Vitamin K: What It Actually Does, and the Honest Read on Those K2 Supplements
Vitamin K: What It Actually Does, and the Honest Read on Those K2 Supplements
A man in his late sixties starts a blood thinner after an irregular heartbeat. Somewhere along the way, maybe from a relative, maybe from something he read, he picks up the idea that his medicine and leafy greens do not mix. So he quietly cuts salad, spinach, and broccoli out of his life. His next blood test comes back off. His dose gets chased around for weeks, and nobody can figure out why his numbers keep swinging.
Here is the quiet part: it was never the greens. It was the change. That small story holds almost everything worth knowing about vitamin K, a vitamin most people have opinions about mainly because they have seen an ad for it. If you have typed "what does vitamin K do" or "does vitamin K2 really help your bones" into a search bar, this is the plain version of the answer. The mechanism is genuinely interesting. The settled part is handled by dinner. And the most heavily marketed part is exactly where the evidence gets thin.
What does vitamin K do
Vitamin K does one chemical trick, and that single trick explains both of its very different jobs.
Think of vitamin K as a key that switches proteins on. Certain proteins in your body are built but arrive switched off. Vitamin K flips them to the active state by attaching a small chemical tag to specific spots on them. Scientists call this carboxylation, and the proteins it activates are called Gla proteins. Without enough vitamin K, those proteins still exist, they just cannot do their job. Two families of them matter, and they are why vitamin K has two completely different reputations.
Job one is blood clotting. Several of the proteins that let your blood clot are Gla proteins, and vitamin K switches them on. This is the classic role: without vitamin K, blood does not clot properly, which shows up as easy bruising and prolonged bleeding. This job happens mainly in the liver, and it has been settled science for the better part of a century.
Job two is directing where calcium goes. The same vitamin K key also switches on two proteins outside the liver. One is osteocalcin, a protein in bone. The other is matrix Gla protein, which sits in blood vessel walls. Switched-on osteocalcin helps bind calcium into the bone structure. Switched-on matrix Gla protein helps keep calcium from settling into artery walls where it does not belong. A 2017 review in the Journal of Nutrition and Metabolism lays this out plainly: carboxylated osteocalcin rises after vitamin K2, uncarboxylated osteocalcin tracks with fracture risk, and matrix Gla protein, once activated, holds back calcification of blood vessels and soft tissue (Schwalfenberg, PMID 28698808). That mechanism is the engine behind the whole "K2 for bones and arteries" wave of the last two decades. The mechanism is real. Hold that thought, because the honest section below is about the gap between this real mechanism and what the supplements actually promise.
One correction while we are here, because it is everywhere online: vitamin K2 does not help you absorb calcium or vitamin D. It does nothing at the gut. It works downstream, on where the calcium you already absorbed ends up. Vitamin D and vitamin K get described as a team, and in one narrow sense they are, D helps you absorb calcium and K helps direct where that calcium goes, but they are teammates on the destination, not on the uptake.
Vitamin K1 vs K2, what is the difference
There are two main forms, and the difference is genuinely useful to understand.
Vitamin K1 (phylloquinone) is the form in leafy greens and brassicas. Your liver takes it up preferentially for the clotting job. It has a short life in the blood, on the order of hours, and it does not reliably travel out to bone and blood vessels.
Vitamin K2 (menaquinones) comes mostly from fermented foods and animal products, plus a limited amount your body converts from K1. The best-studied form, called MK-7, stays in circulation much longer than K1, on the order of days, and reaches those proteins outside the liver more effectively. In one careful study, MK-7 accumulated in the blood to levels roughly 7- to 8-fold higher during daily intake, which is why it is the form supplement makers reach for (Schurgers 2007, PMID 17158229).
There is also a class rule that applies to both forms: vitamin K is fat-soluble, so it needs a little dietary fat in the same meal to absorb well. A naked salad with no oil is one of the least efficient ways to get the K1 out of your greens. A little olive oil on cooked greens changes that. Vitamin K1 is also fairly heat-stable, so normal cooking does not meaningfully cost you the K in your greens.
Does vitamin K2 really help your bones and heart
This is the honest core of the article, because vitamin K2 is the most marketed corner of the whole vitamin world, and the evidence does not match the volume of the marketing.
Start with what is settled: the mechanism. Vitamin K genuinely switches on osteocalcin in bone and matrix Gla protein in your vessels. That is not in dispute. The question is whether taking a K2 supplement measurably changes what happens to real people's bones and arteries. Here the picture gets honest.
