Supplements
Vitamin K2 and Your Arteries: What the Trials Actually Found
K2 does its biochemical job in every study. Two trials found slower artery calcification, three found nothing, and the difference between them says something useful about when to bother.
Here is the idea that sold a million bottles. Calcium is good for your bones and bad in your arteries, and vitamin K2 is the traffic cop that sends it to the right place. Take K2, keep calcium out of your arteries, live longer.
The mechanism behind that story is real and well documented. What is less advertised is that researchers have now run the trials, and the results are split in a way that is genuinely interesting. K2 does its biochemical job in every single study. Whether that job shows up in your arteries seems to depend on how far gone they already are.
This is education, not medical advice.
What K2 actually does
Vitamin K is a cofactor. Your body uses it to activate a handful of proteins through a process called carboxylation. One of those proteins is matrix Gla protein, and its job is to sit in the walls of your blood vessels and stop calcium from crystallizing there.
Without enough vitamin K, matrix Gla protein stays in its inactive form. Researchers can measure that inactive form in your blood, and it goes by the mouthful dp-ucMGP. High levels mean a lot of the protein is sitting there switched off.
Vitamin K comes in two families. K1, phylloquinone, is the one in leafy greens, and it mostly gets used by your liver for clotting. K2, the menaquinones, is found in fermented foods and animal products, hangs around longer in circulation, and is the form that reaches tissues outside the liver. MK-7 is the long-acting menaquinone used in nearly every supplement and nearly every trial.
The observation that started it
The Rotterdam Study followed 4,807 adults with no history of heart attack for about a decade, tracking what they ate. Men and women in the top third for dietary menaquinone intake had a relative risk of coronary heart disease death of 0.43 compared with the bottom third, along with lower all-cause mortality and an odds ratio of 0.48 for severe aortic calcification. Phylloquinone, the leafy green form, was not related to any of it (Geleijnse, 2004).
That is a striking result, and it is also observational. People eating the most fermented food and aged cheese differ from people eating the least in a hundred ways a statistician cannot fully scrub out. It is a reason to run a trial, not a reason to declare victory.
The part that always works
Before the disagreements, note what every trial agrees on. Supplemental MK-7 reliably switches on matrix Gla protein.
Three years of 180 micrograms a day cut inactive dp-ucMGP by 50 percent against placebo (Knapen, 2015). Two years of 720 micrograms dropped it by 212 pmol/L while placebo drifted up (Diederichsen, 2022). Six months of 360 micrograms cut it by 206 pmol/L (Zwakenberg, 2019).
So the supplement is absorbed, it reaches the tissue, and it does the chemistry it is supposed to do. That is more than most things in the supplement aisle can claim. The open question was never whether K2 works on the protein. It is whether working on the protein changes the artery.
Where it moved the artery
Two trials say yes.
The first randomized 244 healthy postmenopausal women to 180 micrograms of MK-7 or placebo for three years and measured arterial stiffness directly. Carotid-femoral pulse wave velocity and the stiffness index both dropped significantly in the K2 group, with the clearest improvements in the women whose arteries were stiffest at the start (Knapen, 2015). Worth labeling plainly: this was a trial in women, and stiffness is a different measure than calcium.
The second is the newest and the most on point. The VitaK-CAC trial randomized 180 patients with symptomatic coronary artery disease and a coronary calcium score between 50 and 400 to 360 micrograms of MK-7 or placebo for two years, with CT scans at baseline, one year, and two years. Placebo calcium scores climbed from a median of 145 to 214 Agatston units. The MK-7 group went from 135 to 184. The difference between groups was significant at p equals 0.02, and it held after adjustment (Vossen, 2026).
Two honest notes on that one. It is a single trial of 180 people, and one of the authors disclosed grants from an MK-7 manufacturer, which the paper states openly. Disclosure is not disqualification, but it belongs in your read.
Where it did not
Three trials say no, or close to it.
