Vitamin K2 (MK-4 / MK-7)

Cartilage contains vitamin-K-dependent proteins, most importantly matrix Gla protein (MGP), which must be carboxylated — a vitamin-K-dependent reaction — to inhibit mineralisation. Mice lacking MGP develop spontaneous calcification of arteries and cartilage and die of vascular rupture within about two months. Inappropriate cartilage calcification is a recognised feature of osteoarthritis, so the hypothesis writes itself: keep vitamin K status adequate, keep MGP active, keep cartilage un-mineralised.

The cohort evidence

The observational record is unusually consistent. In Framingham Offspring (n=672), higher plasma vitamin K tracked with lower prevalence of radiographic hand OA and knee osteophytes. In MOST (n=1,180, 30 months), subclinical vitamin K deficiency predicted incident radiographic knee OA (RR 1.56) and incident MRI cartilage lesions (RR 2.39). In Health ABC (n=791, 3 years), very low plasma vitamin K was associated with progression of cartilage damage (OR 1.7) and meniscus damage (OR 2.6). Three cohorts, three positive associations — with the standard caveat that vitamin K status tracks overall diet quality and health behaviours.

The one trial — and what it didn't show

The only RCT with a joint outcome (n=378, 3 years) supplemented vitamin K1, not K2, as an ancillary to a bone trial. It was null on its primary endpoint: no overall effect on radiographic hand OA. Participants who were insufficient at baseline and reached sufficiency showed about 47% less joint-space narrowing — but that is a post-hoc subgroup, hypothesis- generating only. Narrative reviews in this space tend to be more enthusiastic than this trial record warrants.

Why "K2" specifically? Pharmacokinetics, not outcomes

Here is the honest wrinkle: every human study above measured phylloquinone — vitamin K1. The leap to "take K2" rests on absorption data, not joint trials. In healthy women, a single 420 µg dose of MK-4 was undetectable in serum at every time point, while MK-7 was well absorbed and accumulated with 60 µg/day dosing. So if you supplement at label doses, MK-7 is the form that measurably raises status. But there are no RCTs of MK-4 or MK-7 with any joint or cartilage outcome. Full stop.

Dose & practical notes

No joint trial establishes a dose. Common commercial MK-7 doses are 90–200 µg/day, which normalise the deficiency marker dp-ucMGP in other contexts. The RCT signal, for what it is worth, came from repleting deficiency — not supra-nutritional dosing — which is the best-supported framing: this looks like a nutrient-status story, not a pharmacology story.

Safety

Vitamin K has low toxicity at nutritional and common supplemental doses. The clinically important issue is the warfarin interaction: vitamin K antagonises warfarin and other vitamin-K-antagonist anticoagulants — anyone on them must not add K1 or K2 without medical supervision. Ironically, warfarin itself inactivates MGP and is associated with OA progression in observational data — mechanistically consistent, and also exactly why anticoagulated patients are the wrong people to self-experiment.

What I take from it

A genuinely elegant mechanism, three concordant cohorts, and a trial record that is one null K1 study deep. A K2 (MK-7) RCT with knee imaging in K-insufficient adults is the single study that would move this tier; as of this review, none with published joint outcomes exists.