Two 2026 vitamin K2 artery calcification trials found that high-dose K2 slowed calcium buildup in the heart’s arteries. In the larger trial of 398 people, K2 plus vitamin D3 made the calcium score rise about 21% more slowly over two years (Trial). A second trial of 180 people found a similar slowing (Trial). But a slower rise on a scan does not prove that anyone had fewer heart attacks.
What a Calcium Score Measures
A coronary artery calcium (CAC) score comes from a quick CT scan of the chest. There is no dye and no needle. The scan finds hard, calcified plaque in the arteries that feed the heart and adds it up into one number.
A score of 0 means the scan found no calcium. Higher scores mean more calcified plaque. The main trial below counted a score of 400 or more as severe (Trial).
Doctors use the score because it predicts heart trouble. A US study followed 6,814 adults with no heart disease for about 11 years (Study). With a score of 0, about 1 to 6 in every 100 had a heart attack, stroke or related event within 10 years. With a score above 300, it was about 13 to 26 in every 100. The exact figure depended on age, sex and ethnic group. The link held even after researchers allowed for smoking, cholesterol, blood pressure, diabetes and other usual risk factors.
How fast the score climbs may matter too. In the same US group, people whose scores grew faster over a few years had more heart events later (Study). This held even among people with similar starting scores. It is one reason researchers track how fast the score grows. Still, the authors said they need more data before repeat scans can guide treatment.
A calcium score is a stand-in: a number that tracks the risk of heart attacks, rather than the heart attacks themselves. A German study of 3,281 people found that the latest score mattered most. Once you knew it, how fast the score had grown added little to predicting risk (Study).
How Vitamin K2 Guards Arteries
The case for K2 starts with a protein called matrix Gla protein (MGP). It sits in artery walls and stops calcium from settling there. One review calls it the body’s strongest natural blocker of calcium buildup (Review).
MGP only works after vitamin K helps switch it on. Without enough vitamin K, it stays in an inactive form called dp-ucMGP (Review). That form shows up in the blood, so it can flag low vitamin K.
Mice show how much MGP matters. In 1997, scientists bred mice that lacked it. The pups were born alive, but calcium soon built up in their arteries on its own. They died within two months when blood vessels burst (Study).
Vitamin K comes in two main forms. Vitamin K1 is found mostly in green leafy vegetables. Vitamin K2 shows up in modest amounts in fermented and animal foods. Natto, a dish of fermented soybeans, is the richest source, and cheese has some too (Fact sheet). The heart trials below used a K2 form called MK-7. One study found it stays in the blood longer than K1 (Fact sheet).
In people, extra K2 lowers the inactive form. In a 12-week trial of 60 adults, 180 µg of MK-7 a day cut dp-ucMGP by 31%. A dose of 360 µg cut it by 46% (Trial). People on a dummy pill saw no change. The trial measured only this blood marker, not artery calcium.
Diet studies point the same way. The Rotterdam study followed 4,807 Dutch adults who had no heart disease at the start. People who ate more K2 had less calcium in the aorta, the body’s main artery. They were also less likely to die of heart disease or of any cause. K1 showed no such link (Review). This kind of study can only show a link. People who eat more cheese or natto may differ in other ways, so it cannot prove that K2 did anything.
Years of Flat Results
The theory makes sense, but for more than a decade the big trials mostly came back flat.
An early big test used K1. For three years, 388 healthy adults aged 60 to 80 took a daily multivitamin with or without 500 µg of K1 (Trial). Overall, calcium scores rose about the same in both groups. Only people who took most of their pills, or who started with some calcium, showed a small hint of slowing (Trial). The authors said new studies would need to test that hint.
Next came AVADEC, a Danish trial in 365 men with an average age of 71 (Trial). All of them already had calcium in the aortic valve, where blood leaves the heart. For two years they took either a dummy pill (placebo) or 720 µg of K2 (as MK-7) plus 25 µg of vitamin D a day. Calcium in the valve rose 275 units on the vitamins and 292 on placebo. That gap was small enough that chance could explain it. Calcium buildup in the heart’s own arteries did not differ between the groups either.
The vitamins did change the blood, though. The inactive MGP marker fell sharply on K2 and rose slightly on placebo (Trial). So the pill switched on the protein, but calcium kept building up about equally in both groups.
A second look at AVADEC turned up one clue. Among 304 men with no known heart disease, the overall gap in coronary calcium was small enough that chance could explain it (Trial). But men who started with a score of 400 or more seemed to build calcium more slowly on K2 plus vitamin D. The authors flagged it only as an idea for future trials to test.