The largest is AVADEC. Researchers randomized 365 men, average age 71, with an aortic valve calcification score above 300, to 720 micrograms of MK-7 plus vitamin D or placebo for two years. The difference in valve calcification progression was 17 Agatston units, p equals 0.64. Nothing. Valve area, jet velocity, aortic and coronary calcification all came out the same. And this was the highest dose in the set, with dp-ucMGP confirming everyone took their tablets (Diederichsen, 2022).
A companion analysis looked at coronary calcium in 304 of those men who had no ischemic heart disease. Progression was 203 units on treatment against 254 on placebo, which did not reach significance at p equals 0.089. In the subgroup who started above 400 units, the gap did reach significance, and safety events were fewer on supplementation, 1.9 percent against 6.7 percent. The authors called both findings hypothesis-generating rather than conclusions, and that is the right label (Hasific, 2023).
A third trial gave 360 micrograms to 68 people with type 2 diabetes and known cardiovascular disease for six months. No benefit on CT calcification. If anything the PET signal for active calcification drifted the wrong way, though not significantly, and the authors were candid that nobody is sure how to read that signal yet (Zwakenberg, 2019).
Add it all up
A systematic review pooled 14 randomized trials and 1,533 patients. Vitamin K supplementation slowed coronary calcium progression by a mean difference of 17.4 Agatston units, which cleared significance at p equals 0.04. It cut inactive matrix Gla protein sharply, as expected. Adverse events were no different from placebo. The authors' own summary was that vitamin K may have therapeutic potential and that better designed trials are needed to confirm it (Li, 2023).
That is a fair picture of the whole field. A small, real, statistically detectable effect on artery calcium, sitting on top of a mechanism that unquestionably works, with a safety record that looks clean.
The pattern worth noticing
Line the trials up by how advanced the calcification was and something suggestive appears.
The trial in healthy women with no established disease found improvement. The trial in patients with moderate coronary calcium, 50 to 400 units, found slower progression. The trial in men whose aortic valves were already past 300 units found nothing at all, at double the dose.
That fits how calcification behaves. Matrix Gla protein prevents crystals from forming. It does not dissolve the ones already there, which is why no human trial has shown vitamin K2 reversing arterial calcification. A supplement that keeps mineral out of a vessel wall has more to work with in a 50-year-old with early plaque than in a 71-year-old with a valve that has been calcifying for two decades.
This is a pattern across trials, not something any single trial set out to test. Treat it as a reasonable hypothesis, not a finding.
What nobody has shown
Every result above is an image or a blood marker. A calcium score is a strong predictor of heart attacks, right up there with ApoB and Lp(a) on a blood panel, but no trial has randomized people to K2 and counted heart attacks. Nothing here says K2 prevents anything, and no one should sell it to you as if it does.
The practical read
Food first, and it is easy to miss. Menaquinones show up in fermented foods and animal products: natto, aged and hard cheeses, egg yolks, and organ meats. That was the intake pattern the Rotterdam data tracked. If your diet has none of it, that gap is worth closing before you buy anything.
The trial doses were 180 to 720 micrograms of MK-7 a day. More was not better in this set. The 720 microgram trial found nothing, the 180 and 360 microgram trials found something, and the difference was almost certainly the population rather than the dose.
Give it time. The positive trials ran two and three years. The six month trial found nothing. This is not a supplement where you check a number in a month.
Safety looked clean. Across 14 trials, adverse events matched placebo (Li, 2023). That is a genuine point in its favor.
One hard stop. If you take warfarin or any other vitamin K antagonist, vitamin K directly opposes how the drug works. Do not start K2 on your own. Talk to your prescriber first, and that applies to any blood thinner or heart medication you are on.
The verdict
Vitamin K2 has a real mechanism, an unusual amount of evidence that it engages that mechanism in humans, and a small measurable effect on artery calcium when you pool the trials. That puts it well ahead of most of the shelf.
It also failed its biggest and best-funded test in men with advanced valve calcification, and it has never been tested against the outcome you actually care about. The honest summary is a promising supplement with an open question attached, and the open question is not going to be settled by anyone selling it. That skepticism runs through all our supplement reviews.
