Vitamin K1 for Osteoporosis

Vitamin K1 (phylloquinone) has been tested for bone in several double-blind, placebo-controlled trials, and the results are mostly null. In the ECKO trial, 440 postmenopausal women with osteopenia took 5 mg a day or placebo; over 2 years bone density fell by 1.28% versus 1.22% at the spine and 0.69% versus 0.88% at the hip, with no significant difference at any site over up to 4 years (Cheung 2008). Trials of 500 mcg a day in older men and women (Booth 2008), 1 mg a day in postmenopausal women (Binkley 2009) and 1 mg a day added to bisphosphonates in women with osteoporosis (Moore 2023) also found no difference in bone density.

Two pieces of evidence point the other way, and both are weaker. Fewer women on vitamin K1 had clinical fractures in ECKO (9 vs 20, p = 0.04), but the trial was not powered to examine fractures (Cheung 2008). And people who eat more vitamin K1 have had fewer fractures, including hip fractures, in several cohorts (Feskanich 1999; Booth 2000; Hao 2017), which cannot show whether vitamin K1 itself is the reason. The fracture results usually quoted for "vitamin K" come from trials of high-dose vitamin K2, a different question covered on our vitamin K2 for osteoporosis page.

Vitamin K is the cofactor for an enzyme that carboxylates several proteins involved in blood clotting and bone metabolism, including osteocalcin, one of the main proteins in bone (NIH ODS). When vitamin K is short, more osteocalcin circulates in its undercarboxylated form, which is why vitamin K deficiency has been proposed as a contributor to osteoporosis (NIH ODS).

Vitamin K1 is the main dietary form, found mostly in green leafy vegetables, while vitamin K2 (menaquinones) comes from fermented foods and animal products. Both phytonadione, the synthetic form of K1, and MK-7 supplements are well absorbed, but MK-7 has a longer half-life (NIH ODS).

Every vitamin K1 trial here changed the marker it was meant to change. ECKO raised serum vitamin K1 tenfold and lowered undercarboxylated osteocalcin (Cheung 2008); 200 mcg a day cut it by 51% (Bolton-Smith 2007); and 1 mg a day suppressed it by 81% in people with Crohn's disease (O'Connor 2014). Yet spine and hip bone density did not change in ECKO, in the 500 mcg trial (Booth 2008) or in the Crohn's trial, where only radius bone mass rose modestly (O'Connor 2014).

A better osteocalcin result shows the vitamin is being absorbed and used. It is not evidence of stronger bones, and the trials show the two can come apart.

The bone trials gave vitamin K1 at 200 mcg a day (Bolton-Smith 2007), 500 mcg (Booth 2008), 1 mg (Braam 2003; Binkley 2009; Moore 2023) and 5 mg (Cheung 2008). The adequate intake for adults is 90 mcg a day for women and 120 mcg for men (NIH ODS), so even the lowest trial dose was above it, and the 500 mcg trial's authors described their dose as attainable in the diet (Booth 2008).

Leafy greens are the richest source: half a cup of boiled frozen collards has about 530 mcg and a cup of raw spinach about 145 mcg. Vitamin K1 from vegetables is less well absorbed than from oils or tablets (the body takes up only 4% to 17% as much from spinach as from a tablet), and eating vegetables with some fat improves absorption (NIH ODS). Many of the trials also gave calcium and vitamin D to every participant (Booth 2008; Binkley 2009).

Guidelines for postmenopausal osteoporosis say fracture risk should be assessed to guide treatment, that nutrition, lifestyle and fall prevention should accompany any drug treatment, and that several medicines reduce fracture rates in women at risk (Eastell 2019). Our pages on vitamin D and calcium for osteoporosis cover what those two supplements can and can't do.

The women in ECKO were already vitamin D replete, and the trial had 90% power to detect a 3% difference in bone density. Neither its primary outcome, spine and hip bone density at 2 years, nor bone density at any site over up to 4 years differed between groups (Cheung 2008). In 452 men and women aged 60 to 80 who all took calcium and vitamin D, 500 mcg a day for 3 years made no difference to bone density at any site (Booth 2008).

In 381 healthy postmenopausal women who all took calcium and vitamin D, neither 1 mg of vitamin K1 nor 45 mg of vitamin K2 changed spine or hip bone density or bone turnover over 12 months (Binkley 2009). The only trial in women with diagnosed osteoporosis gave 105 women on oral bisphosphonates and low vitamin K1 levels 1 mg of K1, 45 mg of MK-4 or placebo for 18 months: bone density and turnover markers did not differ, and some hip geometry measures differed only in a per-protocol analysis (Moore 2023).

In 181 healthy women aged 50 to 60, a supplement of calcium, magnesium, zinc and vitamin D plus 1 mg vitamin K1 led to 1.7% less femoral neck bone loss than placebo over 3 years (95% CI 0.35-3.44) and 1.3% less than the same supplement without vitamin K1. Lumbar spine bone density did not differ (Braam 2003).

