Vitamin K2

Builds bone

K2 uses the same chemistry as K1. It's the cofactor that turns specific glutamate amino acids into calcium-binding Gla residues (see K1). The difference is which proteins it reaches. Menaquinones stay in circulation longer and get to tissues outside the liver, where several non-clotting Gla proteins work.¹

In bone, the main one is osteocalcin. Bone-building cells (osteoblasts) make it, and it's the most abundant non-collagen protein in bone. Its three Gla residues line up to match the spacing of calcium in the hydroxyapatite crystal, so carboxylated osteocalcin binds tightly to bone mineral and helps organise how the crystals are laid into the collagen frame. Osteocalcin that isn't carboxylated can't bind and leaks into blood. The share of uncarboxylated osteocalcin (ucOC) in blood is the standard marker of vitamin K status in bone. Most Western adults have a measurable fraction of it, which means the liver's clotting factors are getting enough K while bone isn't.¹

Human trials. In a 3-year double-blind trial, 244 healthy postmenopausal women took 180 mcg/day of MK-7 or placebo. MK-7 improved vitamin K status, slowed age-related loss of bone density at the spine and femoral neck (though not the total hip), improved estimated bone strength and reduced loss of vertebral height.² Fracture outcomes in Western trials are still lacking. The large Japanese trials of high-dose MK-4 (45 mg) that once seemed to show big fracture reductions can no longer be taken at face value. In 2018 the authors of the main 2006 meta-analysis published a caution to readers, because it relied on trials later found to be problematic.⁷

Keeps arteries clear

Blood is saturated with calcium and phosphate, so soft tissues need active protection from calcifying. The main local protection in artery walls is matrix Gla protein (MGP), made by smooth muscle cells in the vessel wall. Once carboxylated, MGP binds calcium-phosphate crystals as they start to form and blocks the signals that would turn smooth muscle cells into bone-like cells.

How essential MGP is shows up in mice bred without it. They develop normally, then their arteries calcify so heavily that the vessels rupture, and they die within about two months.³ Warfarin, which blocks recycling of all vitamin K, causes arterial calcification in animals and is linked to faster vascular calcification in people. Blood levels of the inactive form (dp-ucMGP) track with arterial stiffness and calcification.

Human evidence:

• Observational. In the Rotterdam Study (4,807 people followed about 7 years), people in the top third of K2 intake had 57% lower coronary heart disease mortality, 26% lower all-cause mortality and about half the rate of severe aortic calcification compared with the bottom third. K1 intake showed no link.⁴ That difference between K1 and K2 is what you'd expect from where each one goes in the body.
• Randomised. In the same 3-year trial of 180 mcg/day MK-7 in 244 postmenopausal women, MK-7 lowered inactive MGP and reduced aortic stiffness (pulse wave velocity). The improvement was clearest in women who started with stiffer arteries.⁵

No trial has yet shown that K2 reduces heart attacks or strokes, so this stays at "some human evidence". But the mechanism, the knockout mice, the warfarin data, the cohort and the stiffness trial all point the same way.

Why it pairs with vitamin D

Vitamin D raises calcium absorption and also tells cells to make more osteocalcin and MGP. But those proteins only work once vitamin K has carboxylated them. More vitamin D without enough K means more calcium coming in and more inactive Gla protein to direct it. That's the reason to pair them. It's a strong inference from the mechanism. It hasn't been tested as a combination in outcome trials.

Forms: MK-4 vs MK-7

• MK-4 is the form animals make from K1 in their own tissues (see K1), and it's found in egg yolk, butter and liver. It clears from blood within hours, so supplements use milligram doses.
• MK-7 is made by bacteria and is richest in natto. It has a half-life of about 3 days, so with daily intake blood levels build up 7–8-fold and stay steady. It also carboxylated osteocalcin more completely than K1 at the same dose.⁶ This is the form the Metabolic Health hub recommends. The long half-life doesn't make intermittent dosing a good idea; take it daily for steady levels.

Warfarin users: because MK-7 lasts so long and is so potent, as little as 50 mcg/day can interfere with warfarin dosing.⁶ Don't start or stop K2 without telling whoever manages your INR.

Intake

There's no separate official intake for K2. The 90–120 mcg/day adequate intake covers vitamin K overall. Trials showing effects on bone and arteries used 180 mcg/day of MK-7. Western diets supply little K2 unless they include fermented foods or organ meats.

Top foods

• Natto: about 850 mcg of MK-7 in 3 oz, by far the richest source
• Hard and soft aged cheeses (Gouda, Brie): about 20–80 mcg per oz, mostly longer menaquinones
• Goose and chicken liver: about 10–100 mcg of MK-4 in 3 oz
• Egg yolks: about 5–30 mcg MK-4 per yolk, more from pasture-raised hens
• Butter from grass-fed cows: about 2–4 mcg MK-4 per tbsp

Deficiency effects

There's no obvious deficiency disease for K2, because the liver grabs K first to keep blood clotting normal. Shortage shows up only as inactive osteocalcin and MGP in blood. Over decades, that's linked to lower bone density and more arterial calcification and stiffness.¹