Vitamin K2 for Atherosclerosis

Vitamin K2 is unproven for atherosclerosis: most trials found it did not slow artery calcification, one 2-year trial in coronary artery disease found it did, and none was designed to test heart attacks or strokes. That trial, VitaK-CAC, gave 180 people with symptomatic coronary artery disease 360 mcg of MK-7 a day or placebo for 2 years. Coronary calcium rose less with MK-7 (median 135 to 184 AU vs 145 to 214 AU, P = .02), and the authors say its meaning for plaque stability is still unknown (Vossen 2026). The largest trial, AVADEC, gave 365 older men 720 mcg of MK-7 with 25 mcg of vitamin D for 2 years and found no slowing of valve, aortic or coronary calcification (Diederichsen 2022).

Smaller trials in type 2 diabetes (Zwakenberg 2019), older people with vascular disease (Fulton 2016) and dialysis (Haroon 2023; Levy-Schousboe 2021) were null. Two meta-analyses of vitamin K trials found small reductions in calcification scores (Lees 2019; Li 2023). The strongest-sounding evidence, lower coronary heart disease rates in people who eat more menaquinone, is observational (Geleijnse 2004; Chen 2019). A 2021 critical review concluded that the claim of unique cardiovascular benefits for menaquinones is not supported by the intervention trials (Shea 2021).

Matrix Gla protein (MGP), present in vascular smooth muscle, may help prevent abnormal calcification (NIH ODS), but it only works as a calcification inhibitor once it has been carboxylated, which requires vitamin K (Shea 2021). Low vitamin K status leaves more MGP in its inactive form, measured in blood as dp-ucMGP, and higher dp-ucMGP has been linked to higher cardiovascular and all-cause mortality in cohort studies (Chen 2019).

Supplements reliably fix that marker: MK-7 lowered dp-ucMGP by 50% compared with placebo over 3 years in healthy women (Knapen 2015) and by 40% in dialysis patients (Levy-Schousboe 2021). Both vitamin K1 and menaquinones can carboxylate MGP (Shea 2021), but most artery trials chose MK-7.

A better blood marker did not translate into slower calcification in most trials. In dialysis, MK-7 raised serum MK-7 40-fold and still did not slow the progression of coronary or aortic calcification compared with placebo (Levy-Schousboe 2021). In type 2 diabetes, PET activity in the femoral (thigh) arteries tended to rise rather than fall with MK-7 (P = 0.06) (Zwakenberg 2019).

Calcification is also a double-edged marker. Coronary calcium predicts heart attacks and death (Hasific 2023), yet in VitaK-CAC the rise in calcium tracked non-calcified plaques becoming partly calcified, and the authors could not say whether slowing this is good or bad for plaque stability (Vossen 2026). Prevention guidelines target LDL cholesterol, blood pressure, smoking and diabetes (Arnett 2019), and in the one trial the Cochrane review found, vitamin K2 had no effect on blood pressure or lipids (Hartley 2015).

Nearly every trial here used MK-7, mostly at 180-720 mcg a day, well above the adequate intake for all vitamin K of 120 mcg a day for men and 90 mcg for women (NIH ODS); two small trials used 90-100 mcg (Fulton 2016; Kurnatowska 2015). VitaK-CAC and the diabetes and dialysis trials used 360 mcg a day (Vossen 2026; Zwakenberg 2019; Levy-Schousboe 2021); AVADEC used 720 mcg with vitamin D (Diederichsen 2022); the arterial stiffness trial used 180 mcg for 3 years (Knapen 2015); and the vascular disease trial used 100 mcg (Fulton 2016).

No upper limit has been set for vitamin K because harms from food or supplements have not been reported (NIH ODS). Natto, a fermented soybean food, is the richest dietary source, at about 850 mcg of MK-7 per 3 ounces (NIH ODS).

VitaK-CAC is the only trial in symptomatic coronary artery disease. Among patients with calcium scores of 50-400 AU, 360 mcg MK-7 a day for 2 years slowed the rise in coronary calcium score and calcium mass compared with placebo (P = .02), with no significant adverse effects (Vossen 2026). It did not measure heart attacks or deaths.