In 244 healthy Scottish women aged 60 or over, 200 mcg of vitamin K1 alone did not change bone density, but women taking it together with 400 IU vitamin D3 and 1,000 mg calcium had a sustained rise in bone density at the ultradistal radius, a wrist site, and not at the hip (Bolton-Smith 2007). In adults with Crohn's disease, 1 mg a day for 12 months modestly increased total radius bone mass but not spine or femur bone mass (O'Connor 2014).

Each benefit appeared at one of several sites measured, two of them only when vitamin K1 came with vitamin D and minerals, and none was in people with osteoporosis. They are smaller than, and have not been repeated by, the larger trials above that found no difference, which is why they do not change the verdict.

In ECKO, 9 women in the vitamin K1 group had a clinical fracture, against 20 on placebo (p = 0.04); the authors note the trial was not powered to examine fractures and the numbers were small (Cheung 2008). A UK health technology assessment converted this to a relative risk of 0.46 (95% CI 0.22-0.99), while noting that vertebral fractures were not reported, and judged that first-line alendronate prescribing was unlikely to change (Stevenson 2009).

An earlier review found that all seven trials with fracture data were Japanese trials of vitamin K2 (Cockayne 2006), mostly at 45 mg a day of MK-4 (NIH ODS), and the larger Osteoporosis Fracture study of vitamin K2 later found no reduction in vertebral fractures (Stevenson 2009). Earlier meta-analyses also reported less bone loss with vitamin K (Cockayne 2006; Fang 2012), though Fang's authors warned their modest result may be biased.

A 2019 update of 36 trials of all forms of vitamin K, done partly because the integrity of some earlier evidence had been questioned, found lower odds of clinical fracture in the trials of postmenopausal or osteoporotic patients (OR 0.72, 95% CI 0.55-0.95). The result lost significance when limited to trials at low risk of bias (OR 0.76, 95% CI 0.58-1.01), vertebral fractures did not differ (OR 0.96), and bone density differences shrank once high-risk trials were removed. The authors concluded there is no evidence that vitamin K affects bone density or vertebral fractures (Mott 2019).

In 72,327 women in the Nurses' Health Study, those above the lowest fifth of vitamin K intake (below 109 mcg a day) had a lower risk of hip fracture (RR 0.70, 95% CI 0.53-0.93), and the risk did not fall further at higher intakes (Feskanich 1999). In the Framingham cohort, the highest quarter of intake had fewer hip fractures than the lowest (RR 0.35, 95% CI 0.13-0.94), but vitamin K intake was not linked to bone density (Booth 2000). Among 1,373 Australian women aged 70 or over, intakes of 99 mcg a day or more were associated with fewer hip fracture hospitalisations than intakes below 61 mcg (HR 0.51, 95% CI 0.32-0.79) (Sim 2022).

A meta-analysis of five such studies found a modest link (RR 0.78, 95% CI 0.56-0.99) (Hao 2017), but a Danish cohort of 2,016 perimenopausal women found no association with bone density or fractures (Rejnmark 2006). Vitamin K1 intake largely reflects how many vegetables people eat, and the trials that supplemented it did not reproduce a bone benefit.

No tolerable upper intake level has been set for vitamin K because it has a low potential for toxicity (NIH ODS). In ECKO, 5 mg a day was well tolerated over 4 years, with no difference from placebo in adverse effects or quality of life (Cheung 2008; reviewed in Stevenson 2009).

The real risk is not toxicity but the interaction with warfarin and similar anticoagulants, which work by blocking vitamin K (NIH ODS). Vitamin K is in most multivitamins, and products that combine it with calcium, magnesium or vitamin D vary more in dose, some providing 4,050 mcg, so check the labels of bone formulas too (NIH ODS). People with malabsorption conditions such as cystic fibrosis, celiac disease or ulcerative colitis might not absorb vitamin K properly (NIH ODS). The bone trials enrolled adults, mostly postmenopausal women and older adults, so they say nothing about high-dose supplements in pregnancy or in children.

Vitamin K antagonists such as warfarin deplete vitamin K-dependent clotting factors, so people taking them need a consistent vitamin K intake (NIH ODS). In one trial, 200 patients on phenprocoumon took 100 mcg of vitamin K or placebo daily for 24 weeks under clinic supervision with dose adjustment; time in the target range was similar (89.5% vs 85.5%), with more patients reaching maximal stability (Rombouts 2007). That was a supervised dose in an anticoagulation clinic, not a reason to start vitamin K on your own.

Antibiotics can lower vitamin K status by killing the gut bacteria that make it, possibly more so with cephalosporins such as cefoperazone. Bile acid sequestrants such as cholestyramine can reduce its absorption, though the clinical significance is unclear, and the weight-loss drug orlistat reduces absorption of fat-soluble vitamins; combined with warfarin it might significantly raise prothrombin time (NIH ODS). In women with osteoporosis taking bisphosphonates, adding vitamin K1 did not change bone density (Moore 2023).