On bone, the dramatic numbers you may have seen come from an older Japanese body of research using a pharmaceutical dose. One two-year randomized trial gave osteoporotic patients 45 milligrams a day of a K2 form called MK-4. Lumbar bone density fell by 3.3 percent in the control group but only 0.5 percent in the treated group, and there were fewer fractures in the treated group (Shiraki 2000, PMID 10750566). That sounds impressive, but note the dose: 45 milligrams a day is roughly 250 times the amount in a typical K2 supplement, and it is used as a prescription drug in Japan. A 2006 review of the fracture evidence found that every single trial reporting a fracture benefit was Japanese and used this menaquinone (Cockayne 2006, PMID 16801507). When you move to the modern Western supplement dose, the effect shrinks. A well-run three-year trial gave 244 healthy postmenopausal women 180 micrograms of MK-7 a day. It slowed the age-related decline in bone density at the spine and femoral neck, but not at the total hip (Knapen 2013, PMID 23525894). Real, but modest, measured on bone density rather than actual fractures, and studied almost entirely in postmenopausal women rather than the general population.
On the heart, the honest read is even clearer. The exciting signal is observational. A large Dutch population study found that people eating more dietary K2, though not K1, had less coronary heart disease and less severe aortic calcification over time (Geleijnse 2004, Rotterdam Study). That is an association from following people who already ate a certain way, which cannot prove the vitamin caused the difference. When researchers put it to the test with an actual trial, the result cooled off. A two-year randomized trial gave 365 men a high dose of MK-7 plus vitamin D and measured calcification in the aortic valve. There was no meaningful difference from placebo (Diederichsen 2022, Circulation, PMID 35465686). There is currently no trial showing that a K2 supplement reduces heart attacks, strokes, or deaths.
So here is the calm summary. The mechanism is real and plausible. The supplement pitch works precisely because it describes that real mechanism and then quietly lets you assume the mechanism has been proven to protect bones and arteries in practice. For bone, there is modest supportive evidence at supplement doses, mostly in postmenopausal women. For arteries, the strong data are observational, and the trials at hard endpoints have not delivered. If you bought a K2 supplement on the strength of an ad, you were not foolish, the mechanism on the label is true. It is just that the leap from "this protein does this job" to "this pill measurably protects your heart" is exactly where the marketing stops being careful. And the part of vitamin K that is completely settled, the part your dinner plate already covers, is the part the ad never mentioned.
Vitamin K and blood thinners (warfarin)
Back to the man who cut out his greens. His blood thinner was warfarin, the older medicine, and warfarin works by blocking vitamin K on purpose. That is literally its mechanism. So a person on warfarin has their dose balanced against a steady background level of vitamin K from food.
The point that gets mangled constantly: the goal is consistency, not avoidance. Sudden swings in green-vegetable intake are what move the blood test around, not greens themselves. A big kale week followed by a no-greens week is the problem. Steady greens are not. Two randomized trials back this up. In one, a small steady daily dose of vitamin K did not change overall time in the target range but did cut the number of extreme swings in the blood test (Boonyawat 2016, PMID 27346552). In another, coaching unstable patients to raise and steady their vitamin K intake left more of them in good control, 50 percent versus 19 percent (Ferland 2019). The move a good anticoagulation clinic actually wants is the opposite of what the man did: keep your greens steady, and coordinate with your prescriber, never cut them out. For the record, the newer blood thinners, the ones like apixaban and rivaroxaban, do not work through vitamin K at all, so dietary vitamin K is not a concern with them.
If you are on warfarin, this is a conversation with your own prescriber, not a rule to self-apply from a blog. But the direction is steady, not less.
A calm word on the newborn vitamin K shot
You may know that newborns get a vitamin K shot at birth, and it is worth one plain paragraph. Babies are born low on vitamin K. It crosses the placenta poorly, breast milk is low in it, and a newborn's gut bacteria are not established yet, so a baby cannot make or store much of its own. Because vitamin K runs the clotting job, that shortfall can rarely cause serious bleeding in the first weeks of life. A single vitamin K shot covers that window. This is long-standing routine care, recommended by the American Academy of Pediatrics since 1961 (CDC MMWR, 2013). It is a decades-old, protective, standard practice, not a new intervention and nothing to feel anxious about.
The gut bacteria question, and the E. coli you have heard about
One more real but modest piece. Bacteria in your gut do make vitamin K2. Different bacteria make different forms, and one of them is the famous Escherichia coli, or E. coli, which makes a K2 form called MK-8. That fact sometimes gets stretched into "your gut covers your vitamin K," and it does not. What the research actually shows is that gut bacteria synthesize a fair amount of K2, but most of it stays bound in the gut and is poorly absorbed, so diet remains your principal source (Lai 2022). Your gut bugs contribute a little. Your food does the work.