AVADEC, in 365 men with aortic valve calcification, found no difference in valve, aortic or coronary calcification after 2 years of 720 mcg MK-7 plus vitamin D, and cardiovascular events, a safety outcome, were the same in both groups (10 vs 10) (Diederichsen 2022). A secondary analysis of 304 men without known heart disease found a non-significant trend (mean rise 203 vs 254 AU, P = 0.089), a significant slowing only in men with scores of 400 AU or more (P = 0.047), and fewer heart attacks, revascularisations or deaths (1.9% vs 6.7%, P = 0.048); the authors called these findings hypothesis-generating (Hasific 2023).

In 68 people with type 2 diabetes and heart disease, 360 mcg MK-7 for 6 months did not change femoral artery calcification on CT (P = 0.18) (Zwakenberg 2019). In 80 older people with vascular disease, 100 mcg for 6 months did not improve endothelial function or other vascular measures, including carotid wall thickness (Fulton 2016). In an open-label trial of 60 people with carotid narrowing on statins, adding K2 did not significantly change plaque composition over 12 months (Tshomba 2021).

Kidney disease, where vascular calcification is common, was expected to show the clearest benefit. It did not: Trevasc-HDK in 178 haemodialysis patients, which also found no difference in major cardiac events and death (HR 0.98, 95% CI 0.50-1.94) (Haroon 2023), and RenaKvit in 48 dialysis patients (Levy-Schousboe 2021) found no slowing of coronary calcification. One small study in 42 people with chronic kidney disease not on dialysis found less carotid wall thickening with K2 plus vitamin D than vitamin D alone (P = 0.005), but no difference in coronary calcium (Kurnatowska 2015).

A 2019 meta-analysis of 13 trials (n=2,162) found vitamin K reduced vascular calcification by 9.1% (95% CI -17.7 to -0.5) but not vascular stiffness, and warned of few, heterogeneous studies (Lees 2019). A 2023 meta-analysis of 14 trials (1,533 patients) found a small slowing of coronary calcium (mean difference -17.37 on the calcium score, P = 0.04) (Li 2023). Both analysed vitamin K as a class rather than K2 specifically. The 2015 Cochrane review found only one small trial and no data on cardiovascular events (Hartley 2015).

Arterial stiffness is a related but separate measure. In 244 healthy postmenopausal women, 180 mcg MK-7 for 3 years improved aortic stiffness (Knapen 2015), covered on the vitamin K2 for arterial flexibility page.

In the Rotterdam Study of 4,807 adults, people in the top third of menaquinone intake had a lower risk of coronary death (RR 0.43, 95% CI 0.24-0.77) and less severe aortic calcification (OR 0.48) than those in the bottom third (Geleijnse 2004). Among 564 postmenopausal women, higher menaquinone intake went with less coronary calcification (RR 0.80) (Beulens 2009), and a meta-analysis of cohorts found 30% lower coronary heart disease risk with high menaquinone intake, from two studies, while stressing that causality cannot be established (Chen 2019).

What's missing is a trial with heart attacks, strokes or death as the main outcome, and more trials in women and in people without kidney disease or diabetes. For vitamin K1, one 3-year trial of 500 mcg a day found no overall difference in coronary calcium; less progression appeared only in subgroups (adherent participants and those with existing calcification), which the authors called hypothesis-generating (Shea 2009).

MK-7 was well tolerated in these trials: VitaK-CAC saw no significant adverse effects over 2 years (Vossen 2026), the diabetes trial reported none (Zwakenberg 2019), and the 2023 meta-analysis found no difference in adverse events between vitamin K and control (Li 2023). No upper limit has been set for vitamin K (NIH ODS).

The real risk is with blood thinners that work against vitamin K. People taking warfarin and similar drugs need a consistent vitamin K intake, because sudden changes can weaken or strengthen the anticoagulant effect (NIH ODS). In volunteers on acenocoumarol, as little as 10 mcg MK-7 a day caused a clinically relevant fall in INR in at least 40% of people, and the authors advise avoiding MK-7 supplements with these drugs (Theuwissen 2013).

Vitamin K antagonists are the one serious interaction. Long courses of antibiotics can lower vitamin K status, and bile acid sequestrants such as cholestyramine, used for cholesterol, can reduce vitamin K absorption (NIH ODS). Orlistat can reduce absorption of fat-soluble vitamins and, combined with warfarin, may raise prothrombin time (NIH ODS).