Vitamin K and vitamin K–dependent proteins (also known as Gla proteins) are best known for their role in blood coagulation. However, Gla proteins are also involved in other physiologic processes (e.g. bone metabolism) (1,2). Osteocalcin is a bone Gla protein and synthesized by osteoblasts (3). It constitutes approximately 20% of the noncollagenous protein found in human bone (4).

Vitamin K acts as a cofactor in the posttranslational carboxylation of all Gla proteins. In this process, glutamate residues are converted into γ-carboxyglutamate (Gla) (5). Gla residues have a high affinity for calcium, and their complex formation with calcium ions was demonstrated to be essential for the function of all Gla proteins currently known.

To adequately carboxylate osteocalcin, the osteoblast requires sufficient vitamin K (6). In case of vitamin K deficiency, ucOC will be produced. Bioavailable vitamin K is mainly derived from nutritional sources. The gut flora also produces vitamin K, but at the site of synthesis (colon) its absorption is negligible (7,8). The two most important forms of vitamin K are phylloquinone (vitamin K1) and the group of K2 vitamins (menaquinones). Dietary sources of phylloquinone are green vegetables (spinach, broccoli) and some plant oils (9). Menaquinones are found in meat and fermented foods like cheese (10). Hence, inadequate intake of vitamin K will lead to the production of undercarboxylated (i.e. inactive) Gla proteins.

Methods have been developed to distinguish between cOC and ucOC fractions of osteocalcin (11,12). The ratio between ucOC and cOC (UCR) as well as circulating ucOC levels are used as indicators for the vitamin K status of bone (2,13,14). The UCR is probably the most sensitive marker for bone vitamin K status (2). Serum vitamin K concentrations fluctuate with recent dietary vitamin K intake and are not reliable markers for tissue vitamin K status. In the healthy adult population, osteocalcin is carboxylated to a variable extent, suggesting that the dietary vitamin K intake is insufficient for full osteocalcin carboxylation (15). Markedly higher osteocalcin carboxylation is obtained by increased vitamin K intake (16).

The benefits of vitamin K on bone metabolism and the prevention of fractures are well described in the adult population (1721). Little is known about the vitamin K status in growing bone such as in children. Both the metabolic activity and osteocalcin production (22) are high during bone development and skeletal growth, suggesting that also the vitamin K requirement of bone is high. During recent years, the dietary intake of vitamin K in children (23) has declined gradually. The combination of high requirement and low intake may lead to subclinical vitamin K deficiency in the rapidly growing bone. In a cohort of in 245 healthy girls (ages, 3–16 y), Kalkwarf et al. (24) recently reported a marked variation in the percentage of ucOC. In this study, better vitamin K status (expressed as %ucOC) was associated with decreased bone turnover. Furthermore, no evident correlations were found between %ucOC and dietary intake or serum levels of vitamin K1. This is consistent with the view that serum vitamin K is a marker for recent dietary vitamin K intake but not for tissue vitamin K status.

The aim of this study was to compare the vitamin K status of bone in healthy children and adults in a cross-sectional study design. In addition, we studied possible correlations of vitamin K status with the level of bone turnover markers in children (high bone metabolic activity) and in adults (stable bone metabolic activity).


Study subjects.

From October 2003 through January 2004, 86 healthy children (male and female) between 3 and 18 y of age were recruited from a population of children undergoing minor surgery (e.g. strabismus correction, phimosis correction), and friends or relatives of visiting patients in the Wilhelmina Children's Hospital. The Department of Biochemistry, University of Maastricht, recruited 30 healthy adult volunteers (25–35 y of age) via small ads. Subjects were included if they were within normal ranges for body weight and height according to reference growth charts. Exclusion criteria were current or previous medical diseases such as metabolic, gastrointestinal, or endocrine disease; chronic degenerative diseases; current medication such as anticoagulants and corticosteroids; and vitamin concentrates or food supplements.

Data collection procedure.

This study was approved by the Central Committee on Research Involving Human Subjects (CCMO, The Netherlands). Written informed consent was obtained from all adults, the parents of all children, and also from children when 12 y of age or older.

General data elements.

Body height and weight of all subjects were measured in a standardized manner without shoes and heavy clothing, to the nearest centimeter and 100 g, respectively. From these values, the body mass index (BMI) was calculated. Information about medical history and current use of any medication was ascertained by a short interview. In children, the pubertal stage was determined by a short list of questions about the presence of menarche and breast development (in females) or genital development (in males), according to Tanner's sexual maturity scale (SMS). The groups were divided into three categories: prepuberty (prepubertal stage), puberty (pubertal stages II–IV), and end of puberty (pubertal stage V).

