Two vitamins. Two jobs. Named after the same letter.
Vitamin K was named after the German word Koagulation — blood clotting — when it was first identified in the 1930s. The name stuck to the whole family. But K1 and K2 have different structures, different food sources, and different roles in the body. They are not interchangeable, and they do not behave the same way as supplements.
K1 — Phylloquinone
Blood clotting. From leafy greens.
K1 activates the clotting factors your blood needs to stop a wound from bleeding: factors II, VII, IX, X, and the anticoagulant proteins C and S. It does this through a process called carboxylation — adding a chemical group to these proteins that makes them functional. Without K1, blood cannot clot. With enough K1, it clots normally.
K1 has a short half-life in the blood — roughly 1 to 2 hours. It is tightly regulated, rapidly cleared, and rarely accumulates. Food K1 is extremely safe. There is no documented toxicity from dietary K1.
Food sources: kale, spinach, collard greens, broccoli, Brussels sprouts, parsley, Swiss chard
K2 — Menaquinones (MK-4, MK-7, and others)
Calcium routing. From fermented food and animal fat.
K2 activates a different set of proteins — osteocalcin in bone and matrix Gla protein (MGP) in arterial walls. Osteocalcin helps bind calcium into bone. MGP inhibits calcium from depositing in artery walls and soft tissue. These are real, documented biological functions.
Food K2 comes in amounts calibrated to what traditional diets provided — micrograms from fermented foods and animal fats, arriving alongside vitamins A, D, and E in a fat-soluble matrix. This context matters. The isolated K2 supplement delivers none of it.
Food sources: natto, gouda, brie, aged hard cheese, pastured egg yolks, grass-fed butter, pastured chicken liver, traditionally fermented dairy
What K deficiency actually looks like
True K1 deficiency in adults is rare outside of specific medical situations — fat malabsorption disorders (Crohn's, celiac, short bowel syndrome), biliary obstruction, or long-term antibiotic use that disrupts the gut bacteria that also produce small amounts of K2. The signs are bleeding-related: easy bruising, wounds that bleed longer than expected, heavy menstrual periods.
Warfarin (Coumadin) works by blocking vitamin K recycling. People on warfarin are intentionally kept K-restricted. This is the mechanism of the drug. It is also why K2 supplements are a serious risk for anyone on anticoagulant therapy — any additional K intake competes directly with the drug's intended effect.
Subclinical K2 insufficiency — enough K1 for clotting, but insufficient K2 for full MGP and osteocalcin activation — is likely common in people eating low-fermented-food, low-pastured-fat diets. This is different from clinical deficiency and does not require a supplement. It requires the foods.
Why food K2 is not the same as supplement K2
Natto contains MK-7 at roughly 900–1,000 mcg per 100g serving — but it also contains protein, fiber, probiotics, and the full nutritional context of fermented soybeans. The MK-7 in natto enters the body through the digestive matrix of a whole food, not as an isolated molecule in a gelatin capsule. Aged cheeses contain MK-4 and MK-8 and MK-9 together — the full menaquinone family — alongside fat-soluble vitamins A and D from the milk. The supplement isolates one form and delivers it without context, at doses that were never calibrated to human physiology. The body does not handle isolated MK-7 at 180–360 mcg daily the same way it handles K2 arriving from food.
The K2+D3 teaching. A claim that has never been tested.
The standard advice in functional and integrative medicine is to stack vitamin K2 with vitamin D3. The logic: D3 raises calcium absorption, K2 activates matrix Gla protein (MGP), MGP keeps calcium out of arterial walls and routes it into bone instead. Add K2 to your D3, protect your arteries.
The mechanism is plausible. K2 does activate MGP. MGP does inhibit vascular calcification. This is real biology. The problem is the specific claim: that supplemental K2 protects arteries from the calcium load driven by supplemental D3. That specific claim — D3 supplementation + K2 supplementation versus D3 supplementation alone — has never been tested in a randomized controlled trial.
The protective teaching is extrapolated from mechanism, not from evidence.
