Vitamin K2 and D3: where the calcium in your body goes

You are holding a box. Vitamin D3 plus K2, 5,000 International Units, with 200 micrograms of MK-7 alongside. On the back there is a sentence that sticks: K2 takes the calcium where it belongs.

Sounds logical. Almost too logical.

I found that sentence plausible myself for years. And I still find the biochemistry behind it fascinating. It is just that between what happens inside a cell and what becomes measurable in a large human trial there is often a fairly long stretch of road.

This text walks that stretch. It explains the mechanism as precisely as I wanted to understand it myself. It names the places where the human data contradict each other. And it contains a warning that gets lost in many guides on this topic, even though it is the single most important piece of information in the whole article.

The question the shelf never asks

Imagine you order a delivery of building material for your house. The truck arrives, the material is unloaded, all good. Except nobody said which room it was meant for.

In the end there are bricks in the hallway, in the bathroom and on the stairs. Not because too much was delivered. Because the instruction was missing.

That is roughly the picture used in forums and on packaging to advertise the combination of vitamin D3 and K2. Vitamin D orders the material. Vitamin K2 distributes it. It is a lovely metaphor, and it is not wrong. It is simply incomplete, in a way that can lead you astray.

Because your body does have a site manager. It is called calcium homeostasis, and it works around the clock.

And now you know why I do not start this article with K2, but with the feedback loop.

Calcium balance: a feedback loop with four players

Your blood calcium is one of the most tightly controlled values in the whole body. Regulated more narrowly than your blood sugar. And for good reason: heart rhythm, muscle contraction and the excitability of every single nerve cell hang on this value.

The loop has four participants, and they talk to each other constantly.

The parathyroid glands measure calcium in the blood permanently. If it falls, they release parathyroid hormone. Parathyroid hormone has three addressees. In the kidney it increases the recovery of calcium from the primary urine and drives the conversion of vitamin D into its active form, calcitriol. In bone it mobilises calcium from the depot. And through calcitriol it raises the uptake of calcium from food in the gut.

This is exactly where vitamin D sits. In its active form it is a hormone that can raise calcium uptake in the small intestine, among other things through a calcium channel in the gut wall.

What stands out about that list? Vitamin D appears in three places. Vitamin K2 in none.

That matters and is often overlooked. Vitamin K2 is not regarded as a regulator of the calcium level in blood. On current knowledge it does not measurably change your serum calcium value. It works on an entirely different level: in the tissues where calcium is either meant to be built in or precisely not.

And now you know why the question about K2 only makes sense once you have understood what D3 sets in motion in the first place.

What vitamin K2 does biochemically: carboxylation

Now it gets technical for a moment. I promise it is worth it, because afterwards many advertising claims sound different.

Vitamin K is a cofactor. It is not a building material and not a signal, it is the key for a single enzyme: gamma glutamyl carboxylase. This enzyme attaches an extra chemical group to certain proteins. That process is called carboxylation.

Picture it like fitting magnets to a tool. Without the magnets the tool is there, looks normal, but cannot grip. With the magnets it can bind calcium. Only then does it do its job.

Two of these proteins are decisive for our topic.

Osteocalcin is produced by the bone building cells. Carboxylated, it can bind calcium and build it into the bone matrix. The share of undercarboxylated osteocalcin is therefore used in research as a functional marker for the vitamin K supply of bone.

Matrix Gla protein, MGP for short, sits in the vessel wall and in cartilage. It is regarded as the body's own local inhibitor of soft tissue calcification. Here too vitamin K is needed for activation. The inactive form is called dephosphorylated, undercarboxylated matrix Gla protein, abbreviated dp-ucMGP.

Here comes a detail that explains all the rest. The body prioritises. When vitamin K gets tight, the liver serves the clotting factors first. They are acutely vital, and stopping a bleed tolerates no delay.

Osteocalcin in bone and matrix Gla protein in the vessel wall stand further back in that order. They are the first to be incompletely carboxylated when supply is tight.

And now you know why the building material metaphor has a genuine biochemical core. The only question is whether that core also leads to measurable results in humans.

The calcium paradox hypothesis: what it claims and where it stands

Now comes the part where I risk disappointing you. I am writing it anyway, because I do not want to offer you a text that only confirms what sounds good.

The hypothesis runs roughly like this. When vitamin K supply is tight, matrix Gla protein stays incompletely activated. The vessel wall then brakes calcification less well. At the same time osteocalcin stays incompletely activated, and bone builds calcium in less efficiently. The result would be the paradoxical combination of calcifying vessels and weakening bones. Vitamin K2 should be able to shift that state.