Assessment of vitamin K status and bone turnover markers.

After blood sampling, all samples were frozen and kept at −80°C until use.

In serum, two biochemical markers of bone turnover were measured. Bone alkaline phosphatase (BAP) (Hybritech, Belgium) was measured as marker of bone formation by radio immunoassay. As a marker of bone resorption, N-telopeptide cross-links of collagen breakdown (NTX) (Ostex, Washington) were determined. Undercarboxylated and carboxylated fractions of osteocalcin (Takara, Japan) were used as indicators of vitamin K status and measured by enzyme-linked immunosorbent assay. The UCR is a sensitive indicator of the vitamin K status of bone (2,13). Elevated levels of UCR are indicative of an inferior vitamin K status of bone.

Statistical analysis.

Normality of distribution for all subjects was checked by histograms for all study parameters. Independent t tests and analysis of variance were used for comparison of parameters between children and adults. A χ2 test was performed for the differences in gender between adults and children. Pearson's correlation tests were performed to investigate possible correlations between demographic variables, bone markers, and vitamin K parameters in all groups. Stepwise multivariate linear regression analysis for the difference between children and adults was performed, using ucOC, cOC, and UCR dependent variables with adjustments for the confounders age, gender, puberty, height, and weight. The statistical tests were performed using a two-sided significance level of 5%. A p value <0.05 was considered to be statistically significant. SPSS Base 12.0.2 for Windows (SPSS Inc, Chicago, IL) was used for all analyses.



Eighty-six children (37 males, 49 females) and 30 adults (14 males, 16 females) were included in the study. Baseline characteristics of the subjects are shown in Table 1. The groups of adults and children differed in age, height, and weight, as might be expected. In the children's group, boys and girls did not differ in age (mean: 10.06 versus 10.01 y; p = 0.958), weight (mean: 40.8 versus 40.2 kg; p = 0.889), height (mean: 155.6 versus 150.8 cm; p = 0.324), and pubertal stage (p = 0.103). Furthermore, no differences in these variables between male and female in adults were present.

Table 1 Demographic and anthropometric data, bone markers, and vitamin K parameters of children and adults

Bone turnover parameters.

The levels of BAP and NTX were elevated in children compared with adults (Table 1). This is consistent with the higher metabolic activity in growing bone in children. No differences in bone markers were found between boys and girls (data not shown). Both BAP and NTX significantly increased during puberty and declined by the end of puberty (data not shown).

Vitamin K status of bone.

The levels of ucOC and cOC were elevated in children as well (Table 1). In addition, the UCR in children was significantly higher than in adults, suggesting a poor vitamin K status of growing bone. Figure 1 shows the variation of the ucOC and cOC levels in children compared with the adult group. No differences in vitamin K status were found between boys and girls: ucOC (mean: 36.5 versus 27.4 ng/mL; p = 0.082), cOC (mean: 16.9 versus 14.3 ng/mL; p = 0.110), and the UCR (mean: 2.46 versus 2.11; p = 0.445).

Figure 1
figure 1

Scatterplot of ucOC (A) and cOC (B) in children (n = 86) and adults (n = 30). The p value for the difference in the level of ucOC (t test) in children vs adults is <0.001 (A). The p value for the difference in the level of cOC (t test) in children vs adults is <0.001 (B).

To investigate whether this apparent vitamin K deficiency was associated with certain stages of development, we subdivided the children according to age and pubertal stage (Table 2). In all age groups, the UCR was above the adult level. A similar conclusion was obtained after subdivision according to pubertal stage. ucOC and cOC were highest during puberty and declined at the end of puberty. In postpubertal children, the cOC levels had almost reached adults values. Remarkably, the UCR remained high during all stages. Multiple linear regression analysis with adjustment for age, gender, and pubertal stage showed that ucOC, cOC, and the UCR in children were significantly higher than in adults (Table 3). When height and weight were also introduced in the equation, only ucOC and the UCR remained significantly different from the adult levels (Table 3).

Table 2 Vitamin K parameters in children of different age groups and pubertal stages
Table 3 Differences in vitamin K status between adults and children (multivariate analysis)

Correlation of vitamin K status and bone turnover parameters.

In the total group of children, a marked correlation was found between the markers for bone metabolism (BAP and NTX) and both ucOC and cOC. The bone markers did not correlate with the UCR (Table 4). In adults, NTX, but not BAP, was correlated with the osteocalcin markers. In the total group (including adults and children), both BAP and NTX showed significant correlations with ucOC, cOC, and the UCR.

