2 to vitamin D improves a clinical outcome in people with normal kidney function.
For a British reader there is a practical layer. The NHS advises everyone to consider a daily supplement containing 10 µg of vitamin D during the autumn and winter, because UK sunlight is too weak between October and March to make enough. That advice is for vitamin D on its own. If you buy a combined product, you are buying the recommended vitamin D alongside an optional extra, and the K2 in it is the part that needs checking against your prescriptions.
The downside: warfarin, and who should not take K2
Warfarin works by blocking vitamin K. That is its entire mechanism. Taking a vitamin K supplement while on warfarin therefore works directly against the drug, and MK-7 is the form most likely to do so because it stays in the blood longest.
Schurgers and colleagues put a number on it in Blood in 2006, and it is the most useful sentence in this whole field: preparations supplying 50 µg a day or more of MK-7 may interfere with oral anticoagulant treatment in a clinically relevant way.
of MK-7 per day is the level above which the authors warn of clinically relevant interference with oral anticoagulants. Common UK supplements contain 90 to 180 µg.
Schurgers L.J. et al., Blood, 2006, DOI: 10.1182/blood-2006-08-040709
Set that against a British supplement shelf and the problem is obvious. A typical UK vitamin K2 capsule contains 90 µg or 100 µg. The dose used in the Knapen trials was 180 µg. Every one of those is above the threshold in that warning, by a factor of two to four.
If you take warfarin, the rule is simple and it is not negotiable: do not start a vitamin K2 supplement without speaking to your anticoagulation clinic or your GP. The issue is not that vitamin K is forbidden, it is that your warfarin dose is calibrated to your usual vitamin K intake, and changing that intake changes your INR. Consistency matters more than avoidance, and only your clinic can manage the adjustment.
The direct oral anticoagulants that have replaced warfarin for many patients, such as apixaban, rivaroxaban and edoxaban, do not work through vitamin K and are not affected in the same way. If you are unsure which category your medicine falls into, your pharmacist can tell you in a minute.
Beyond anticoagulation, vitamin K2 is well tolerated and no consistent adverse effect has been reported at supplement doses. The honest caution is that there is not enough evidence about high doses taken for years, which is exactly what the NHS says. That gap is the rule rather than the exception: very few of the compounds attached to the biology of ageing have a randomised trial with a long-term outcome behind them.
One boundary worth drawing. Vitamin K2 acts on calcium handling in bone and vessel wall, and it does nothing for the inflammatory side of joint ageing, which is a separate process with separate targets. That terrain belongs to a well-absorbed anti-inflammatory formula and to support built for ageing joints rather than to a vitamin. Neither treats osteoporosis, and neither replaces a DXA scan and a conversation with your GP if you have been told your bone density is low.
Frequently asked questions
What is vitamin K2 taken for?
It is taken to activate two calcium-handling proteins: osteocalcin, which binds calcium into bone, and matrix Gla protein, which inhibits calcification of the artery wall. In practice that means bone density and arterial stiffness. The biochemistry moves reliably in trials. The clinical outcomes are less consistent.
What are signs of K2 deficiency?
There is no recognised deficiency syndrome for K2 specifically. Frank vitamin K deficiency shows as easy bruising and bleeding, and it is rare in adults eating normally. What researchers measure instead is functional insufficiency: high levels of uncarboxylated osteocalcin or of dephosphorylated uncarboxylated matrix Gla protein, neither of which is a routine NHS test.
Should you take K2 with D3?
The pairing is mechanistically logical, because vitamin D increases calcium absorption and K2 influences where that calcium goes. There is no large randomised trial showing that adding K2 to vitamin D improves a clinical outcome in people with normal kidney function. The NHS advice to take 10 µg of vitamin D through autumn and winter stands on its own, with or without K2.
Which food has K2 vitamin?
Natto, fermented soybeans, by a very wide margin, and it is the only significant food source of MK-7. After that: aged hard cheeses such as Gouda and Edam, egg yolk, liver and organ meats, butter and full-fat dairy, and sauerkraut in modest amounts. Green vegetables supply K1, not K2.
Are bananas high in vitamin K2?
No. Bananas contain essentially no vitamin K2 and very little vitamin K of any kind. Fruit is not a source. If you are looking for K2 in the fresh aisle you are in the wrong aisle: it comes from fermentation and from animal tissue.
Is there a downside to taking vitamin K2?
The significant one is warfarin. Warfarin works by blocking vitamin K, and MK-7 stays in the blood longest of all the forms. Preparations supplying 50 µg a day or more of MK-7 have been reported to interfere with oral anticoagulant treatment in a clinically relevant way, and typical UK supplements contain 90 to 180 µg. Beyond that, the NHS notes there is not enough evidence about the effects of high doses taken long term.
How much vitamin K2 should I take?
The UK reference is for vitamin K as a group, not K2: approximately 1 microgram a day per kilogram of body weight, so about 70 µg for a 70 kg adult. There is no separate UK figure for K2. The trials that measured bone and arterial outcomes used 180 µg of MK-7 daily for 3 years. Check any dose against your prescriptions first.
Sources
- Schurgers L.J., Teunissen K.J.F., Hamulyák K., Knapen M.H.J., Vik H., Vermeer C. Vitamin K-containing dietary supplements: comparison of synthetic vitamin K1 and natto-derived menaquinone-7. Blood, 2006. DOI: 10.1182/blood-2006-08-040709
- Knapen M.H.J., Drummen N.E., Smit E., Vermeer C., Theuwissen E. Three-year low-dose menaquinone-7 supplementation helps decrease bone loss in healthy postmenopausal women. Osteoporosis International, 2013. DOI: 10.1007/s00198-013-2325-6
- Knapen M.H.J., Braam L.A.J.L.M., Drummen N.E., et al. Menaquinone-7 supplementation improves arterial stiffness in healthy postmenopausal women. Thrombosis and Haemostasis, 2015. DOI: 10.1160/TH14-08-0675
- Geleijnse J.M., Vermeer C., Grobbee D.E., et al. Dietary intake of menaquinone is associated with a reduced risk of coronary heart disease: the Rotterdam Study. The Journal of Nutrition, 2004. DOI: 10.1093/jn/134.11.3100
- Mott A., Bradley T., Wright K., et al. Effect of vitamin K on bone mineral density and fractures in adults: an updated systematic review and meta-analysis of randomised controlled trials. Osteoporosis International, 2019. DOI: 10.1007/s00198-019-04949-0
- Vlasschaert C., Goss C.J., Pilkey N.G., McKeown S., Holden R.M. Vitamin K supplementation for the prevention of cardiovascular disease: where is the evidence? A systematic review of controlled trials. Nutrients, 2020. DOI: 10.3390/nu12102909
This article is general information, not medical advice. If you take warfarin or any other anticoagulant, have been diagnosed with osteoporosis, or have kidney disease, speak to your GP, your anticoagulation clinic or your pharmacist before taking vitamin K2.
