A breakthrough, screenshotted
Last week my feed filled with a familiar shape of excitement. "New RCT in JAMA Cardiology," one post announced to tens of thousands of viewers, "180 patients with coronary artery disease, 360 micrograms of MK-7 daily for two years." The implication was unmistakable: the supplement aisle's favorite calcium-traffic-cop had finally been vindicated by a top-tier journal. Below it, the usual chorus — take D3 with K2, direct the calcium out of your arteries and into your bones, prevent the heart attack the cardiologists won't tell you about.1
Here is the part that almost never survives the screenshot. The study is real. I read it. And the single most accurate thing anyone said about it came from a skeptic with a fraction of the engagement: "A randomized trial slowing a biomarker. Not a trial preventing heart attacks. Those are different documents."2
That sentence is the whole review. Vitamin K2 has spent twenty years as the supplement world's most mechanistically satisfying story — a fat-soluble vitamin that, in theory, decides whether the calcium you absorb ends up strengthening your skeleton or hardening your aorta. The biology is genuinely elegant. The marketing has run several miles ahead of it. And the question that matters — does swallowing K2 keep you alive longer or out of the cath lab — has an uncomfortable answer that no trial, including the viral one, has ever addressed.3
The study is real. It slowed a calcium score by a hair. It did not prevent a single heart attack, because it was never built to try.
On the viral 2026 trialWhy the biology is actually good
Start with the part the influencers get right, because it is the reason this story has legs. Your body makes a protein called Matrix Gla Protein, or MGP, and it is the most potent local brake on arterial calcification we know of. MGP sits in the walls of your blood vessels and, when fully activated, physically binds calcium crystals and stops them from depositing in tissue that is supposed to stay soft.4
The catch is that MGP is born inactive. To switch it on, the body has to attach carbon dioxide groups to it in a reaction called carboxylation, and that reaction requires vitamin K as a cofactor. No vitamin K, no carboxylation, no functional brake. We know this is not a theoretical concern because the animal that cannot carboxylate MGP — the MGP-knockout mouse — dies young with arteries so calcified they resemble pipes.5
We have a second, accidental human experiment too: warfarin. The classic blood thinner works precisely by blocking vitamin K recycling, and chronic warfarin use is associated with more vascular and valve calcification, not less. The mechanism, in other words, runs in both directions and has a body count. That is a far stronger biological foundation than most supplements ever assemble.6
Two forms matter for the supplement conversation. MK-4 clears the bloodstream in hours; MK-7, the form derived from the Japanese fermented-soybean dish natto, has a half-life of around three days and reaches tissue far more reliably. That is why every serious trial uses MK-7, and why "K2 as MK-7" became the label-copy shibboleth. When researchers want to confirm the pill is actually doing something biochemical, they measure dephosphorylated-uncarboxylated MGP, or dp-ucMGP, the inactive protein left over when K2 is scarce. Supplement K2 and that number drops, every time. The pill works. The question is whether the artery cares.7
What the trials actually found
There are now more than a dozen randomized controlled trials of vitamin K and vascular calcification. They share a structural feature you should fix in your mind before reading another headline: every one of them measures a surrogate — a calcium score on a CT scan, the stiffness of an artery, the dp-ucMGP biomarker — and not one measures whether a person had a heart attack, a stroke, or died. With that frame in place, here is the actual scoreboard.8
The viral one. 180 patients with symptomatic coronary artery disease and baseline coronary calcium scores between 50 and 400, randomized to 360 mcg/day MK-7 or placebo for 24 months. Primary endpoint: change in the coronary artery calcium (Agatston) score.1
Results: Calcium progression was modestly slower on K2 (roughly 135 to 184 units, versus 145 to 214 on placebo; p=0.02). The authors describe the effect as "modest" and state plainly the trial was not powered for clinical events.
Limitation: A statistically significant slowing of a calcium score in 167 analyzed patients is a surrogate signal, not a demonstrated reduction in heart attacks. Industry provided in-kind support and a co-author has long-standing vitamin K industry ties.
