Vitamin K2: Between Real Vitamin and Big Promise

Vitamin K is a family: K1 (phylloquinone) sits in green vegetables and mainly serves clotting; K2 (menaquinones, including the long-lived MK-7) comes from fermentation and animal products and stars in the supplement narrative — because it circulates longer in blood and activates proteins meant to deposit calcium in bone and clear it from vessel walls (osteocalcin, matrix Gla protein). The biochemistry is real and elegant — which is exactly what makes it so sellable.

The catch sits one level up: that a mechanism exists does not mean capsules improve measurable health endpoints. This gap between beautiful mechanism and hard outcomes is the recurring pattern of this series — from collagen to NMN.

Bones: K2's best card is the three-year trial by Knapen and colleagues: 180 µg MK-7 daily measurably slowed bone density loss at lumbar spine and femoral neck in healthy postmenopausal women. A respectable signal — with the usual footnotes: one trial of this size, bone density as a surrogate (density is measured, fewer fractures are promised — the latter is unproven), and the Japanese high-dose studies often co-cited used pharmacological doses (45 mg MK-4 — 250-fold) that have nothing to do with drugstore K2. A 2025 meta-analysis confirms: K2 improves bone-turnover markers in the blood, but here too fracture endpoints are missing — markers are not the same as fracture protection. For bones the priority list stands unchanged: loading exercise (see also LIFTMOR in the menopause dossier), protein, calcium from food, vitamin D by status. A 2026 RCT post-hoc analysis also shows: vitamin K status does not change how vitamin D acts on bone — another argument against the "K2 is mandatory alongside D3" claim.

Heart/vessels: the Dutch observational cohorts (Prospect-EPIC; high dietary K2 intake ↔ fewer cardiovascular events) founded the category — but observation is not intervention: people who eat a lot of K2 eat and live differently overall. The randomised K2 trials run so far in risk groups long failed to convincingly confirm the hoped-for braking of vascular calcification. That changed for the first time in 2026: a randomised trial with a hard imaging endpoint (JAMA Cardiology, n=180, people with symptomatic coronary artery disease, MK-7 360 µg/day over 2 years) showed significantly slower progression of coronary calcium scores than under placebo. An honest verdict is still needed: a surrogate endpoint, a single trial, a heart-diseased population — no proof for healthy people, but the first hard intervention pointing in this direction.

• Most adults: supply via the plate (green vegetables daily, plus cheese/eggs) — no preparation needed. Taking vitamin D creates no K2 accompaniment duty.
• The osteoporosis topic: treatment belongs in medical hands — there, training, calcium/vitamin D and, where indicated, medication are the proven building blocks. K2 as an add-on is a case-by-case conversation, not a no-brainer.
• If you supplement anyway: MK-7 at trial-like doses (typically 90–180 µg) is well tolerated in healthy people by current knowledge — just match the expectation to the evidence: a possible bone plus, an unproven vessel story.
• THE warning — anticoagulants: anyone taking vitamin K antagonists (phenprocoumon, warfarin) must not take vitamin K preparations on their own initiative — vitamin K is these drugs' direct antagonist; even moderate extra amounts shift the anticoagulation setting. Any change (including drastic dietary shifts in green vegetables) belongs in coordination with the treating practice. Modern DOAC anticoagulants do not carry this interaction — when in doubt, the rule still holds: ask, don't guess.

Consistently, no product link here either — for a preparation whose main promise is unproven and whose main risk group is real, it would be the wrong signal.