Why the mechanism argument fails in practice
- MGP has a saturation ceiling. Matrix Gla protein requires K2 for carboxylation — the activation step that makes it functional. But there is a finite amount of MGP in arterial walls, and a finite carboxylation capacity. High-dose D3 drives calcium absorption and mobilization continuously, around the clock. A daily K2 supplement dose cannot keep pace with a continuous calcium load from a fat-soluble hormone that accumulates in tissue and releases slowly over months or years. You cannot manage steady overflow by adding a fixed bucket.
- K2 does not lower stored D3. K2 does not lower serum 25-hydroxyvitamin D. It does not slow the release of D3 from adipose tissue. It does not reduce the amount of calcium being pulled into circulation. Every D3 supplement dose adds to the fat-tissue accumulation (Drincic et al., Obesity 2012, PMID 22475077). K2 attempts to manage downstream calcium movement while the upstream driver continues compounding unchecked.
- The stacking logic is circular. The argument "take K2 to manage the side effects of D3" is the same logic as taking one drug to manage the side effects of another. It does not address why the D3 is creating excess calcium in the first place. It adds a second supplement to a problem created by the first.
- D3 can stay in fat tissue for years. Vitamin D3 is fat-soluble and lipophilic — it stores in adipose tissue and the liver. After stopping supplementation, stored D3 continues releasing into circulation for months, and in some cases years, depending on how long and at what dose supplementation occurred. K2 taken daily provides no protection against this slow, continuous release from tissue stores.
What the evidence actually shows — and doesn't show
The most-cited K2 trial is Knapen et al. (2015, Thrombosis and Haemostasis, PMID 25694037): 244 healthy postmenopausal women, 180 mcg/day MK-7 (commercial product MenaQ7, NattoPharma) for 3 years. Findings: reduced dp-ucMGP (a vascular K-deficiency marker) and reduced carotid-femoral pulse wave velocity in the group with the highest baseline arterial stiffness.
What this trial did not show: reduced heart attacks, reduced strokes, reduced cardiovascular mortality. No hard outcome. The authors had disclosed ties to NattoPharma, the commercial supplier of the supplement used. A 2023 systematic review of 14 RCTs (1,533 patients) found inconsistent results across broader arterial calcification outcomes — 6 of 8 trials showed no significant benefit.
Critically: none of these trials were conducted in people taking high-dose D3 supplements. The K2 evidence, such as it is, was generated in general populations. Whether K2 prevents the specific arterial calcium load driven by supplemental D3 is an untested question.
The honest answer
K2 from food — natto, aged cheeses, pastured egg yolks, grass-fed butter, pastured liver — supports normal MGP and osteocalcin function as part of a nutrient-dense diet. That is documented. What is not documented is that K2 supplements, taken alongside high-dose D3 supplements, prevent arterial calcification. The supplement industry sells both together. The research to justify selling them together does not exist.
For the full mechanism breakdown, see the K2+D3 section of the Vitamin D article.
MK-7 is not natto. The half-life changes everything.
K2 from food is safe. This is well-established. K2 supplements — specifically MK-7 in the doses sold commercially — are a different matter. The distinction is the half-life.
MK-7 extracted from natto and put into a capsule has a half-life of approximately 70–72 hours. This means it takes three days for half of a single dose to clear the body. With daily dosing, it accumulates. Food-form K2 from natto arrives embedded in a food matrix and is absorbed and cleared differently. A supplement is not a food.
The warfarin interaction — specific numbers
Warfarin (Coumadin) and other vitamin K antagonists work by blocking the recycling of vitamin K — intentionally impairing the activation of clotting factors. Any additional vitamin K intake counteracts this mechanism. This is why patients on warfarin are told to keep their vitamin K intake consistent.