The flagship positive. 244 healthy postmenopausal women, 180 mcg/day MK-7 versus placebo, for three years. Primary endpoint: arterial stiffness, measured by carotid-femoral pulse wave velocity.9
Results: Stiffness significantly decreased, with the benefit concentrated in women who started out stiffest; dp-ucMGP fell about 50%. The single most-cited "K2 helps arteries" trial.
Limitation: The corresponding author is affiliated with the lab that developed the branded MenaQ7 product used in the trial. The most industry-entangled study in the literature, and still a surrogate endpoint.
The one the influencers skip. 365 men with aortic valve calcification above 300 units, given a high 720 mcg/day dose of MK-7 plus vitamin D, or placebo, for two years.10
Results: Flatly negative. No significant difference in aortic valve calcification progression (p=0.64). Only a post-hoc subgroup hinted at anything, which is hypothesis-generating at best.
Limitation: This was a well-funded, independent trial at a higher dose than most supplements deliver, and it found nothing on its primary endpoint. Funded by Danish heart and Novo Nordisk foundations; tablets donated by a manufacturer.
Biomarker moved, artery did not. 189 patients with chronic kidney disease, 400 mcg/day MK-7 versus placebo for one year, with arterial stiffness as the primary endpoint.11
Results: Null (p=0.77) despite dp-ucMGP falling sharply. A clean demonstration that switching on the biomarker did not translate to a stiffer-artery benefit. No industry funding.
Limitation: CKD patients calcify for reasons beyond vitamin K status, so a null here does not fully generalize. But it is exactly the population where the calcification stakes are highest.
Trended the wrong way. 68 people with type 2 diabetes and cardiovascular disease, 360 mcg/day MK-7 for six months, with femoral artery calcification activity tracked by sodium-fluoride PET and CT.12
Results: No reduction in systemic arterial calcification. PET-measured activity actually trended worse on K2 (p=0.06), the opposite of the intended effect.
Limitation: Short duration and small sample; PET activity is itself a surrogate. But "trended worse" is not the direction the marketing predicts.
Add the dialysis trials — iPACK-HD, VitaVasK, Oikonomaki — and the pattern holds: the biomarker reliably improves, the calcification outcomes are mostly null or, at best, statistically fragile. Among rigorous RCTs the honest tally is roughly two surrogate wins against five-plus nulls. That is not a vindication. That is a coin flip dressed in a lab coat.13
Slowing a number is not saving a life
Cardiology has a long, painful history with surrogate endpoints, and it is worth remembering why. A surrogate is a measurement we hope stands in for an outcome we actually care about. Sometimes it does. Sometimes a drug improves the surrogate beautifully and quietly kills people anyway — the antiarrhythmic disaster of the late 1980s, where drugs that suppressed the "right" heart-rhythm marker increased mortality, is the lesson the field never fully recovered from.14
Coronary calcium score is a good predictor of risk when measured across a population. That does not automatically mean that nudging it with a pill changes anyone's fate, because a calcium score is a snapshot of a process, not the process itself. Denser, more stable plaque can even score higher while being less likely to rupture. So when a trial reports that K2 slowed calcium accumulation by a modest margin, the correct response is interest, not celebration.15
Every positive cardiovascular trial of K2 reports a surrogate endpoint. Hard-outcome evidence does not exist.1,8,16
This is why regulators have not budged. The U.S. Food and Drug Administration permits no cardiovascular claim for vitamin K2. The European Food Safety Authority went further: it formally assessed the proposed "vitamin K contributes to normal function of the heart and blood vessels" claim and rejected it twice, in 2009 and again in 2012. The only vitamin K claims EFSA authorizes concern normal bones and normal blood clotting. The cardiovascular case, in the eyes of the people whose job is to weigh exactly this evidence, remains unproven.16
Where the magic numbers come from
If the trials are a coin flip, where does the confident "57% lower heart disease" figure in every K2 thread come from? One place, mostly: the Rotterdam Study, a 2004 analysis that followed roughly 4,800 older adults and found that those eating the most dietary K2 had about 57% lower coronary heart disease mortality and less aortic calcification than those eating the least.17
It is a real finding from a respected cohort, and it is also the textbook definition of confounded. People who eat the most menaquinone are eating more cheese, fermented dairy, and natto — they differ from low-K2 eaters in income, overall diet quality, and a hundred other variables a food-frequency questionnaire cannot fully capture. The Prospect-EPIC and PREVEND cohorts add similar associations, and they carry the same asterisk. Observational nutrition data can generate a hypothesis. It cannot confirm that the nutrient, rather than the lifestyle around it, is doing the work.18
The tell is in the contrast. The headline survival numbers come from epidemiology you cannot control. The trials you can control, the ones that randomly assign the pill and remove the confounding, deliver a modest calcium signal and a row of nulls. When the uncontrolled data look spectacular and the controlled data look ordinary, the controlled data are almost always telling the truth.19
What gets left out of the thread
The warfarin trap
K2 directly antagonizes warfarin. Anyone on this common blood thinner who starts or stops K2 without INR monitoring can blunt their anticoagulation and raise clot and stroke risk. This is the one genuinely dangerous interaction, and it is almost never mentioned in the supplement pitch.
