Natto for Bone Density

The cleanest population-level evidence for natto's skeletal benefit comes from inside Japan itself. Hip fracture incidence in Japanese women is approximately 1.7 times higher in the western prefectures (where natto consumption is low) than in the eastern Kanto and Tohoku prefectures (where natto is a breakfast staple). The Kaneki 2001 Nutrition paper traced this gradient to circulating MK-7 levels — eastern Japanese women had 5-7 times higher serum MK-7 than western Japanese women, who in turn had 4-6 times higher serum MK-7 than European or American women. The intervention trials that followed — the Knapen 3-year postmenopausal MK-7 trial, the Iwamoto Japanese menatetrenone trials, the Inaba osteocalcin gamma-carboxylation studies — demonstrated that supplemental MK-7 reduces age-related vertebral and femoral neck bone loss and partially restores undercarboxylated osteocalcin to the carboxylated form that actively recruits calcium into hydroxyapatite. This page walks through the regional Japanese data, the postmenopausal bone-loss trials, the dual osteocalcin / matrix Gla protein mechanism, the synergy with Vitamin D3 and magnesium, and why isolated calcium supplementation has been disappointing without adequate K2 co-administration.

Table of Contents

  1. Japanese Regional Hip Fracture Gradient
  2. Osteocalcin Mechanism (Bone Calcium In)
  3. Matrix Gla Protein Mechanism (Arterial Calcium Out)
  4. The Knapen 3-Year Postmenopausal MK-7 Trial
  5. Iwamoto Japanese Menatetrenone Fracture Trials
  6. Cochrane and Cockayne Meta-Analyses
  7. The K2 + D3 + Magnesium Triumvirate
  8. The Calcium Paradox — Why Isolated Calcium Disappoints
  9. Clinical Application — Osteopenia and Osteoporosis
  10. Key Research Papers
  11. Connections

Japanese Regional Hip Fracture Gradient

Hip fracture is the most consequential osteoporotic fracture — it carries approximately 20-25% one-year mortality and produces permanent functional disability in a majority of survivors. Population-level data on hip fracture incidence within Japan have shown a striking east-west gradient: eastern Japanese prefectures (Ibaraki, Tochigi, Fukushima, Yamagata, Akita) have substantially lower age-adjusted hip fracture rates than western prefectures (Hyogo, Okayama, Hiroshima, Yamaguchi, Fukuoka).

The Kaneki 2001 Nutrition paper systematically investigated this gradient. The authors documented that:

This is one of the strongest natural-experiment associations in nutritional epidemiology. The genetic and cultural backgrounds of eastern and western Japanese populations are similar; the major diet-composition difference is the eastern habit of natto with breakfast. The MK-7 from natto plausibly explains the bone phenotype.

Subsequent international comparisons have reinforced the conclusion. American and European women have serum MK-7 levels that are an order of magnitude lower than even western Japanese women, and have correspondingly higher age-adjusted hip fracture rates. The Yamaguchi 2006 Yakugaku Zasshi review summarized the cross-population evidence and made the case for K2 as the most important under-recognized determinant of postmenopausal bone health.

Back to Table of Contents

Osteocalcin Mechanism (Bone Calcium In)

Osteocalcin is the second-most-abundant protein in bone (after type I collagen) and the most-abundant non-collagen bone protein. It is secreted by osteoblasts during the bone-mineralization phase, contains three glutamic acid residues that require gamma-carboxylation to function, and after carboxylation binds calcium ions with high affinity and orients them onto the growing hydroxyapatite crystal lattice.

The functional state of osteocalcin can be measured directly. Total osteocalcin is a marker of bone-formation activity. Undercarboxylated osteocalcin (ucOC) is the form that lacks the gamma-carboxylation needed for calcium binding — an elevated ucOC, or an elevated ratio of ucOC to total osteocalcin, indicates inadequate Vitamin K2 status at the bone-osteoblast level. Conversely, carboxylated osteocalcin (cOC) is the functional form that actively recruits calcium into bone.

