Do the nattokinase and vitamin K2 in natto lower blood pressure and protect bones?

Most fermented soy foods — miso, tofu, tempeh — share similar isoflavone profiles. Natto is different because fermentation with Bacillus subtilis produces two things found nowhere else in significant amounts:

Nattokinase is an enzyme secreted by the fermenting bacteria. In the lab it acts directly on fibrin (the protein scaffold of blood clots) and also stimulates the body's own clot-dissolving pathways, raising tissue plasminogen activator (tPA) levels and degrading plasminogen activator inhibitor⁴. How much active enzyme actually reaches the bloodstream after eating natto or taking a supplement is still not well characterized in people¹⁴. In the Takayama cohort, the cardiovascular association was seen with natto but not with other soy foods¹.

Vitamin K2 as MK-7 is the long-chain form of menaquinone, primarily found in fermented foods. MK-7 has a half-life of roughly three days, compared to hours for K1, meaning a single daily serving keeps tissues saturated. Its critical role is activating two proteins: osteocalcin (which binds calcium into bone matrix) and matrix Gla protein (which prevents calcium from depositing in artery walls). In practice, this means MK-7 simultaneously supports bone strength and arterial flexibility. See our Vitamin K2 page for more on this routing mechanism.

Cardiovascular Benefits

A 2023 meta-analysis pooling six randomized controlled trials and 546 participants found that nattokinase supplementation reduced systolic blood pressure by 3.45 mmHg and diastolic by 2.32 mmHg². A 2016 double-blind RCT in 79 adults with elevated blood pressure (74 completed it) found that 100 mg of nattokinase a day for eight weeks lowered both systolic and diastolic pressure, most clearly in men, whose diastolic pressure fell from 86 to 81 mmHg. A fall in von Willebrand factor — a clotting-risk marker — appeared in women, but only at a lenient significance threshold (P < 0.1)³.

In a larger clinical study of 1,062 people with atherosclerosis or high cholesterol, 10,800 fibrinolytic units (FU) a day for 12 months was followed by thinner carotid artery walls, smaller carotid plaques, and better lipid profiles, while 3,600 FU a day did nothing; the authors also reported no noticeable adverse effects¹¹. This is a single clinical study, so placebo-controlled replication is still needed.

It is worth noting one important null result: a three-year RCT in 265 healthy adults (median age 65) with no clinical cardiovascular disease found that 2,000 FU a day of nattokinase had no effect on carotid intima-media thickness or arterial stiffness (measures of arterial plaque progression)⁸. The likely explanation is that this population had low baseline cardiovascular risk — and the dose was far below the 10,800 FU/day used in the positive study¹¹. Nattokinase appears to lower blood pressure modestly, but it has not been shown to slow plaque in low-risk people. The observational evidence from Japan, where natto is eaten regularly from early adulthood, may reflect long-term primary prevention rather than therapeutic reversal.

A second large Japanese cohort, of 92,915 adults aged 45–74 followed for a median of 14.8 years, found that eating more fermented soy (mainly natto and miso) was linked to about 10% lower all-cause mortality (highest versus lowest fifth: hazard ratio 0.90 in men and 0.89 in women), while total soy intake showed no link with mortality. Natto in particular was linked to lower cardiovascular mortality in both sexes. The authors caution that unadjusted residual confounding may explain part of the association⁹.

Bone Health

A three-year placebo-controlled trial in 244 postmenopausal women found that 180 mcg of MK-7 daily significantly slowed bone mineral density loss at the lumbar spine and femoral neck (though not at the total hip), with measurable improvements in bone strength indices⁵. A serving of natto supplies more MK-7 than that: the regular natto tested in one Japanese study contained 775 mcg per 100 g, and eating it raised blood MK-7 and the active, carboxylated form of osteocalcin¹⁰.