And since E. coli usually shows up in scary headlines, worth a quick reframe: the World Health Organization notes that E. coli is normally found in the gut of healthy people and that most strains are harmless. Only specific strains, like the Shiga-toxin producers behind food-poisoning outbreaks, make people sick. The E. coli quietly living in your gut is mostly a useful neighbor, and some of it even makes a little of your vitamin K.
How to get vitamin K from food
Here is the reassuring headline: K1 is one of the easiest vitamins in this whole series to get from food. The settled job of vitamin K runs on K1, and K1 is abundant in greens. The adult target is 120 micrograms a day for men and 90 for women, and a single serving of cooked greens covers that comfortably. The numbers below are here to show how easily greens handle it, not a checklist to hit or tally.
Vitamin K1, from greens. Rough per-serving amounts:
- Cooked kale, 1/2 cup: about 565 micrograms
- Cooked collard greens, 1/2 cup: about 530 micrograms
- Cooked spinach, 1/2 cup: about 445 micrograms
- Cooked Brussels sprouts, 1/2 cup: about 150 micrograms
- Cooked broccoli, 1/2 cup: about 110 micrograms
- Romaine lettuce, 2 cups: about 100 micrograms
A half-cup of cooked leafy greens clears the whole adult daily target, often many times over. Add a little olive oil so it absorbs. For almost everyone, that is all it takes.
Vitamin K2, from fermented and animal foods. This form is genuinely harder to source, and that is okay:
- Natto (fermented soybeans), 1 ounce: about 300 micrograms, by far the highest food source
- Aged hard cheese like gouda or edam, 1 ounce: about 25 micrograms
- Pasture-raised egg yolks, 2 large: about 15 to 30 micrograms
- Dark-meat chicken, 3 ounces: about 10 micrograms
- Beef liver, 3 ounces: about 10 micrograms
Natto is technically the best K2 source by a wide margin, and practically almost nobody who did not grow up with it is going to eat it. That is fine. The realistic K2 rotation for most people is a little aged cheese, some dark-meat poultry, and pasture-raised eggs. And here is the deeper reassurance: the part of vitamin K that is settled science does not hinge on hitting a K2 target, so a person who never touches natto is not failing at anything.
How we think about this at Anchor & Apex
The ANCHOR read on vitamin K is calm and specific. Vitamin K does two real jobs, clotting and directing where calcium goes, and both run on a mechanism that is genuinely settled. The settled part is easy to cover: a serving of greens with a little fat handles it, no test and no purchase required. The marketed part, the K2 supplement for bones and arteries, rests on a real mechanism whose clinical payoff is modest for bone and unproven for the heart, which is worth knowing before you spend on it.
And if you are on warfarin, the whole message flips from avoid to steady: keep your greens consistent and let your clinic guide the dose. Back to the man who cut out his salad. His eating was never the failure. The change was, and that is not something a person can feel or see. The food was fine all along. That is the spirit of this vitamin: the mechanism is interesting, the greens are enough, and the honest answer stays quiet.
About the evidence
This article is the durable, evergreen companion to The Vitamin Letters, Issue 014, and draws on Anchor & Apex's curated micronutrient science (Domain 02, Stephanie Clarke) and food-science library (Domain 15, Zoe Olsen), with fresh citation pulls by Giada. The carboxylation mechanism and the two cofactor roles are mechanism-level science (Schwalfenberg 2017, PMID 28698808). The bone evidence spans a high-dose Japanese MK-4 trial (Shiraki 2000, PMID 10750566), a fracture review of Japanese menaquinone trials (Cockayne 2006, PMID 16801507), and a modern low-dose MK-7 trial that slowed bone-density loss at some sites but not the total hip (Knapen 2013, PMID 23525894). The heart evidence contrasts an observational signal (Geleijnse 2004, Rotterdam Study) with a randomized trial that found no effect on aortic valve calcification (Diederichsen 2022, PMID 35465686). The warfarin guidance rests on two randomized trials supporting steady intake (Boonyawat 2016, PMID 27346552; Ferland 2019). The newborn practice and its history come from a CDC report (2013), and the gut-synthesis and E. coli points from a 2022 review (Lai 2022) and the World Health Organization. This is general education, not individualized medical or nutrition advice. Warfarin, supplement, and testing decisions belong with your own clinician.
Next in the series: vitamin E, staying on the fat-soluble side of the house.
This article is for educational purposes and does not constitute medical advice. If this is relevant to you, please work with your physician and a registered dietitian on an individualized approach.
Be gentle with yourself out there. You are doing better than you think. ♡
Maggie Cooper
Written by Maggie Cooper · Blog & Website Copywriter · Anchor & Apex
"Smart science, soft landing. Honest words that leave you a little wiser and a lot kinder to yourself."