If you eat none of the foods that contain it, fix that first, the same food-first rule that applies to vitamin D. If you want to supplement, the trial-backed range is 180 to 360 micrograms of MK-7 a day, taken for years rather than weeks, and cleared with your doctor if you are on any blood thinner.
That is the verdict.
This article is for general education and is not medical advice. Nothing here is intended to diagnose, treat, cure, or prevent any disease. Talk to your doctor before starting any supplement, especially if you take medication.
References
- Geleijnse JM, Vermeer C, Grobbee DE, et al. Dietary intake of menaquinone is associated with a reduced risk of coronary heart disease: the Rotterdam Study. J Nutr. 2004. https://pubmed.ncbi.nlm.nih.gov/15514282/
- Knapen MH, Braam LA, Drummen NE, Bekers O, Hoeks AP, Vermeer C. Menaquinone-7 supplementation improves arterial stiffness in healthy postmenopausal women. A double-blind randomised clinical trial. Thromb Haemost. 2015. https://pubmed.ncbi.nlm.nih.gov/25694037/
- Vossen LM, de Leeuw PW, Schurgers LJ, et al. Two Years of Menaquinone-7 Supplementation and Coronary Artery Calcification: A Randomized Clinical Trial. JAMA Cardiol. 2026. https://pubmed.ncbi.nlm.nih.gov/42268593/
- Diederichsen ACP, Lindholt JS, Moller S, et al. Vitamin K2 and D in Patients With Aortic Valve Calcification: A Randomized Double-Blinded Clinical Trial. Circulation. 2022. https://pubmed.ncbi.nlm.nih.gov/35465686/
- Hasific S, Oevrehus KA, Lindholt JS, et al. Effects of Vitamin K2 and D Supplementation on Coronary Artery Disease in Men: A RCT. JACC Adv. 2023. https://pubmed.ncbi.nlm.nih.gov/38938724/
- Zwakenberg SR, de Jong PA, Bartstra JW, et al. The effect of menaquinone-7 supplementation on vascular calcification in patients with diabetes: a randomized, double-blind, placebo-controlled trial. Am J Clin Nutr. 2019. https://pubmed.ncbi.nlm.nih.gov/31387121/
- Li T, Wang Y, Tu WP. Vitamin K supplementation and vascular calcification: a systematic review and meta-analysis of randomized controlled trials. Front Nutr. 2023. https://pubmed.ncbi.nlm.nih.gov/37252246/
Common questions
Does vitamin K2 really keep calcium out of your arteries?
Vitamin K2 activates matrix Gla protein, which is the protein that stops calcium from crystallizing in blood vessel walls, and every trial that measured it confirmed the supplement does that job in humans. Whether that translates into less artery calcification is the split part. A trial in 180 patients with moderate coronary calcium found significantly slower progression over two years (Vossen, 2026), while a larger trial in 365 older men with advanced valve calcification found no difference at all (Diederichsen, 2022).
How much MK-7 did the trials use?
Between 180 and 720 micrograms a day. The two trials that found benefit used 180 and 360 micrograms, and the trial that used 720 micrograms found nothing, so more was not better here. The likely explanation is the population rather than the dose: the higher dose was tested in men whose valves were already heavily calcified.
How long do you have to take it?
The trials that found something ran two and three years. The six month trial found no effect on CT calcification (Zwakenberg, 2019). This is a slow-acting supplement measured in years, not a supplement where you check a number after a month.
Is vitamin K2 safe to take?
In a pooled analysis of 14 randomized trials and 1,533 patients, adverse events were no different from placebo (Li, 2023). The one hard exception is warfarin and other vitamin K antagonists, where vitamin K directly opposes how the drug works. If you take any blood thinner, talk to your prescriber before starting K2.
Can I get K2 from food?
You can get K2 from food. Menaquinones are in fermented foods and animal products, mainly natto, aged and hard cheeses, egg yolks, and organ meats. That dietary pattern is what the Rotterdam Study tracked when it found lower coronary heart disease mortality in the highest intake group (Geleijnse, 2004). If you eat none of those, closing that gap is the first move.