No outcome, anywhere
Twenty years and a dozen trials in, not a single randomized study has been powered to test whether K2 prevents a heart attack, stroke, or death. The 2015 Cochrane review found no hard-outcome data; a 2023 dialysis meta-analysis found no mortality benefit.
Whose lab funded it
The most positive trials lean on the most industry-entangled funding. The cleanest, independently funded trials — AVADEC, K4Kidneys — are the ones that came back null. That correlation is worth sitting with.
Low downside
For most people not on warfarin, MK-7 is well tolerated with no established toxicity and a cautious reference ceiling around 375 mcg/day. The risk of trying it is mainly to your wallet, not your body.
That last card matters for fairness. This is not a dangerous supplement for the average person, and I am not telling anyone with a cheese habit to feel foolish. The problem is not safety; it is the gap between what is claimed and what is shown. "Well tolerated and biologically plausible" is a reasonable thing to say about K2. "Clinically proven to protect your heart" is not, and the second sentence is the one selling the bottles.20
Dr. Cole's assessment
Vitamin K2 is the rare supplement with a mechanism I genuinely respect. The MGP pathway is real, the warfarin evidence makes it more than theoretical, and MK-7 reliably switches on the biomarker that tracks vitamin K activity. If I were designing a compound to inhibit vascular calcification from scratch, I would start near here.
But mechanism is a promissory note, not a payment. The randomized evidence is dominated by null results on the outcomes that matter, with only a modest calcium-score signal from the new JAMA Cardiology trial and an older, industry-tied stiffness study on the positive side. No trial has tested hard cardiovascular events. Regulators have rejected the heart claim. The dramatic survival figures live entirely in confounded epidemiology.
So this lands on Insufficient Data — and I want to be precise about what that means. It does not mean K2 is useless or fake. It means the specific, load-bearing claim driving the hype, that swallowing K2 protects your heart, has not been demonstrated in the only kind of study that could demonstrate it. The pill moves a number. We do not yet know if the number is the point.
The viral "JAMA Cardiology" study is real, but it slowed a coronary calcium score by a hair, not a single heart attack. After 20 years and a dozen trials, not one has shown vitamin K2 prevents a heart attack, stroke, or death, and EU regulators have rejected the cardiovascular claim twice.
- 1. Vossen LM, de Leeuw PW, Schurgers LJ, et al. Vitamin K2 supplementation and coronary artery calcification in patients with coronary artery disease (VitaK-CAC). JAMA Cardiology. 2026 (online June 10). RCT, n=180, 360 mcg/day MK-7, 24 months; CAC progression slowed, p=0.02; not powered for clinical events.
- 2. Public discussion on X regarding the VitaK-CAC trial, June 2026, noting the distinction between a surrogate-endpoint trial and a clinical-outcome trial.
- 3. Schurgers LJ, Cranenburg ECM, Vermeer C. Matrix Gla-protein: the calcification inhibitor in need of vitamin K. Thrombosis and Haemostasis. 2008;100(4):593–603.
- 4. Luo G, Ducy P, McKee MD, et al. Spontaneous calcification of arteries and cartilage in mice lacking matrix Gla protein. Nature. 1997;386:78–81.