The Inaba 2015 trial established the dose-response curve. Postmenopausal women given placebo had stable elevated ucOC levels over 4 weeks. Women given 50 µg/day MK-7 had measurable reduction in ucOC by week 4 and continued declining through week 12. Women given 100 µg/day had near-complete normalization by week 4, with 99% gamma-carboxylation by week 12. Women given 200 µg/day reached saturation at week 4 with no further benefit at 12 weeks — defining 200 µg/day as the upper end of the dose-response for osteocalcin activation.

Once bound to calcium, carboxylated osteocalcin participates in the orderly assembly of hydroxyapatite crystals on the type I collagen scaffold. The geometry of crystal deposition determines bone strength: bone with disorganized hydroxyapatite is brittle (low fracture toughness) even at adequate calcium content; bone with well-organized hydroxyapatite is mechanically resilient. The K2-deficient state produces measurable changes in bone microarchitecture on micro-CT, with reduced trabecular thickness and reduced connectivity even at preserved bone mineral density.

This explains a long-standing puzzle in osteoporosis research: bone mineral density (BMD) as measured by DEXA scan correlates only moderately with fracture risk. Some patients with apparently normal BMD have multiple low-trauma fractures; some with low BMD have remarkable structural resilience. Vitamin K2 status is one of the determinants of bone quality (microarchitecture and fracture toughness) that is not captured by BMD alone, and it likely accounts for a meaningful fraction of the BMD-fracture-risk dissociation.

Back to Table of Contents

Matrix Gla Protein Mechanism (Arterial Calcium Out)

The second Vitamin-K2-dependent protein relevant to skeletal health is matrix Gla protein (MGP), although its primary action is in the arterial wall rather than bone. MGP synthesized in vascular smooth muscle cells, secreted into the arterial wall, and (in carboxylated form) binds calcium ions before they crystallize, actively dissolving small calcium deposits in the arterial wall.

The relevance to bone health is the "calcium paradox" phenomenon: postmenopausal women who supplement calcium without adequate K2 often have stable or declining BMD and measurable progression of arterial calcification. The same calcium that is failing to enter bone is being deposited in the arterial wall. The mechanism is that osteocalcin without K2 cannot recruit calcium into bone hydroxyapatite, so circulating calcium remains available for inappropriate soft-tissue deposition; matrix Gla protein without K2 cannot prevent that deposition in the artery wall.

The integrated mechanism is therefore: K2-activated osteocalcin pulls calcium into bone; K2-activated matrix Gla protein keeps calcium out of artery wall. Both processes serve the same physiologic goal of partitioning calcium to its proper location. Both fail simultaneously under K2 deficiency, producing the dual phenotype of low bone density and high arterial calcification that characterizes many postmenopausal women.

The Knapen 2015 Thrombosis & Haemostasis arterial stiffness trial demonstrated the arterial side of this directly — 180 µg/day MK-7 for 3 years reduced arterial stiffness in postmenopausal women by approximately 12% (carotid-femoral pulse wave velocity), with placebo showing no change. The 3-year Knapen 2013 bone trial (same population) demonstrated the bone side — vertebral bone loss reduced from approximately 1.0% per year (placebo) to approximately 0.5% per year (MK-7).

For the cardiovascular calcification angle in depth, see our Atherosclerosis page. For the bone calcium pathway, see Osteoporosis.

Back to Table of Contents

The Knapen 3-Year Postmenopausal MK-7 Trial

The Knapen et al. 2013 Osteoporosis International trial is the most-cited modern intervention trial of MK-7 for bone health. The design was clean and the results were definitive:

Results:

The trial is notable not only for the bone benefit demonstrated but for the modest dose used (180 µg/day MK-7) — approximately one-third of what a single 50 g natto serving delivers. This means the bone benefit observed in the trial is well below the natural intake of regular natto consumers, suggesting the food intervention is likely at least as effective as the supplement intervention tested.

