Vitamin D is transformed into its active hormonal conditions through hydroxylation of hormone precursors, e.g. into 25-hydroxyvitamin D in the liver and then into 1α, 25-dihydroxyvitamin D [1.25-(OH)2D] in the kidneys. This active hormonal state of vitamin D plays an essential role in bone homeostasis. Upon reaching the bones as one of its target tissue, several functions are discussed. The leading effect of the vitamin D hormone on bones is indirect and mediated through its endocrine function on mineral homeostasis . However, also binding its receptor (vitamin D receptor or VDR) a receptor-regulated pathway acts regulating transcription of target genes responsible for conducting the physiological process of 1.25-(OH)2D [24, 25].
The physiological level of 25(OH)D3 still remains unclear and ranges from 12 ng/ml to 40 ng/ml mean [3, 22]. The observed mean level of 25(OH)D3 in our study ranged from 16.3 ± 5.2 ng/ml to 18.3 ± 5.1 ng/ml and is similar to mean values found with other authors, describing the mean 14.7 ± 6.4 ng/ml .
Concerning fracture repair, the time course of 25(OH)D3 was examined in chicks over periods of maximal 21 days. Increasing levels of 25(OH)D3 were observed in the chicks due to an increase in CYP24A1 activity, whereas serum calcium levels revealed the same. Exposure of kidney cells in cultures to serum obtained from chicks with fractures showed an increase in CYP24A1 activity compared to control serum. This suggests that 25(OH)D3 is involved in the early stages of fracture repair and that there is some form of physiological communication between the fractured bone and the kidneys leading to an increase of renal 24-hydroxylase and the circulating concentration of this metabolite .
In mice CYP24a1 mRNA levels were significantly elevated in the fracture callus compared to the callus of the undamaged contralateral bone  and in CYP24a1-/- mice a delay in the mineralization of the cartilagous matrix of the soft callus and an altered expression of differentiation marker genes were found .
The presence of a nonnuclear membrane receptor for 25(OH)D3 in the fracture healing callus has been suggested [27, 28] as well as a receptor/binding protein for 25(OH)D3 in the fracture healing callus membrane fraction from 25(OH)D3 depleted chicks .
Serum 25(OH)D3 levels after sustaining a fracture have only been measured in few studies with humans [10–12]. Significantly reduced levels were found 28 hours after sustaining the fracture.
There are two interesting studies from Meller et al. in humans in 1982 and 1984 (11,12). In the first study, regarding 25(OH)D3, 13 young adults who had sustained fractures of a long bone, pelvis and vertebral bodies were included. Blood samples were taken within 12 hours after sustaining the fracture and also after six to eight weeks (with the sign of primary callus). During the healing period of the fractures no significant changes could be detected in the selected two pints of time. In the second study he investigated a population of 41 geriatric patients with 43 fractures of the proximal femur. 31 fractures were treated operatively while 12 non-operatively. Again blood samples were taken at two points in time, i.e. 48 hours after administration and after 8 weeks. Concerning the time course of 25(OH)D3 again no significant changes could be demonstrated.
Our study is the first study to examen the time course of 25(OH)D3 during human fracture healing in normal BMD level bone and low BMD level bone in a time course with blood samples taken at four different points in time. We could not confirm the increase of 25(OH)D3 during the healing period, which we had expected, but rather a slight decrease. This heterogeneity may be explained by the different study designs, meaning results of animal studies may be different than human situation. Comparing our results with the studies of Meller, it only remains for us to show these first results as a basis for further investigations.
In addition, no significant differences could be detected between the normal and low BMD level groups in our study with regard to levels of 25(OH)D3. This contradicts reports of low levels of 25(OH)D3 in osteoporotic patients in the literature so far [1, 29]. In fact, there are other authors, who failed to show any significant correlation between 25(OH)D3 and BMD, e.g. Garnero et al. in the OFELY (Os de Femmes de Lyon) study  and others [11, 12, 31, 32] after adjusting for age.
The observed mean level of PTH ranged in our study from 34.6 ± 10.9 pg/ml to 40.8 ± 23.3 pg/ml and is similar to the mean values found of other authors, describing the mean as 47.9 ± 30.4 ng/ml .
Potentially, PTH can exert anabolic and catabolic effects on osteoblasts by initiating different signalling cascades . To date, three receptors for PTH/PTHrP are known on osteoblasts (PTH1R-PTH3R)  inducing formation of cyclic 3’.5’-adenosine monophosphatase (cAMP) by activating the adenylate cyclase through stimulatory G-alpha proteins coupled to the receptor . Studies in vitro and in vivo have shown that the N-terminal 1-34 synthetic fragment of the 84 amino acids of PTH mediates full PTH activity , eliciting a cAMP response and stimulation of bone formation . Thus we propose activation of the bone anabolic effects of PTH especially in the first period after fracture, when most important stabilization processes occur.
In recent literature significant negative correlations between 25(OH)D3 and PTH have been described . These findings support concur with the results of our study, where an increase of PTH in first week after fracture healing is accompanied by a slight decrease of 25(OH)D3.
One lack of our study is of course a missing control group with matching demographics without fractures, since especially 25(OH)D3 levels are dependent on multiple effects like oral supply or exposure to the sun. We had to accept this problem in our study design, as we rarely see patients without fractures in our trauma department.