In the early 1990s, a team of scientists went door to door through one suburb of Rotterdam and asked nearly five thousand people over the age of fifty-five a simple question. What do you eat?
Then they waited. For the next seven years they counted who died of heart disease.
When they sorted the answers by one nutrient, the pattern was hard to believe.
The third of people eating the most of it had 57 percent fewer deaths from coronary heart disease than the third eating the least. The middle group, 27 percent fewer. This was a vitamin most people have never heard of, found in cheese and egg yolks [1].
Those results were published in 2004. And still, no guideline mentions it. Cardiologists aren't prescribing it.
So why not?
Table of Contents
- It isn't for lack of a reason it could work
- The Danish valve trial
- Two trials built to answer the question
- Calcium is not how a blocked artery starts. It's how it ends
- We don't yet know that treating calcium saves lives
- But what about the fewer events?
- What I tell my patients
- References
It isn't for lack of a reason it could work
The wall of an artery makes a protein called matrix Gla protein. Its job is to stop calcium settling into the wall. To do that job it has to be switched on, and the switch is vitamin K2. Without enough vitamin K2 the protein sits there switched off [1].
That's great, because we have data showing an association between higher calcium build-up in our arteries and higher rates of heart attacks. In a study of nearly seven thousand people, the faster the calcium score climbed, the more heart attacks followed [2].
So if we have data showing that higher dietary vitamin K2 intake is associated with lower deaths from coronary artery disease, and we have a plausible mechanism for how vitamin K2 might be doing this, why aren't cardiologists prescribing it?
The Danish valve trial
Especially when we have this trial in Denmark, published in 2022, involving 365 men aged 65 to 74.
Every one of them had calcium in the aortic valve, and the researchers wanted to see if vitamin K2 could prevent or even reverse this calcification. Half got 720 micrograms of vitamin K2 plus vitamin D every day for two years. Half got a placebo. Why the vitamin D? Vitamin K2 switches on the matrix Gla proteins, while vitamin D has been shown to increase the amount of matrix Gla protein [3].
It failed, but something interesting was found when the data was looked at carefully [4].
During the trial, the researchers looked at the coronary arteries, and for those who already had lots of calcium in their arteries, a score above 400, vitamin K2 appeared to slow down further build-up compared to the placebo group [5].
And serious events, meaning heart attacks, stents or bypasses, and deaths from any cause: three on vitamin K2, ten on placebo [5].
The Danish Heart Foundation and other non-commercial funders paid for the trial. The tablets were supplied free by the company that makes the vitamin K2 [5].
So, we have a link between higher dietary vitamin K2 and lower deaths from coronary artery disease, a plausible mechanism for how it could do this, a randomized trial where vitamin K2 appeared to slow down further artery calcification, and the K2 group had fewer serious events than the placebo group.
So why aren't cardiologists prescribing vitamin K2?
Two trials built to answer the question
Well, ideally we want a study that looks specifically at coronary calcium, rather than relying on a study that was primarily looking at aortic valve calcification.
Which is exactly what the VitaK-CAC trial did, and the study was done in the Netherlands.
Published on the 10th of June, 2026, it took 180 adults with a coronary calcium score between 50 and 400. So moderate disease. Half had 360 micrograms of MK-7 and the other half had a placebo.
After the two-year follow-up period, the K2 group's calcium score rose less than the placebo group's, by 22 units, and that was statistically significant [6].
This finding was backed up by another randomized controlled trial published on the 28th of August, called the DANCODE trial.
It was done by the same Danish team whose valve trial had failed. They went back. This time they took only people with severe coronary artery calcification, a score above 400, the exact group where their first trial had hinted at something. They were betting that the hint was real.
398 people. Half had vitamin K2 MK-7 720 micrograms and vitamin D 1,000 IU, the other half a placebo.
After the two-year follow-up, the K2 group's calcium scores rose less than the placebo group's, by 52 units [3].
So we have two dedicated randomized trials. One showing that for moderate disease, vitamin K2 slows further calcification. The other showing that for severe disease, vitamin K2 also slows down further calcification.
So why aren't cardiologists prescribing it? Then I'll explain the recommendation that I give to my patients.
Calcium is not how a blocked artery starts. It's how it ends
It may sound obvious that if vitamin K2 reduces calcium build-up in arteries, that would reduce heart attacks.
But it's not that simple.
Calcium is not how a blocked artery starts. It's how it ends.
A plaque begins when cholesterol slips into the wall of the artery and gets stuck. The body sends immune cells to clean it up, they fail, and they die there. What's left is a soft pool of fat and dead cells with a thin skin stretched over it. That is a young plaque. It's soft, it's alive, and it's dangerous, because it can rupture and cause a heart attack.
Over years, the body tries to wall the mess off. It lays down scar tissue, and then calcium, the same way it hardens a healing bone. A calcified plaque is old, stiff and much less likely to tear. Cardiologists have known this for decades: the plaques that kill are usually the soft ones you cannot see on a calcium scan.
So a calcium scan measures the end of disease. The score predicts risk because people with a lot of scar usually also have a lot of soft, vulnerable plaque.
Look for example at what statins do to coronary plaque.
Statins add calcium. In ultrasound studies of 3,500 people's arteries, statins shrank the plaque burden and made it denser and more calcified at the same time. Soft plaque into dense, stable plaque. This may be part of how statins lower heart attack risk [7].