Rombouts et al. (2013, PMID 23530987) tested MK-7 in 18 patients on acenocoumarol (a warfarin-class anticoagulant):
- At 10 mcg/day MK-7: INR was significantly altered in approximately 50% of subjects (Rombouts et al., 2013, PMID 23530987)
- At 45 mcg/day MK-7: INR decreased by an average of 37% (Rombouts et al., 2013, PMID 23530987)
- Most commercial K2 supplements contain 100–200 mcg MK-7 per dose — 2 to 20 times above the level that altered INR in half of anticoagulated patients
This is not a fringe interaction. It is pharmacokinetically predictable from MK-7's half-life. Anyone on warfarin, acenocoumarol, apixaban (Eliquis), rivaroxaban (Xarelto), or any anticoagulant therapy should not take MK-7 supplements without direct medical supervision and frequent INR monitoring.
Accumulation with daily dosing
Schurgers et al. (2007, Blood, PMID 17158229) established the pharmacokinetics: MK-7 half-life is approximately 70 hours versus 1–2 hours for K1. Daily MK-7 supplementation produces steadily increasing serum levels that plateau after several days of continuous dosing. The resulting sustained elevation is dramatically greater than what food K2 produces. Bioavailability is approximately 2.5 times greater than K1 at equivalent doses.
This accumulation profile is what makes MK-7 supplements a different risk category than food K2 or even K1 supplements. The supplement industry uses MK-7's long half-life as a selling point — "stays active longer." That same property is what makes it dangerous for anyone on anticoagulant therapy, and what distinguishes it from the K2 in fermented food.
The thin evidence base — what "K2 protects arteries" is actually based on
The evidence for K2 supplementation and cardiovascular protection rests on two surrogate biomarkers:
- dp-ucMGP — dephosphorylated-uncarboxylated Matrix Gla Protein, a marker of vascular K deficiency. K2 supplements reliably lower dp-ucMGP. What dp-ucMGP reduction means for actual cardiovascular events is unclear.
- Pulse wave velocity and coronary artery calcification score — arterial stiffness and calcification imaging markers. Evidence is inconsistent across trials.
No randomized controlled trial has demonstrated that K2 supplementation reduces myocardial infarction, stroke, cardiovascular death, or all-cause mortality. The 2023 systematic review of 14 RCTs and 1,533 patients found inconsistent results for hard arterial calcification outcomes. The "K2 protects your arteries" teaching has moved from surrogate biomarker to clinical recommendation without the clinical trial evidence to support that transition.
The safer approach
Regular consumption of K2-rich foods — hard aged cheeses, natto if tolerated, pastured egg yolks, grass-fed butter, pastured liver — supports normal MGP and osteocalcin function without accumulation, without anticoagulant interaction, and without the need to calculate dose. Traditional diets that included fermented dairy, aged cheese, and organ meats provided K2 in context. The supplement isolates one form, removes the context, and delivers it at doses calibrated to a sales model, not to human physiology.
Research & References
Key studies on K2 pharmacokinetics, the anticoagulant interaction, the vascular calcification evidence base, and D3 fat-tissue storage.
K2 Pharmacokinetics & Accumulation
Schurgers LJ, et al. (2007). Vitamin K-containing dietary supplements: comparison of synthetic vitamin K1 and natto-derived menaquinone-7. Blood 109(8):3279–83. PMID 17158229.
Foundational pharmacokinetics. MK-7 half-life ~70 hours vs K1 at 1–2 hours. MK-7 approximately 2.5× more bioavailable than K1. Accumulates with daily dosing and produces greater, more sustained interference with vitamin K antagonists than K1.
Halder M, et al. (2019). Vitamin K: Double Bonds beyond Coagulation — Insights into Differences between Vitamin K1 and K2 in Health and Disease. International Journal of Molecular Sciences 20(4):896. PMID 30791399.
Comparative review of K1 and K2 tissue distribution, biological roles, and safety profiles. Documents how K2 (MK-4 and MK-7) preferentially distributes to extrahepatic tissues — bone, vasculature, brain — while K1 stays predominantly hepatic. Summarizes pharmacokinetic differences explaining why MK-7 carries a different risk profile than dietary K2 forms.
