requires further investigation as significant inverse relationships have also been observed between BMD and total osteocalcin (Liu and Peacock, 1998; Ravn et al., 1996) and between BMD and the active (carboxylated) form of osteocalcin (Knapen et al., 1998).

Undercarboxylated osteocalcin has also been associated with increased risk of hip fracture. In a series of reports involving institutionalized elderly women studied for periods of up to 3 years, women with elevated ucOC at the start of the study had a three- to six-fold higher risk of suffering a hip fracture during the follow-up period (Szulc et al., 1993, 1996). It is of interest that in these studies the concentration of carboxylated osteocalcin, presumably the biologically active form, also was highest in the hip fracture group. Similar results subsequently were observed in a 22-month follow-up study involving a group of 359 independently living women (104 women having suffered a hip fracture and 255 controls) (Vergnaud et al., 1997). When the risk of hip fracture was related to levels of ucOC, increased baseline ucOC levels were associated with increased hip fracture risk with an odds ratio of 2. Although it is not possible to calculate carboxylated osteocalcin by quartiles from the data presented, this biologically active form of osteocalcin was not reduced in the hip fracture group. These studies are of interest with respect to a potential role of vitamin K in bone health, but they should be interpreted with caution given that in most cases they did not control for confounding factors such as overall quality of the diet or for nutrients known to influence bone metabolism (i.e., vitamin D and calcium). The increased concentration of circulating carboxylated osteocalcin in the fracture-prone population would also suggest that if vitamin K status has a role in bone health, it is not mediated through the action of osteocalcin.

Vitamin K intake has been associated with bone health in an epidemiological study. Utilizing the Nurse’s Health Study cohort, researchers found that vitamin K intakes were inversely related to the risk of hip fractures in a 10-year follow-up period (Feskanich et al., 1999). Vitamin K intakes of 71,327 women aged 38 to 63 years were assessed through the use of a food frequency questionnaire. Women in quintiles two through five of vitamin K intake had a lower age-adjusted relative risk of hip fracture (relative risk, 0.70; 95 percent confidence interval, 0.53–0.93) than women in the lowest quintile (vitamin K intake less than 109 μg/day). Risk did not decrease between quintiles two and five, a finding that should be explored further.

Intervention studies using different K vitamers in physiological and pharmacological dosages have also been performed. In a study