Back to Table of Contents

Iwamoto Japanese Menatetrenone Fracture Trials

The Japanese pharmaceutical regimen for osteoporosis includes menatetrenone (MK-4 form of Vitamin K2) at 45 mg/day — a dose roughly 1,000 times the dietary intake from natto in MK-7 terms, but using the short-half-life MK-4 form that requires high pharmacologic dose to maintain steady-state activation. The drug was approved for osteoporosis in Japan in 1995 and has been the subject of multiple Japanese clinical trials.

Iwamoto J et al. (2005) summarized the Japanese fracture data in Nutrition Research: across multiple Japanese trials totaling several thousand patients, menatetrenone reduced vertebral fracture incidence by approximately 60% and hip fracture incidence by approximately 70% in postmenopausal women with osteoporosis, with parallel reductions in non-vertebral fractures. These are effects sizes comparable to the bisphosphonates (alendronate, risedronate) without the bisphosphonate-class adverse events.

Notably, the menatetrenone trials demonstrated fracture reduction without producing a proportional increase in BMD — consistent with the bone-quality interpretation that K2 improves the architecture and toughness of bone independently of its mineral density. This is a different mechanism than the antiresorptive bisphosphonates, which suppress osteoclast activity and increase BMD; the two interventions are mechanistically complementary and could theoretically be combined, although such combination trials have not been done at scale outside Japan.

Outside Japan, menatetrenone (MK-4) is not licensed as an osteoporosis drug, partly because of the patent landscape and partly because the short MK-4 half-life requires three-times-daily dosing which is impractical compared to once-weekly bisphosphonate dosing. MK-7 supplementation at 100-200 µg/day in the Western market is the practical equivalent path to the same biochemical end (osteocalcin gamma-carboxylation).

Back to Table of Contents

Cochrane and Cockayne Meta-Analyses

The Cochrane Database has reviewed Vitamin K supplementation for fracture prevention twice. Cockayne et al. 2006 Archives of Internal Medicine meta-analysis pooled 13 randomized controlled trials, predominantly using MK-4 in Japanese populations, and concluded that Vitamin K supplementation reduced vertebral fracture incidence by approximately 60% (OR 0.40, 95% CI 0.25-0.65) and hip fracture incidence by approximately 77% (OR 0.23, 95% CI 0.12-0.47). These are very large effect sizes — comparable to or larger than the bisphosphonates — for a nutritional intervention with minimal toxicity.

A more conservative subsequent meta-analysis restricted to high-quality studies and conducted in non-Japanese populations found smaller but still meaningful effects. The discrepancy may reflect baseline K2 status differences between Japanese and Western populations — the effect of repletion is larger in more deficient populations.

The Booth 2008 Journal of Clinical Endocrinology & Metabolism trial in 452 elderly American men and women, by contrast, found no significant effect of 500 µg/day phylloquinone (K1) on bone loss over 3 years. The K1 finding is biologically consistent — K1 is preferentially taken up by the liver for clotting-factor synthesis and produces limited extra-hepatic gamma-carboxylation; the trial does not test K2.

The take-away: K2 (particularly MK-7) is well-supported for fracture prevention; K1 is not. Trials that fail to distinguish between K1 and K2 forms produce ambiguous results and should be interpreted with caution.

Back to Table of Contents

The K2 + D3 + Magnesium Triumvirate

Optimal bone mineralization requires the coordinated action of three fat-soluble or related micronutrients that the modern American diet does not reliably supply: Vitamin D3, Vitamin K2, and magnesium. Each plays a non-redundant role.

The integrated supplementation strategy for bone health is therefore: D3 (2,000-5,000 IU/day, sufficient to maintain serum 25-OH-D above 30 ng/mL), MK-7 (100-200 µg/day, equivalent to the natto serving discussed throughout), magnesium (400-800 mg/day elemental as glycinate, citrate, or malate — avoiding oxide for absorption reasons). The triumvirate is more important than calcium supplementation per se, because most adults consume adequate dietary calcium but inadequate D3, K2, and magnesium.