K2 & Anticoagulant Interaction
Rombouts EK, et al. (2013). Daily vitamin K supplementation and INR in anticoagulated patients: a dose-response study. PMID 23530987.
18 subjects on acenocoumarol. At 10 mcg/day MK-7: INR significantly altered in ~50% of subjects (PMID 23530987). At 45 mcg/day: INR decreased by average 37% (PMID 23530987). Commercial D+K2 supplements contain 100–200 mcg — well above both thresholds.
Theuwissen E, et al. (2012). Low-dose menaquinone-7 supplementation improved extra-hepatic vitamin K status, but had no effect on thrombin generation in healthy subjects. British Journal of Nutrition 108(9):1652–7. PMID 22313576.
Healthy adults, MK-7 at 10 or 20 mcg/day for 12 weeks (PMID 22313576). Extra-hepatic K status improved; no effect on thrombin generation in this healthy group. Confirms the distinction between healthy subjects and those on anticoagulant therapy — the Rombouts 2013 data showing INR disruption in anticoagulated patients cannot be extrapolated from healthy-subject trials.
K2 & Vascular Calcification — The Evidence
Knapen MH, et al. (2015). Menaquinone-7 supplementation improves arterial stiffness in healthy postmenopausal women. Thrombosis and Haemostasis 113(5):1135–44. PMID 25694037.
3-year RCT, 244 postmenopausal women, 180 mcg/day MenaQ7 (NattoPharma) (PMID 25694037). Reduced dp-ucMGP and cfPWV. Significant arterial effects only in highest-baseline-stiffness subgroup. Industry-tied authors. No hard cardiovascular endpoints measured.
Systematic review (2023). Vitamin K supplementation, vascular calcification, and arterial stiffness. Frontiers in Cardiovascular Medicine. 14 RCTs, 1,533 patients.
VK consistently reduces dp-ucMGP (surrogate). Broader arterial/valve calcification: inconsistent — 6 of 8 trials showed no significant benefit. No trial demonstrates reduction in MI, stroke, or cardiovascular mortality. Knapen 2015 flagged for unclear allocation concealment.
Beulens JW, et al. (2009). High dietary menaquinone intake is associated with reduced coronary calcification. Atherosclerosis 203(2):489–93. PMID 18722618.
Dietary (not supplement) K2 intake associated with reduced coronary calcification in observational data. Observational study — cannot establish causation. Supports food-form K2 intake, not supplementation.
Geleijnse JM, et al. (2004). Dietary intake of menaquinone is associated with a reduced risk of coronary heart disease: the Rotterdam Study. Journal of Nutrition 134(11):3100–5. PMID 15514282.
Rotterdam Study: dietary K2 intake (from food) associated with 57% reduction in coronary heart disease mortality over 10 years. Dietary association, not supplement trial. The association is with food-form K2, not isolated MK-7.
K2 & Bone
Iwamoto J, et al. (2009). High-dose vitamin K supplementation reduces fracture incidence in postmenopausal women. Nutrition Research 29(4):221–8. PMID 19410972.
Review of Japanese trials using 45 mg/day MK-4 (pharmacological dose, far above any supplement). MK-4 at 45 mg is used as a prescription drug in Japan for osteoporosis — not comparable to the 100–200 mcg in commercial K2 supplements.
D3 Fat-Tissue Storage — Why K2 Can't Fix the Problem
Drincic AR, et al. (2012). Volumetric dilution, rather than sequestration best explains the low vitamin D status of obesity. Obesity 20(7):1444–8. PMID 22475077.
D3 distributes into fat tissue in proportion to body adiposity — the more body fat, the more D3 is stored rather than remaining in circulation. Every dose of supplemental D3 adds to this fat-tissue pool (PMID 22475077). K2 does not affect the stored D3 reservoir, does not accelerate clearance, and does not change the rate at which fat-stored D3 is released. Adding K2 to a D3 supplement stack addresses the downstream calcium routing — not the source of excess calcium mobilization from continuous D3 release. The two problems are independent.