Vitamin D3 + K2: a sound mechanism, an unproven payoff
Published September 11, 2026. This page argues a position. Every other page on this site computes its answer from data and does not.
Run and checked by Dan, a registered nurse who trains and uses these supplements. Every figure names its source, and links to it where the source is public. About this site. Not medical advice.
Every study linked below was opened and checked against what this page says about it, by a person, on September 11, 2026.
Most vitamin D3 shelves now carry a D3 + K2 version beside the plain one, sold on a simple idea: D3 helps you absorb calcium, and K2 makes sure that calcium ends up in bone instead of in your arteries.
Price is barely the question. At BulkSupplements the 5,000 IU softgel is $24.97 for 365 plain and $22.97 for 365 with K2. The question is whether the K2 does anything you would notice.
The mechanism is real
Vitamin K is a cofactor the body needs to activate two proteins: osteocalcin in bone and matrix Gla protein in the vessel wall. That is textbook biochemistry, not marketing. The long-chain K2 in most of these products, MK-7, stays in the blood much longer than K1, builds up to 7 to 8 times higher levels with daily use, and activated osteocalcin more completely in a direct comparison (Schurgers 2007).
K2 supplements measurably do this in people. In every trial below that measured them, the inactive forms of these proteins fell in the K2 group. So when a label says K2 “activates” these proteins, that part checks out.
What happened when trials measured what people care about
A marker moving is not the same as a benefit. The trials that looked at bone density and calcification mostly came back empty.
| Trial | Who | What they took | Result |
|---|---|---|---|
| AVADEC, Circulation 2022 | 365 men, average age 71, with aortic valve calcification | 720 mcg MK-7 plus 25 mcg (1,000 IU) vitamin D daily, 2 years | No effect on valve, aortic or coronary calcification |
| AVADEC coronary analysis, JACC Advances 2023 | 304 of those men with no prior ischemic heart disease | Same | No significant reduction in coronary calcification |
| Rønn, Osteoporosis International 2021 | 142 postmenopausal women with osteopenia | 375 mcg MK-7 on top of vitamin D3 and calcium, 3 years | Bone density fell in both groups, no difference |
| Emaus, Osteoporosis International 2010 | 334 early postmenopausal Norwegian women | 360 mcg MK-7 (natto capsules), 1 year | No difference in bone loss |
| Knapen, Osteoporosis International 2013 | 244 healthy postmenopausal women | 180 mcg MK-7, 3 years | Slower loss of bone density at the spine and femoral neck, not the total hip |
AVADEC matters most here, because it tested the actual combination: K2 plus vitamin D against placebo. Its vascular protein marker moved sharply (down 212 pmol/L, against a rise of 45 on placebo), yet the valve calcification it was meant to slow progressed about the same in both groups (275 against 292 units, P = 0.64). It also used about seven times the K2 in the 100 mcg BulkSupplements softgel. The coronary analysis of the same trial found no significant difference in its main measure; its authors called the differences they saw in one subgroup and in safety events “hypothesis-generating”, meaning worth a future trial, not proof. The study tablets were supplied free by two K2 companies, and the paper states the funders had no role in the analysis.
The one clearly positive trial, Knapen 2013, is real, and it should be said plainly. The same research group later reported less arterial stiffness in the same 244 women (Knapen 2015). That is encouraging, but it is a measurement of stiffness, not of heart attacks or strokes.
Rønn 2021 is the closest bone trial to what a D3 + K2 softgel is: K2 added on top of vitamin D3, compared with vitamin D3 alone. After three years there was no difference.
The fracture evidence comes with a caution attached
The most-cited positive result for vitamin K and fractures is a 2006 meta-analysis in Archives of Internal Medicine. It found large reductions in fractures, but all seven trials with fracture data were Japanese and used menaquinone-4, a different form of K2 from the MK-7 in most D3 + K2 softgels. In 2018 JAMA Internal Medicine published a notice from one of its authors cautioning readers that the review included problematic trials. When a review's own author asks readers to be careful with it, I would be too.
The one thing that is not uncertain: blood thinners
Warfarin and similar drugs work by blocking vitamin K, so extra vitamin K pushes against them. A 2007 study warned that 50 mcg a day or more of MK-7 may interfere with anticoagulant treatment in a clinically relevant way (Schurgers 2007). A 2013 dose-finding study in healthy volunteers on acenocoumarol found that as little as 10 mcg a day caused a clinically relevant drop in INR in at least 40% of them, and its authors concluded that MK-7 supplements should be avoided by patients on this kind of drug. Every D3 + K2 product we list contains 50 to 200 mcg. If you take one of these medicines, this is a question for whoever manages it, not a shopping decision.
What I would actually say
Not that D3 + K2 is a scam. The biochemistry is sound, the markers move, and one decent trial found slower bone loss. But the payoff the combination is sold on, stronger bones and cleaner arteries, has not shown up in most of the trials that measured it, including a two-year trial of K2 plus vitamin D in 365 men.
So buying it is a bet on a mechanism. That is why this site compares D3 + K2 on its D3 content alone, lists the K2 as a stated spec, and never lets the K2 move one product above another.
As far as I can find, no large trial in the general population has compared D3 + K2 against D3 alone on fractures or heart events. If one exists, or appears later, the corrections page is how to tell us. It would change the argument above rather than decorate it.
Sources
- Schurgers et al., Blood 2007. Vitamin K1 vs natto-derived MK-7: absorption, half-life, osteocalcin carboxylation, anticoagulant warning. PubMed 17158229
- Diederichsen et al., Circulation 2022. AVADEC: vitamin K2 and D in aortic valve calcification. PubMed 35465686
- Hasific et al., JACC Advances 2023. AVADEC coronary artery calcification. Funding and tablet supply stated in the paper. PubMed 38938724
- Rønn et al., Osteoporosis International 2021. MK-7 plus vitamin D3 and calcium in osteopenia, 3 years. PubMed 33030563
- Emaus et al., Osteoporosis International 2010. MK-7 in early menopause, 1 year. PubMed 19937427
- Knapen et al., Osteoporosis International 2013. MK-7 and bone loss, 3 years. PubMed 23525894
- Knapen et al., Thrombosis and Haemostasis 2015. MK-7 and arterial stiffness. PubMed 25694037
- Cockayne et al., Archives of Internal Medicine 2006. Vitamin K and fractures, meta-analysis. PubMed 16801507
- Torgerson, JAMA Internal Medicine 2018. Notice cautioning readers about the 2006 review. PubMed 29582044
- Theuwissen et al., Journal of Thrombosis and Haemostasis 2013. MK-7 dose and anticoagulant stability. PubMed 23530987
Prices quoted move; the D3 + K2 and plain D3 pages have the current figures. This is not medical advice, and nothing here says how much vitamin D or K2 anyone should take.
What this rests on: prices we last recorded at the retailer named on each product page, which go stale, and each page says when it was checked; the products we track, not every product sold, and Amazon and Walmart are not among the stores; listings and labels read at the source, each linked from the page. Found an error? Tell us. The full method: How we compare.