For a deeper dive into the magnesium piece, see our Magnesium page. For the D3 piece, see Vitamin D3.

Back to Table of Contents

The Calcium Paradox — Why Isolated Calcium Disappoints

Calcium supplementation alone for osteoporosis prevention has been disappointing in large randomized trials. The Women's Health Initiative calcium-plus-D arm (36,000 postmenopausal women) found a modest 12% reduction in hip fracture risk that was not statistically significant in intention-to-treat analysis. Some observational analyses have suggested that calcium supplementation without K2 may increase cardiovascular events — the most controversial finding being the Bolland 2010 meta-analysis suggesting a 27% increased MI risk in calcium-supplemented older adults.

The mechanism, as discussed in the matrix Gla protein section above, is that calcium absorbed in the absence of adequate K2 has nowhere properly to go. Bone osteoblasts cannot recruit it (undercarboxylated osteocalcin), so it circulates and is deposited in soft tissue — arterial wall, kidney, breast, and other ectopic sites. The K2 co-administration is what redirects absorbed calcium back to bone.

This is also why dietary calcium from food is generally safer than supplemental calcium. Most calcium-rich foods (dairy, leafy greens, sardines with bones, almonds) also supply some K2 (dairy) or other co-factors that modulate the bone-vs-artery partition. Isolated calcium carbonate tablets do not.

The practical implication: for an individual concerned about osteoporosis who is not consuming high-calcium foods, calcium supplementation is reasonable — but it should be paired with K2 (100-200 µg/day MK-7, or regular natto consumption) plus D3 plus magnesium. Calcium alone is not just suboptimal — it may be counterproductive.

Back to Table of Contents

Clinical Application — Osteopenia and Osteoporosis

For patients with established osteopenia (T-score -1 to -2.5) or osteoporosis (T-score below -2.5), the K2 intervention is a low-risk, high-evidence adjunct to whatever pharmaceutical or lifestyle management is being pursued:

For comprehensive osteoporosis management, see our Osteoporosis page.

Back to Table of Contents

Key Research Papers

  1. Kaneki M et al. (2001). Japanese fermented soybean food as the major determinant of the large geographic difference in circulating levels of vitamin K2. Nutrition. — PubMed
  2. Yamaguchi M (2006). Regulatory mechanism of food factors in bone metabolism and prevention of osteoporosis. Yakugaku Zasshi. — PubMed
  3. Knapen MH et al. (2013). Three-year low-dose MK-7 supplementation reduces bone loss in postmenopausal women. Osteoporosis Int. — PubMed
  4. Knapen MH et al. (2015). MK-7 supplementation improves arterial stiffness in postmenopausal women. Thromb Haemost. — PubMed
  5. Iwamoto J et al. (2005). High-dose vitamin K supplementation reduces fracture incidence in postmenopausal women. Nutrition Research. — PubMed
  6. Cockayne S et al. (2006). Vitamin K and the prevention of fractures: systematic review. Arch Intern Med. — PubMed
  7. Inaba N et al. (2015). Low-dose daily MK-7 intake improves osteocalcin gamma-carboxylation. J Nutr Sci Vitaminol. — PubMed
  8. Plaza SM, Lamson DW (2005). Vitamin K2 in bone metabolism and osteoporosis. Alt Med Rev. — PubMed
  9. Booth SL et al. (2008). Vitamin K supplementation and bone loss in elderly. JCEM. — PubMed
  10. Tsugawa N et al. (2006). Vitamin K status of healthy Japanese women. AJCN. — PubMed
  11. Bolland MJ et al. (2010). Effect of calcium supplements on risk of myocardial infarction and cardiovascular events. BMJ. — PubMed
  12. Schurgers LJ et al. (2008). Matrix Gla protein carboxylation. Thromb Haemost. — PubMed

PubMed Topic Searches

Back to Table of Contents

Connections

Back to Table of Contents