A separate 12-month RCT in 148 postmenopausal women with osteopenia confirmed the mechanism: MK-7 reduced undercarboxylated osteocalcin — the inactive form that means calcium is not being properly incorporated into bone — by 65.6%, versus only 6.4% in the placebo group. Trabecular bone number in the tibia held stable in the MK-7 group while declining 3.5% in placebo, although bone mineral density itself did not differ between the groups⁶.

Studies of people who actually eat natto point the same way, but they are observational. In the Japanese Population-based Osteoporosis cohort, postmenopausal women who ate more natto lost less bone at the femoral neck over three years¹³. In a follow-up of 1,417 postmenopausal women (median 15.2 years), those who ate seven or more packs a week (a pack is about 40 g) had about half the risk of an osteoporotic fracture of those who ate less than one pack a week (hazard ratio 0.51, 95% CI 0.30–0.87, adjusted for age and hip bone density)¹².

Gut Microbiome Effects

Bacillus subtilis spores are heat-resistant and survive both cooking and stomach acid. In a Japanese study of a natto-powder capsule containing B. subtilis var. natto spores, the relative abundance of Bifidobacterium and Blautia rose in men and of Bifidobacterium in women, and the size of the change depended on how much Bifidobacterium each person had at baseline⁷. The authors describe it as a case study of the gut microbiota, so it shows a shift in bacteria counts rather than a health benefit.

Practical Notes

Natto is sold fresh in most Japanese or Asian grocery stores, typically in small polystyrene cups (50 g portions). Traditional preparation involves stirring it vigorously, which is thought to enhance the sticky mucilage. It is commonly eaten over rice with soy sauce and mustard, or mixed into miso soup.

For those unfamiliar with the flavor — which is earthy, slightly ammonia-like, and intensely savory — starting with small portions mixed into other foods (avocado toast, eggs, grain bowls) can ease the transition.

Nattokinase supplements are available for those who cannot tolerate the taste, though whole natto delivers MK-7, probiotics, protein (around 18 g per 100 g), iron, manganese, and copper alongside the enzyme.

People taking warfarin or other anticoagulants should consult their doctor before eating natto regularly, as the high vitamin K2 content may interfere with dosing.

Large Prospective Cohorts: Natto and Mortality

The strongest evidence for whole natto comes from the Takayama study (Nagata et al., 2017) — a 16-year prospective cohort of 29,079 Japanese adults (13,355 male, 15,724 female) that recorded 1,678 CVD deaths, including 677 from stroke and 308 from ischemic heart disease¹. Comparing the highest versus lowest quartile of natto intake, the total CVD mortality hazard ratio was 0.75 (95% CI: 0.64–0.88, p-trend=0.0004); the hazard ratio was 0.68 (95% CI: 0.52–0.88) for total stroke mortality and 0.67 (95% CI: 0.47–0.95) for ischemic stroke mortality. Total soy protein and isoflavone intake, and soy protein and isoflavones from soy foods other than natto, were not associated with total CVD mortality. That suggests, but cannot prove, that fermentation-specific compounds such as nattokinase and MK-7 account for the pattern.

The Japan Public Health Center-based Prospective Study (Katagiri et al., 2020) followed 92,915 adults aged 45–74 (42,750 men and 50,165 women) for a median of 14.8 years, during which 13,303 deaths were identified⁹. Total soy product intake was not associated with total mortality (highest versus lowest fifth: hazard ratio 0.98 in both men and women). Fermented soy intake (natto and miso) was inversely associated with all-cause mortality: 0.90 (95% CI 0.83–0.97) in men and 0.89 (0.80–0.98) in women. Natto showed significant inverse associations with total cardiovascular mortality in both sexes. The authors caution that the association for fermented soy might be attenuated by unadjusted residual confounding.

Both studies are observational. Natto is more commonly eaten by older, rural, and traditional Japanese adults — populations that may differ in other health behaviors — and confounders such as diet quality, physical activity, and socioeconomic status were adjusted for but cannot be fully excluded.