Table 4 Pearson correlation and significance level of relation between bone metabolic activity and the vitamin K status in different groups


In this article, we report that high circulating levels of ucOC are found in the majority of children between 6 and 18 y of age. Moreover, the UCR was significantly higher in children than in adults. These findings suggest a poor vitamin K status of bone during growth. The extent of vitamin K deficiency shows a substantially interindividual difference when considering the large variations in the UCR. In the elderly, high serum levels of ucOC have repeatedly been mentioned to be associated with low bone mass and high fracture risk. Whether children would benefit from increased dietary vitamin intake remains to be seen.

Osteocalcin is a vitamin K–dependent protein almost exclusively synthesized by the osteoblasts in bone (4,25). On a molecular level, its function is still obscure, but the circulating antigen is frequently used as a marker for osteoblast activity. One of the characteristics of vitamin K–dependent proteins is that during episodes of low vitamin K intake, they are synthesized as undercarboxylated species. Osteocalcin is the only Gla protein known of which a relatively large fraction (20%–30%) circulates in an undercarboxylated form in healthy adults (15). The major part of newly synthesized osteocalcin is absorbed to the hydroxyapatite in bone, but a part of it leaks into the circulation where it can be detected (26).

The elevated levels of osteocalcin, BAP, and NTX in children, as found in the current study, are indicative of high metabolic activity in bone. Moreover, the high levels of ucOC suggest that the vitamin K requirement of bone during growth is not met by the dietary vitamin K supply. It may be argued that high ucOC levels are representative for the high bone turnover state found in adolescence rather than the result of vitamin K deficiency. Indeed, the serum ucOC level is determined by both the total osteocalcin production and the fraction thereof that has been carboxylated by vitamin K action. Therefore, the UCR (or total OC), which is independent of bone turnover, is a more reliable marker for vitamin K status of the bone. In our study, we found very high UCR values in children compared with adults, suggesting subclinical vitamin K deficiency during high bone turnover in children. This is consistent with observations in another population characterized by high bone metabolism, i.e. postmenopausal women (19,21,27,28). Several authors have demonstrated that, in the latter group, low dietary vitamin K intake, high circulating ucOC, and high UCR are associated with low bone mineral density and increased fracture risk. In these women, the serum ucOC levels are higher than in young adults.

From our study, it is clear that both the absolute concentration of ucOC and the UCR in children are much higher than in healthy adults or even in postmenopausal women. Here we postulate that the metabolic activity of bone is associated with its vitamin K requirement. Children during growth have a much higher bone metabolism than adults. This may be a possible explanation for the apparent vitamin K deficiency in children of all ages.

In children, the importance of vitamin K status in bone health has only been studied by a small number of research workers (29). Kalkwarf et al. (24) described that in healthy girls (3–16 y), greater vitamin K status was associated with decreased bone metabolic activity. Furthermore, in children with long-standing vitamin K deficiency caused by the anticoagulation drug warfarin, a reduced bone density was found (30). This study illustrates the potential consequences of vitamin K deficiency in growing bone.

However, the question still remains whether improved bone health in children can be achieved by higher vitamin K intake. The value of vitamin K supplements has been shown in several studies in postmenopausal women. In this group, vitamin K supplementation results in decreased bone loss (31) or even gain of bone mass (3234). Besides an effect on bone mineral density, recent data suggest that the most important benefit of vitamin K in bone is improved geometry. In an article that has only been published as an abstract thus far, Kaptoge et al. showed that in elderly women vitamin K supplementation resulted in increased femur neck width and increased bone strength at this site (Kaptoge S, Dalzell N, Welch A, Shearer MJ, Khaw KT, Reeve J 2005 Vitamin K and fracture risk: an effect on bone not BMD? Bone and Tooth Society Annual Meeting, July 4–5, 2005, Birmingham, UK, Abstract No. 12). Recently, this was confirmed in galactosemic children (B. Panis et al., unpublished results), a group known to have low BMD.

Our report describes the finding of a possible vitamin K deficiency in bone. The major limitation of this study is that we were only able to link these results to markers of bone metabolism and not to assessment of bone mass or strength. The confirmation that this apparent deficiency may affect bone quality can only be established in prospective pediatric studies. Another limitation concerns the cross-sectional design of our study. Prospective studies in which possible changes in UCR are monitored in a time-dependent manner may demonstrate whether vitamin K status is fluctuating during different episodes of growth or whether it is consistently associated with young age.

Our finding that, based on osteocalcin carboxylation, most of the children in a healthy population must be defined as vitamin K–deficient is a point of interest and justifies follow-up studies in which bone quality is monitored as a function of vitamin K–intake.