That is mechanistically elegant. And it has so far not been secured as an endpoint result in humans.

Let us look at what four larger randomised trials found. One found a positive signal. Three found no effect, one of them even a tendency in the other direction. That is the honest state of things, and it is more sober than the advertising suggests.

And now the other side, in the same level of detail.

Two further trials round out the picture. In a Dutch study, 68 people with type 2 diabetes and known cardiovascular disease received 360 micrograms of MK-7 or placebo for six months. Calcification in the leg artery was measured with a special PET method. dp-ucMGP fell clearly, the calcification marker on PET did not improve, and there was even a tendency in the other direction, which was not statistically supported and which the authors themselves classified as hard to interpret. On conventional CT there was no difference at all.

In a Scottish trial, 90 people after kidney transplantation received vitamin K or placebo for one year. Neither vascular stiffness nor calcification changed measurably.

What remains on the positive side? The observational data. In the Rotterdam study, 4,807 people were followed over years. Those in the upper third of menaquinone intake had a lower relative risk of death from coronary heart disease and severe aortic calcification less often than those in the lower third. For vitamin K1 this association did not appear. Cohort data like these show a link, not a cause. People who regularly eat ripened cheese and fermented foods usually differ in other ways too from people who do not.

A cross-sectional study in 564 women after the menopause likewise found less coronary calcium with higher menaquinone intake, again not with vitamin K1.

And now you know why I consistently speak of a hypothesis with the calcium paradox, and not of a mode of action.

MK-4 or MK-7: the difference you can read off the dose

You are standing in front of the shelf again. One product says 100 micrograms of MK-7. Another says 15 milligrams of MK-4. That is a difference of a factor of 150. Two products, the same vitamin, completely different numbers.

The reason lies in the chemistry. Vitamin K2 is not a single substance but a family. All members have the same head, but side chains of different lengths. MK-4 has four chain links, MK-7 has seven. This chain determines how the molecule is transported in blood and how long it stays there.

The dose difference in the trials follows directly from this pharmacokinetics. MK-4 disappears quickly, which is why it was used in milligram amounts in Japanese osteoporosis trials, classically 45 milligrams per day. MK-7 stays a long time, which is why microgram amounts are enough in the European trials, typically 180 to 720 micrograms per day.

Both figures are study data. They are explicitly not a recommendation for you, and certainly not one to adopt without medical guidance.

There are human data for MK-4 as well. In an open randomised trial, 109 women after the menopause with osteoporosis received either 45 milligrams of MK-4 daily or calcium as a comparison for six months. Undercarboxylated osteocalcin fell clearly in the MK-4 group after just one month, and the carboxylated share rose. The turnover marker in urine, however, also rose after six months. The authors wrote themselves that further studies would be needed to clarify whether fewer fractures can be derived from this.

And now you know why comparing two K2 products by the plain number on the tin does not work.

The most important warning in this article: vitamin K and blood thinners

If you read only half of this article, then please read this section.

There is a group of prescription-only medicines whose entire mechanism consists of blocking the vitamin K cycle. They are called vitamin K antagonists for that reason. In Germany this is above all the active ingredient phenprocoumon, known under the trade names Marcumar and Falithrom. In the English speaking world it is warfarin. These medicines are prescribed by a doctor and steered via the INR, because both too strong and too weak an effect can be dangerous.

These medicines inhibit the enzyme that returns vitamin K to its usable form after every use. Without that recycling step the liver can produce fewer functional clotting factors. That is precisely the desired effect: blood clots more slowly, and thrombosis and stroke are meant to become rarer as a result.

And now you see the problem. Additional vitamin K refills exactly the store that the medicine deliberately keeps small.

Interestingly, the interaction is not only a risk. It has also been studied deliberately.

What about the newer blood thinners? The direct oral anticoagulants, that is the prescription-only agents apixaban, rivaroxaban, edoxaban and dabigatran, inhibit individual clotting factors directly and, on current knowledge, do not run through the vitamin K cycle. An interaction along this pathway is therefore not to be expected, and the INR is not used as a steering value under them.

That is not an all-clear for supplements in general. Other interaction routes remain, for example via transport proteins and liver enzymes, and these medicines carry their own risks, above all bleeding. So tell your doctor what else you are taking under a DOAC as well.

And now you know why with this topic I ask first about the medication list and not about the product.