Nattokinase Trials and Meta-Analysis

Li et al. (2023) synthesized six RCTs with 546 total participants in a systematic review and meta-analysis². The pooled estimate showed that nattokinase significantly reduced systolic BP by 3.45 mmHg and diastolic BP by 2.32 mmHg. The lipid results were not favorable at the doses studied: relatively low total doses were associated with higher total and LDL cholesterol and lower HDL cholesterol; relatively high total doses were associated with higher total cholesterol but no difference in HDL or LDL; triglycerides were unchanged; and blood glucose rose slightly (mean difference 0.40). The authors concluded that nattokinase can serve as an adjunctive therapy for hypertension, that low-dose supplementation probably has no lipid-lowering effect, and that whether its effects are dose-dependent needs further work. No notable adverse events were reported.

Jensen et al. (2016) conducted a randomized double-blind placebo-controlled multicenter trial in 79 North American adults with elevated blood pressure (74 completed it) over eight weeks, using 100 mg a day of a nattokinase concentrate from which vitamin K2 had been removed³. Systolic pressure fell in both sexes, more robustly in males. Diastolic pressure decreased from 87 to 84 mmHg overall (p<0.05 compared with placebo, which stayed at 87), with a stronger reduction in males (86 to 81 mmHg, p<0.006). Von Willebrand factor — a clotting risk marker — declined in females, but only at a significance level of P<0.1. In the subgroup with low baseline plasma renin activity, 66% showed renin increases in the nattokinase group versus 8% with placebo (P<0.1). The authors interpret the results as suggesting sex-specific mechanisms.

Chen et al. (2022) reported a clinical study of 1,062 participants with atherosclerosis or hyperlipidemia who took nattokinase for 12 months¹¹. At 10,800 FU/day the lipid profile improved significantly, and carotid artery intima-media thickness and carotid plaque size fell; at 3,600 FU/day nattokinase was ineffective. The lipid-lowering effect was more prominent in smokers, in people who drank alcohol, and in people with a higher BMI, and the authors report that regular exercise improved the effect and that co-administration of vitamin K2 and aspirin produced a synergistic effect. No noticeable adverse effects were recorded. The authors argue that their results challenge the commonly recommended 2,000 FU/day dose. The published abstract describes dose and lifestyle comparisons rather than a placebo-controlled trial, so the findings need independent confirmation.

Weng et al. (2017) reviewed nattokinase as an oral antithrombotic agent: a blood-clot-dissolving protein produced by B. subtilis during fermentation, whose safety had been evaluated and which was then in a Phase II trial for atherothrombotic prevention in the United States⁴. Its proposed actions include direct fibrin breakdown and effects on the plasminogen-activator system, which differ from those of antiplatelet drugs. How much active enzyme survives digestion and reaches the circulation in people is still an open question¹⁴.

The important null result comes from Hodis et al. (2021): 265 adults without clinical cardiovascular disease (median age 65.3) were randomized to 2,000 FU/day of nattokinase or matching placebo and followed for a median of three years⁸. Annualized change in carotid intima-media thickness and carotid arterial stiffness did not differ between nattokinase and placebo, and nattokinase had no effect on blood pressure or on any laboratory measure, including blood coagulation, fibrinolysis, and inflammatory markers. The authors concluded that nattokinase has a null effect on subclinical atherosclerosis progression in healthy individuals at low risk of CVD. That does not settle whether higher doses or higher-risk people would benefit, as the Chen study suggests¹¹.

Vitamin K2 (MK-7) and Bone Outcomes

Knapen et al. (2013) conducted a three-year placebo-controlled trial in 244 postmenopausal women randomized to 180 mcg MK-7 daily or placebo⁵. MK-7 improved vitamin K status and significantly attenuated the age-related decline in bone mineral content and density at the lumbar spine and femoral neck, but not at the total hip. Bone strength indices at the femoral neck were favorably affected, and loss of vertebral height in the lower thoracic region was reduced. Three years is a useful duration because bone remodeling cycles require extended observation.

