Vitamin K1 for Bone Mineral Density

Vitamin K1 (phylloquinone) is the form found mainly in leafy greens and vegetable oils (NIH ODS). For bone mineral density (BMD, the mineral content of bone measured by a DXA scan), the largest trials say no: 3 years of 500 micrograms a day in 452 adults aged 60-80 (Booth 2008), up to 4 years of 5 mg a day in 440 women with osteopenia (Cheung 2008), and 12 months of 1 mg a day in 381 postmenopausal women randomised to K1, MK-4 or placebo (Binkley 2009) all found no difference from placebo at the spine or hip.

Two smaller trials found a gain at one skeletal site when K1 was given with other nutrients: less femoral neck loss with K1, minerals and vitamin D in 181 women, 155 of whom completed (Braam 2003), and a gain at the ultradistal radius, a wrist site, with K1, vitamin D and calcium in 244 older women (Bolton-Smith 2007). A third trial, in adults with Crohn's disease, found a modest gain at the total radius but not the spine or femur (O'Connor 2014). A 2019 pooled review that mixed K1 with vitamin K2 found no evidence of an effect on BMD in postmenopausal or osteoporotic people (Mott 2019). For fractures and osteoporosis itself, see our page on vitamin K1 for osteoporosis.

Vitamin K is a cofactor for carboxylating osteocalcin, one of the main proteins in bone, and some research links high levels of undercarboxylated osteocalcin (ucOC, the form without that activation) to lower bone density (NIH ODS). K1 supplements clearly act on this step. In the ECKO trial, 5 mg a day raised serum vitamin K1 about 10-fold and lowered the percentage of ucOC (Cheung 2008). In the Scottish trial, 200 micrograms a day lowered it by 51% over 2 years (Bolton-Smith 2007).

In adults with Crohn's disease, 1,000 micrograms of K1 a day lowered ucOC by 81% after 2 weeks, and 2,000 micrograms lowered it no further, yet 12 months of 1,000 micrograms did not change lumbar spine or femur density (O'Connor 2014). A 3-year trial of 375 micrograms of the K2 form MK-7 cut ucOC by 65% after one year, with no difference in bone density or microarchitecture at 3 years (Rønn 2021). A 2024 pooled analysis, which found higher lumbar spine BMD (especially with K2), concluded that vitamin K's beneficial effects on bone mainly involve osteocalcin carboxylation rather than total osteocalcin (Xie 2024).

A marker that moves shows the vitamin reached its target. It does not show denser or stronger bone. NIH ODS says it is not clear whether vitamin K supplementation affects the risk of osteoporosis.

Trials gave 200 micrograms (Bolton-Smith 2007), 500 micrograms (Booth 2008), 600 micrograms (Volpe 2008), 1 mg (Braam 2003; Binkley 2009; O'Connor 2014; Moore 2023) or 5 mg of K1 a day (Cheung 2008). For comparison, the Adequate Intake is 90 micrograms a day for adult women and 120 micrograms for adult men, and NIH ODS reports average US food intakes of 122 micrograms for women and 138 micrograms for men. Because no dose raised spine or hip density in the large trials, a bigger dose is not a better one.

Phylloquinone from plant foods is less bioavailable than that from oils or supplements, and eating vegetables with fat improves absorption (NIH ODS). Some supplements contain very high amounts, up to 4,050 micrograms, so compare the number on the label with the doses above.

The widest trial randomised 452 men and women aged 60-80 to a multivitamin with 500 micrograms of K1 or without it, both with 600 mg of calcium and 400 IU of vitamin D, for 3 years. Femoral neck, spine and total-body BMD did not differ between groups, although the K1 group had higher blood vitamin K1 and a lower percentage of ucOC (Booth 2008).

The ECKO trial gave 440 postmenopausal women with osteopenia 5 mg of K1 or placebo for 2 years, extended for earlier participants up to 4 years. It had 90% power to detect a 3% difference in BMD. Lumbar spine BMD fell 1.28% with K1 and 1.22% with placebo (difference -0.06%, 95% CI -0.67% to 0.54%), and total hip BMD fell 0.69% and 0.88% (difference 0.19%, 95% CI -0.37% to 0.75%). There was no difference at any site over 2 to 4 years (Cheung 2008).

In 381 postmenopausal North American women, 1 mg of K1 or 45 mg of MK-4 for 12 months, with calcium and vitamin D3 for everyone, lowered ucOC but did not change bone turnover markers, spine or proximal femur BMD, or hip geometry (Binkley 2009).

In 181 healthy postmenopausal women aged 50-60 (155 completed), 1 mg of K1 added to a supplement of calcium, magnesium, zinc and vitamin D reduced femoral neck bone loss over 3 years. The difference was 1.7% (95% CI 0.35-3.44) against placebo and 1.3% (95% CI 0.10-3.41) against the supplement without K1, with no difference at the lumbar spine (Braam 2003). Because the benefit appeared when K1 was combined with minerals and vitamin D, it cannot be credited to K1 alone.