A 2023 review pooled 14 trials with 1,533 people (Meta-analysis). It found that vitamin K may slow coronary calcium a little. But that result rested on just four trials and 424 people. It also barely cleared the bar for chance. Most other artery and valve measures showed no effect.
The Vitamin K2 Artery Calcification Trials
The same Danish team then tested its own clue. DANCODE enrolled 400 people at several Danish hospitals (Trial). All had severe calcium in their heart arteries, meaning a score of 400 or more. Their typical age was 71, and 30% were women. The typical starting score was 903, more than twice the cutoff for severe.
A coin flip decided who took the vitamins and who took a dummy pill. The daily dose was 720 µg of K2 (as MK-7) plus 25 µg (1,000 IU) of vitamin D3. Neither the patients nor the doctors knew who got what. After two years, the team had results for 398 people.
On average, calcium scores rose by 196 units on the vitamins and by 248 on placebo (Trial). So scores grew about a fifth more slowly. The effect looked much the same in men and women, and in people with higher or lower starting scores. Scores still went up in both groups. The vitamins slowed the climb but did not reverse it.
A more detailed scan showed a limit. It was done in 143 people whose scores were under 1,000. In them, total plaque grew by about the same amount in both groups (Trial). Soft plaque, the kind with no calcium in it, did not differ either. Only the calcified part grew more slowly. So the vitamins seemed to change how much calcium the plaque held, not how much plaque there was.
The second trial, VitaK-CAC, came from the Netherlands. It enrolled 180 people with heart symptoms and scores of 50 to 400. They took 360 µg of MK-7 a day or a dummy pill (Trial). Over two years, the typical score rose from 135 to 184 on MK-7. On placebo it rose from 145 to 214. That is a rise of about 49 units against 69, which the authors called a modest slowing. This trial gave MK-7 alone, with no added vitamin D. That hints that K2 may do at least part of the work.
Both trials have weak spots. VitaK-CAC gathered its data from 2012 to 2022, with a few breaks along the way (Trial). About one in six people dropped out. Roughly 150 finished the full two years (Report). One senior author reported grants from Gnosis by Lesaffre for work outside this study. The same company gave the trial its MK-7 tablets for free (Trial). The paper says the funders had no role in the design, analysis or write-up.
DANCODE was paid for by Kappa, a company that makes K2. Another firm made the tablets (Report). The university says neither company shaped the study or how the results were read. The K2 in the tablets is a branded ingredient owned by Balchem. Balchem has talked up the result in the supplement trade press (Report). The DANCODE researchers themselves say they cannot yet recommend the vitamins. They add that nobody knows yet whether the vitamins prevent blood clots, including in the heart, or help people live longer (Report).
That leaves the evidence split. AVADEC used the same dose as DANCODE and found no overall slowing, while the two newer trials found a modest one. One way to square them is that DANCODE picked only people with severe calcium. That is the group where AVADEC saw its hint. Still, two fairly small positive trials after years of flat ones leave the picture mixed.
Less Calcium Is Not Always Safer
Statins are among the best-tested heart drugs. They cut heart events even in people over 70, as our piece on statins over 70 covers. Yet statins tend to push coronary calcium up.
A later analysis of 8 trials that scanned plaque from inside the arteries showed the pattern (Study). High-dose statins shrank plaque slightly. Calcium in plaque rose in every group, but it rose most on statins. The authors concluded that statins add calcium to plaque, separately from shrinking it.
A study of 1,255 people with repeat heart scans found a similar pattern (Study). On statins, total plaque grew more slowly. Soft plaque grew about half as fast, and fewer high-risk plaque features appeared. But calcified plaque grew faster. The authors read this as plaque hardening into a less risky form. This study only watched people, so it shows a link rather than proof.
The scoring method has a quirk. It gives extra points for denser calcium (Study). Yet in a US study of 3,398 adults, denser calcium went with fewer heart attacks and strokes. That held when the amount of calcium was the same. So a lower score does not automatically mean a safer artery.
The editorial published beside VitaK-CAC asks in its title whether calcium growth is a useful goal for a trial at all (Commentary). It called the slowing small, though unlikely to be down to chance alone. Its lead author, Dr. Blaha of Johns Hopkins, told a cardiology news site that the score stands in for total plaque. He and his co-author argue that little is known about what a change in calcium means for heart events. Recommending vitamin K to patients now, he said, might be “a bridge too far” (Report).
None of the trials here was built to test whether K2 prevents heart attacks, strokes or deaths. In AVADEC, 10 people on the vitamins and 10 on placebo had a heart or blood-vessel event (Trial). One AVADEC sub-study did see fewer heart attacks, artery procedures and deaths on K2 plus vitamin D: 3 men against 10 (Trial). But only 13 men had an event, so the authors said it could be coincidence. VitaK-CAC measured calcium on scans. Its authors call for a larger trial that counts heart events (Trial).