Bone: what the vitamin K2 trials support, and what they do not

That leaves the second half of the story. If osteocalcin needs vitamin K in order to build calcium into the bone matrix, surely that should show up in bone density.

Here too the answer is an honest both-and, and the difference often lies in details of the study design.

Sounds convincing. And now the Norwegian trial that arrived at the opposite result in the same field.

How do you put that together? For that there is a meta-analysis I consider particularly straight.

A research group from York reworked the entire body of evidence in 2019, explicitly also because doubts had arisen about the integrity of part of the older work. Across 36 studies, a lower risk of clinical fractures under vitamin K emerged for women after the menopause or with osteoporosis. If the analysis was restricted to studies with a low risk of bias, the statistical support disappeared. For vertebral fractures no difference appeared at all. For bone density, only small and mostly unsupported differences remained after the weaker studies were excluded.

A later systematic review across 20 studies arrived at somewhat more favourable numbers for the fracture rate, but likewise found no supported effect on bone density at the femoral neck.

And now you know why with this topic I so often say: let us first look at where you stand before we add anything.

Food first, dose second, measuring third

I do not start with the capsule on this topic. Not out of stubbornness on principle, but because the food sources for vitamin K2 are unusually well studied.

Vitamin K1, phylloquinone, sits above all in green leafy vegetables, kale, spinach, broccoli and herbs. Vitamin K2, the menaquinones, arise above all through bacteria. That is why they sit in fermented foods and in animal products.

And how much of it do you need? That too is more honest than many guides make out.

Covering a requirement is not the same as therapy

Here is a distinction that is made too rarely in the whole supplement world, and that I consider central.

On one side stands covering a requirement. These are the doses you buy online or in the drugstore. They fill gaps. No therapeutic effect is to be expected from them, and that is not a shortcoming, it is their purpose.

On the other side stands the orthomolecular or therapeutic high dose. Considerably higher amounts, used deliberately, limited in time, under medical supervision and with laboratory monitoring. Only here can a nutrient act like a medicine, and only here do the corresponding risks arise.

The clearest example of this is, of all things, vitamin D, the partner in our topic. The reference value of the professional societies for vitamin D intake is 20 micrograms per day, which corresponds to 800 International Units. Some commercial products sit well above that. Which amount fits your situation is a question for the practice and for your 25-OH vitamin D value, not for a rule of thumb.

Against that stand therapeutic high dose concepts such as the so-called Coimbra protocol. There a multiple of this amount is used, exclusively under close medical monitoring of calcium, parathyroid hormone and kidney values and with strict calcium restriction. I deliberately do not name concrete figures here, because without examination and laboratory supervision they are dangerous.

Why topping up is more often justified today than it used to be

I am no friend of capsule reflexes. And still I believe that targeted topping up is more often justified today than fifty years ago. There are several reasons for that, and I want to name them together with their limits.

Soils, varieties, harvest times and storage have changed. For some nutrients, comparative data across decades show lower contents in fruit and vegetables. These data are methodologically contested, because old and new analytical methods are hard to compare, and part of the effect is explained by higher-yielding varieties with a dilution effect. It is a hint, not a scandal.

Heavily processed foods deliver energy but little micronutrient density. Their share of energy intake is high in Western countries. Environmental exposures and chronic stress can raise the consumption of protective substances. And medicines can change the nutrient balance. For proton pump inhibitors this is described for vitamin B12 and magnesium, for metformin for B12, for diuretics for potassium and magnesium, for hormonal contraception for B6, B12, folate and zinc. That is a reason to look closely and, where appropriate, to measure. It is explicitly not a reason to stop a prescribed medicine. The right route is to keep an eye on the nutrient, not to drop the drug.

For vitamin K specifically there is one further point. Part of the menaquinones comes from the gut flora. Antibiotics, chronic bowel inflammation and fat digestion disorders can change that contribution. How large this contribution actually is in humans in quantitative terms has never been cleanly clarified. That is precisely why EFSA derived no reference value for menaquinones.

And now you know why with me the plate comes first, then the question of dose, and only then the laboratory.

Three levers you can pull today

No protocol, no weekly plans, no brands. Just three things that lie in your hands.

And now you know why the most honest answer to the question about D3 and K2 is not a dose figure, but three questions in return: which medicines are you taking, what does your plate look like, and what did you last measure?

Common questions about vitamin K2 and vitamin D3

How this topic connects to the others

Vitamin K2 rarely stands alone. Calcium balance hangs on vitamin D, on magnesium, on the hormones and on what is on your plate. These paths lead onward.