In 244 healthy women aged 60 and over, 200 micrograms of K1 with vitamin D3 and calcium increased BMD and bone mineral content at the ultradistal radius over 2 years, but not at hip or other radius sites. The authors called it a modest synergy and suggested the extra osteocalcin carboxylation might be reached through diet rather than supplements (Bolton-Smith 2007).

Three other trials were null for the spine and hip. In 86 adults with Crohn's disease, 1,000 micrograms for 12 months raised total-radius bone mass modestly (P below 0.05) but did not change spine or femur density (O'Connor 2014). In 105 women with osteoporosis taking a bisphosphonate, randomised to K1, MK-4 or placebo, 1 mg of K1 for 18 months did not change total hip, femoral neck or lumbar spine BMD, and the modest hip geometry differences appeared only in a per-protocol analysis (Moore 2023). In 14 pre- and perimenopausal women, 600 micrograms for 6 months did not improve regional BMD (Volpe 2008).

A 2019 review of 36 studies concluded that, for postmenopausal or osteoporotic people, there is no evidence that vitamin K affects BMD or vertebral fractures; it may reduce clinical fractures, but the evidence is insufficient to confirm this. In the main analysis the lumbar spine difference was significant at 2 years (+1.63%, 95% CI 0.10-3.16) but not at 1 year (+0.93%, 95% CI -0.02 to 1.89), and 6-month hip (+0.42%) and femur (+0.29%) differences were significant; the lumbar differences shrank and lost significance when high-risk-of-bias trials were removed (Mott 2019). Its authors later adjusted their estimates after concerns about the integrity of one included trial, the Yamaguchi Osteoporosis Prevention Study (Mott 2020).

Earlier and other pooled analyses lean the other way. A 2012 analysis found a lumbar spine gain of 1.27% (95% CI 0.47-2.06) but no femoral neck effect (0.17, 95% CI -0.21 to 0.54), and its authors warned the modest effects may be biased (Fang 2012). A 2022 analysis found a non-significant effect on femoral BMD (p = 0.08), and a 2024 analysis found higher lumbar spine BMD (p = 0.035), especially with vitamin K2 (Salma 2022; Xie 2024). The 2006 review reported less bone loss in 12 of 13 trials, but all 7 trials with fracture data were Japanese and used menaquinone-4 (Cockayne 2006). These pooled signals come largely from K2 and older, smaller trials, not from the large K1 trials above.

K2 results are also mixed. A 3-year trial of 180 micrograms of MK-7 in 244 healthy postmenopausal women reduced the age-related decline in BMD at the lumbar spine and femoral neck, but not the total hip (Knapen 2013). A 1-year trial of 360 micrograms in 334 early postmenopausal Norwegian women (Emaus 2010) and a 3-year trial of 375 micrograms in 142 women with osteopenia (Rønn 2021) found no difference from placebo. NIH ODS notes that a 45 mg dose of MK-4 is used as an osteoporosis treatment in Japan and other parts of Asia, that the European Food Safety Authority has approved a claim linking dietary vitamin K to maintenance of normal bone, and that the US FDA has not authorised a health claim. See our vitamin K2 page for the K2 evidence.

The trials enrolled postmenopausal women, older men and women, adults with Crohn's disease and a small group of premenopausal women; none was in men alone or in children. In 245 healthy girls, vitamin K status was not consistently associated with bone mineral content or its gain over 4 years (Kalkwarf 2004).

Everyone in the Booth and Binkley trials took calcium and vitamin D, and the ECKO women were vitamin D replete, so these trials show that K1 added nothing on top of an adequate vitamin D status, not what K1 does when that is lacking. The Booth authors concluded that a dietary dose of phylloquinone gives no extra bone benefit at the spine or hip when taken with recommended calcium and vitamin D. Even in people with likely low vitamin K status, such as the Crohn's trial and the 2023 trial (serum K1 of 0.4 micrograms/L or less), spine and hip BMD did not improve.

Vitamin K has no Tolerable Upper Intake Level because of its low potential for toxicity, and the National Academies stated that no adverse effects of vitamin K from food or supplements have been reported in humans or animals (NIH ODS). In ECKO, 5 mg of K1 a day for up to 4 years was well tolerated, with no significant differences from placebo in adverse effects or quality of life (Cheung 2008).

The practical risks are an interaction with warfarin and the temptation to let a supplement stand in for treatment. The Bone Health and Osteoporosis Foundation pairs medicine for people at high fracture risk with calcium, vitamin D, exercise, avoiding smoking and excessive alcohol, and fall prevention; vitamin K is not among its recommendations (LeBoff 2022). The Adequate Intake in pregnancy and lactation is 90 micrograms a day for adults and 75 micrograms for ages 14-18 (NIH ODS); none of these trials tested K1 for bone density in pregnancy.

Warfarin is the interaction that matters. Several other medicines change how much vitamin K your body absorbs or makes, and NIH ODS advises anyone taking them regularly to discuss vitamin K status with a clinician. The 2023 trial gave K1 to women already taking an oral bisphosphonate with calcium and/or vitamin D and found no difference in BMD from placebo (Moore 2023).