So if vitamin K2 slows down coronary artery calcification, does that mean that the vulnerable soft plaque isn't being stabilized? That even though there's less calcium on a scan, the actual heart attack risk is higher?
To help answer this, we can look at what vitamin K2 did to non-calcified plaque in the randomized trials.
For the VitaK-CAC trial, there was no significant difference in the number of soft plaques between the vitamin K2 and placebo groups [6].
Which matches the findings of the new DANCODE trial, where the total volume of plaque and noncalcified plaque were not significantly different [3].
So we now have a potential situation where the ratio of soft versus calcified plaque is actually worse with vitamin K2. Less calcium on the scan, the same amount of the dangerous stuff.
And I'm quoting from Dr Diederichsen, the cardiologist who ran the aortic calcification study and the DANCODE study: one scenario might be that patients are left with "more soft plaque, with more of this lipid core, with more lipids in the walls, and then the patient would be at a larger risk" [8].
We don't yet know that treating calcium saves lives
We don't know. And there's a bigger problem underneath. We are not even sure yet that finding coronary calcium and treating it saves lives.
Denmark ran the biggest experiment ever on finding calcium early and treating it with the drugs we trust most. 31,000 men. Seven years later it still can't show fewer deaths. If treating a known score with statins can't yet prove itself, changing the score with a vitamin has a long way to go [9].
The Circulation editorial printed beside the DANCODE trial put it in one sentence: "there remains no direct, prospectively validated evidence that reduced CAC progression reduces cardiovascular events" [10].
So why aren't cardiologists prescribing vitamin K2?
Because nobody has shown whether using vitamin K2 at these high doses actually reduces heart attacks or not.
But what about the fewer events?
Ah, but what about the lower event rate in the aortic calcification trial that we went through? Fewer heart attacks, stents, and all-cause deaths?
Here's what the trial authors say: it must be emphasized that this finding was a safety endpoint. Thus, the study was not powered for this outcome, leaving this a possible coincidental finding [5].
Thirteen events in total. And in the DANCODE trial, five. Two on vitamin K2, three on placebo. You cannot learn anything about heart attacks from so few events.
Now look at the vitamin K2 doses.
The doses that moved a scan were 360 to 720 micrograms a day. That is three to eight times the adequate intake of 90 to 120 micrograms, and roughly ten to twenty times what a high-K2 diet delivers [1].
And remember the Rotterdam study that started this article. The group with the lowest heart deaths were eating about 33 micrograms of vitamin K2 a day. 720 is around twenty times that. At that dose we should not think of K2 as a vitamin, but as a drug. And a drug has to prove two things before anyone prescribes it: that it is safe, and that it actually reduces heart attacks.
Are we being pedantic?
I don't think so. Medicine is littered with examples of how promising early trials and markers have failed to deliver results.
Look at drugs that raise HDL-cholesterol. The so-called "good cholesterol", which is a misnomer. The drugs raised HDL-cholesterol, but patients did worse [11].
So what do I tell a patient sitting across from me if they ask about vitamin K2?
What I tell my patients
I tell them this.
Like all vitamins and minerals, we want to reach the recommended daily intake, but be very cautious of megadosing.
For total vitamin K, the adequate intake is 90 to 120 micrograms a day. How much of that should be K1 versus K2 is up for debate, and we lack data. Leafy greens carry K1. Cheese, natto and eggs carry K2. For people who don't eat those, to fill that nutritional gap a low-dose vitamin K2 supplement could be considered.
But I would advise against taking more than, say, 90 to 120 micrograms.
Meeting the daily requirements for vitamin K is a good idea for bone strength.
The European Food Safety Authority's authorised claim: "a cause and effect relationship has been established between the dietary intake of vitamin K and the maintenance of normal bone." Dietary intake. Normal bone. An adequacy claim [1].
Vitamin K2 at 180 micrograms a day for three years slowed bone-density loss at the spine and the neck of the hip, in postmenopausal women. But that trial measured density only. Fractures, the thing that changes a life, have not been shown to fall [12].
One safety point, and it's the one I check with every patient. If you take warfarin, be extremely cautious with vitamin K2. The MK-7 form, the one in every coronary trial in this story, moved the clotting test in at least 40 percent of volunteers at just 10 micrograms a day, and some specialists advise avoiding it entirely. The newer blood thinners, apixaban, rivaroxaban and dabigatran, are not affected; the Danish trial let people on those take part [13][3].
The advice is the same. Meet your daily nutrition requirements, be very cautious about megadosing.
References
1. https://ods.od.nih.gov/factsheets/VitaminK-HealthProfessional/
2. https://doi.org/10.1016/j.jacc.2012.12.035
3. https://www.ahajournals.org/doi/10.1161/CIRCULATIONAHA.126.082363
4. https://doi.org/10.1161/CIRCULATIONAHA.121.057008
5. https://doi.org/10.1016/j.jacadv.2023.100643
6. https://jamanetwork.com/journals/jamacardiology/fullarticle/2850256
7. https://doi.org/10.1016/j.jacc.2015.01.036
8. https://www.tctmd.com/news/vitamin-k-supplementation-may-reduce-coronary-calcification-vitak-cac
9. https://doi.org/10.1093/eurheartj/ehaf704
10. https://doi.org/10.1161/CIRCULATIONAHA.126.082640
11. https://doi.org/10.1056/NEJMoa0706628