Rønn et al. (2016) studied 148 postmenopausal women with osteopenia, all supplemented with calcium and vitamin D, randomized to 375 mcg MK-7 or placebo for 12 months⁶. Undercarboxylated osteocalcin — a biomarker of inadequate K2 status — fell 65.6% in the MK-7 group vs. 6.4% in placebo after 3 months (p<0.01). High-resolution peripheral quantitative CT imaging at the tibia showed that trabecular number was unchanged in the MK-7 group (−0.1%) but declined in placebo (−3.5%), trabecular spacing was unchanged in MK-7 (+1.2%) versus worsening in placebo (+4.5%), and trabecular thickness was unchanged in MK-7 (+0.2%) versus increased in placebo (+4.0%). There were no differences between groups in bone mineral density at any site or in radius microarchitecture. Fracture outcomes were not measured.

Natto Itself: Biomarker and Observational Bone Data

Tsukamoto et al. (2000) gave eight male volunteers, in sequence a week apart, natto containing 775, 1,298, or 1,765 mcg of MK-7 per 100 g¹⁰. Serum MK-7 and gamma-carboxylated osteocalcin (the active form) rose significantly with the 1,298 and 1,765 mcg natto, and undercarboxylated osteocalcin fell significantly with the 1,765 mcg natto. In 134 healthy adults, men who ate regular natto occasionally or frequently had significantly higher serum MK-7 and carboxylated osteocalcin than men who did not eat it.

Ikeda et al. (2006) analyzed 944 Japanese women (aged 20–79) in the JPOS cohort¹³. Among postmenopausal women, more habitual natto intake was associated with higher total hip bone density at baseline and with less bone loss over three years at the femoral neck (P < 0.0001) and the distal third of the radius (P = 0.0002); the femoral neck association persisted after adjusting for covariates. Tofu and other soy products showed no such association.

Kojima et al. (2020) followed 1,417 postmenopausal women from the same cohort for a median of 15.2 years (17,699 person-years), during which 172 women had osteoporotic fractures¹². Compared with less than one pack (about 40 g) a week, the hazard ratios were 0.72 (95% CI 0.52–0.98) for 1–6 packs and 0.51 (0.30–0.87) for 7 or more packs after adjustment for age and total hip bone density; with further adjustment for BMI, fracture history, cardiovascular disease, diabetes, smoking, alcohol, other soy foods, and dietary calcium they were 0.79 (0.56–1.10) and 0.56 (0.32–0.99). Tofu and other soy products were not associated with fracture risk.

Gut Microbiome: Human Data

Kono et al. (2022) examined changes in the gut microbiota after ingestion of a processed natto food containing Bacillus subtilis var. natto SONOMONO spores (a natto-powder capsule), using 16S rRNA sequencing and a large Japanese intestinal microbiota database as a reference⁷. The relative abundance of Bifidobacterium and Blautia increased in males and of Bifidobacterium in females in the group taking it, and the effect on both genera depended on the baseline abundance of Bifidobacterium. The authors describe the work as a case study, and the genera that changed may be associated with lifestyle-related diseases such as diabetes. It does not show a clinical benefit.

Overall Confidence Assessment

• Cardiovascular mortality with regular natto intake: Consistent observational evidence from two large Japanese cohorts (29,079 and 92,915 participants, roughly 15 years of follow-up, hard endpoints), but limited to Japanese populations with lifelong consumption patterns and open to residual confounding.
• Blood pressure reduction from nattokinase: Moderate confidence — a modest effect (about 3 mmHg systolic) across several RCTs and a meta-analysis, with a biologically plausible mechanism. Not demonstrated to slow atherosclerotic plaque in low-risk adults; the one positive plaque study used a much higher dose and needs confirmation.
• Bone protection from MK-7: Moderate — two RCTs of MK-7 supplements show slower bone loss at some sites and better trabecular structure, but the trials are small, one found no difference in bone density, and no trial has shown fewer fractures. Observational data link regular natto eating with higher bone density and fewer fractures in Japanese women.
• Gut microbiome support: Preliminary — a single case study of a natto-powder product showing shifts in bacterial abundance, with no controlled trial of clinical outcomes.