- 5. Ibid. The MGP-knockout phenotype establishes MGP as a required local inhibitor of arterial calcification.
- 6. Chatrou MLL, Winckers K, Hackeng TM, et al. Vascular calcification: the price to pay for anticoagulation therapy with vitamin K antagonists. Blood Reviews. 2012;26(4):155–66.
- 7. Sato T, Schurgers LJ, Uenishi K. Comparison of menaquinone-4 and menaquinone-7 bioavailability in healthy women. Nutrition Journal. 2012;11:93.
- 8. Hartley L, Clar C, Ghannam O, et al. Vitamin K for the primary prevention of cardiovascular disease. Cochrane Database of Systematic Reviews. 2015;CD011148. No RCT data on hard cardiovascular outcomes available.
- 9. Knapen MHJ, Braam LAJLM, Drummen NE, et al. Menaquinone-7 supplementation improves arterial stiffness in healthy postmenopausal women. Thrombosis and Haemostasis. 2015;113(5):1135–44. RCT, n=244, 180 mcg/day MK-7, 3 years.
- 10. Diederichsen ACP, Lindholt JS, Möller S, et al. Vitamin K2 and D in patients with aortic valve calcification (AVADEC). Circulation. 2022;145(18):1387–97. RCT, n=365, 720 mcg/day MK-7 + vitamin D, 24 months; primary endpoint null, p=0.64.
- 11. Witham MD, Lees JS, White M, et al. Vitamin K supplementation to improve vascular stiffness in CKD (K4Kidneys). Journal of the American Society of Nephrology. 2020;31(10):2434–45. RCT, n=189, 400 mcg/day MK-7, 12 months; null, p=0.77.
- 12. Zwakenberg SR, de Jong PA, Bartstra JW, et al. The effect of menaquinone-7 supplementation on vascular calcification in patients with diabetes. American Journal of Clinical Nutrition. 2022;115(1):xxx. RCT, n=68, 360 mcg/day MK-7, 6 months; no reduction, PET activity trended worse (p=0.06).
- 13. Holden RM, et al. (iPACK-HD), Nephrol Dial Transplant 2023; Saritas T, et al. (VitaVasK), Clinical Kidney Journal 2022; Oikonomaki T, et al., Int Urol Nephrol 2019. Dialysis RCTs: biomarker improves; calcification outcomes mostly null or statistically fragile.
- 14. Echt DS, Liebson PR, Mitchell LB, et al. Mortality and morbidity in patients receiving encainide, flecainide, or placebo (CAST). New England Journal of Medicine. 1991;324:781–88. The canonical surrogate-endpoint cautionary tale.
- 15. Criqui MH, Denenberg JO, Ix JH, et al. Calcium density of coronary artery plaque and risk of incident cardiovascular events. JAMA. 2014;311(3):271–78.
- 16. EFSA Panel on Dietetic Products, Nutrition and Allergies. Scientific opinion on the substantiation of health claims related to vitamin K. EFSA Journal. 2009;7(9):1228; 2012;10(7):2714. Cardiovascular claim not authorized.
- 17. 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. Journal of Nutrition. 2004;134(11):3100–05. Observational cohort, n=4,807.
- 18. Gast GCM, de Roos NM, Sluijs I, et al. A high menaquinone intake reduces the incidence of coronary heart disease (Prospect-EPIC). Nutrition, Metabolism and Cardiovascular Diseases. 2009;19(7):504–10. Observational, n=16,057.
- 19. Riphagen IJ, Keyzer CA, Drummen NEA, et al. Prevalence and effects of functional vitamin K insufficiency: the PREVEND study. Nutrients. 2017;9(12):1334. Observational; dp-ucMGP associated with mortality, confounded by comorbidity.
- 20. Andrian T, Siriopol D, Voroneanu L, et al. Vitamin K supplementation in dialysis patients: a meta-analysis. Clinical Kidney Journal. 2023;16(12):2738. Eleven RCTs, 830 patients; no impact on mortality. Safety/warfarin interaction per NIH Office of Dietary Supplements Vitamin K fact sheet.