Doses, Food and Warfarin
The doses that slowed calcium were high. VitaK-CAC used 360 µg a day, and DANCODE used 720 µg. The official target for adults, called the adequate intake, is 120 µg a day for men and 90 µg for women (Fact sheet). On average, US women already get about 122 µg a day from food, and men about 138 µg. So the trial doses were roughly 3 to 8 times the daily target. There is no official upper limit, because vitamin K has a low risk of being toxic.
Supplement doses vary widely. In 2013, one study described 45 µg as the usual MK-7 dose sold in stores (Study). Some vitamin K supplements contain as much as 4,050 µg (Fact sheet).
The biggest safety issue involves blood thinners. Vitamin K works against warfarin (Coumadin). A sudden change in how much you get can make the drug stronger or weaker (Fact sheet). People on warfarin are told to keep their vitamin K intake steady and to talk with their care team.
Even small K2 doses can throw off a blood thinner. Fifteen healthy volunteers took a blood thinner similar to warfarin. At 45 µg of MK-7 a day, a standard clotting test showed the drug’s effect falling by about 40% (Study). Two blood specialists judged that even 10 µg a day upset clotting control in at least 4 in 10 of them. The authors advised people on these drugs to avoid MK-7.
The link also runs the other way. Warfarin blocks vitamin K, so it may switch off the proteins that guard against calcium. A 2026 review pooled three small trials of 272 people with an irregular heartbeat called atrial fibrillation. Those on newer blood thinners built calcified plaque more slowly than those on warfarin (Meta-analysis). Total plaque did not differ. This evidence is thin and not final. If you take warfarin, raise it with your doctor. Do not switch drugs on your own.
Keep in mind who took part, too. DANCODE enrolled older adults with heavy artery calcium, at a typical age of 71. VitaK-CAC’s volunteers were around 60 and had heart symptoms (Report). None of these trials tested healthy young people with a score of 0.
Frequently Asked Questions
What does the vitamin K2 artery calcification evidence show?
Two 2026 trials found that K2 slowed the rise in coronary calcium scores over two years (Trial). Earlier trials, including one at the same dose, found no overall effect. So the evidence is mixed. None of these trials was built to test whether K2 prevents heart attacks or deaths.
Can vitamin K2 reverse artery calcification?
The trials so far do not show that. In DANCODE, scores still rose in both groups. They just rose about a fifth more slowly on K2 plus vitamin D3 (Trial). VitaK-CAC saw the same pattern: scores rose more slowly, but they still rose.
Is it safe to take vitamin K2 with warfarin or statins?
Not with warfarin unless your clinician is managing it, because vitamin K can weaken the drug (Fact sheet). About 3 in 4 VitaK-CAC volunteers took statins (Report). The trial saw no major side effects from MK-7 (Trial). That is not a formal safety test, though. Either way, never drop a statin for K2.
How much vitamin K2 did the trials use?
VitaK-CAC used 360 µg of MK-7 a day. DANCODE used 720 µg of K2 with 25 µg of vitamin D3 (Trial). That is several times the 90 to 120 µg adults are advised to get each day (Fact sheet).
Key Takeaways
- Two trials slowed the score: In 2026, DANCODE and VitaK-CAC both found that coronary calcium rose more slowly over two years on K2. DANCODE gave it with vitamin D3 (Trial).
- Earlier trials were flat: A K1 trial found no overall slowing, and AVADEC found none at DANCODE’s 720 µg dose (Trial).
- A score is a stand-in: The score predicts risk. But these trials were not built to show fewer heart attacks, strokes or deaths (Report).
- Less calcium is not always safer: Statins cut heart events even as they add calcium to plaque. So a lower score is not automatically better (Study).
- The doses were high: The trials used 360 to 720 µg a day in people who already had artery calcium. The daily target is 90 to 120 µg (Fact sheet).
- Warfarin users must check first: Vitamin K weakens warfarin and similar drugs, so talk to a clinician before taking K2 (Study).
Ask Before You Buy K2
After years of flat results, two trials finally showed K2 slowing calcium on a scan. The next step is a trial that counts heart attacks and deaths. Until one reports, the benefit to your heart is unproven.
If you worry about your arteries, ask your doctor whether a calcium score makes sense for you. The proven steps still come first: a statin when one is advised, blood pressure control and not smoking. If you want to try K2 anyway, raise it at that visit, especially if you take any blood thinner.
The K2 evidence is finally moving, so it is worth watching. Talk to your doctor before you buy.
This article is for educational purposes and is not medical advice. Talk to a qualified clinician before changing your health regimen.
Leave a Reply