Interplay of physical activity and vitamin D receptor gene polymorphism on bone mineral density.

The interplay of physical activity and vitamin D receptor (VDR) gene polymorphism in their effects on bone mineral density (BMD) was studied for 120 Japanese girls aged 18-19 years. BMD at distal radius in the group with the VDR genotype at the Apal site of Aa was significantly higher than that in the aa group, but this association disappeared in a group having the habit of physical activity. The Aa genotype gave a higher BMD than the aa genotype only in the group without the habit of physical activity. The habit of physical activity was associated with a higher BMD only in the aa genotype group. The similar interplay was observed in the VDR genotype at the TaqI site. We thus suggest that physical activity and VDR genotypes affect BMD in independent mechanism to give a saturated level of BMD. Higher dietary calcium intake tended to be associated with higher BMD only in the Aa genotype, suggesting that the calcium intake and VDR genotype affect BMD in a synergistic mechanism.

The interplay of physical activity and vitamin D receptor (VDR) gene polymorphism in their effects on bone mineral density (BMD) was studied for 120 Japanese girls aged 18-19 years. BMD at distal radius in the group with the VDR genotype at the Apal site of Aa was significantly higher than that in the as group, but this association disappeared in a group having the habit of physical activity. The Aa genotype gave a higher BMD than the as genotype only in the group without the habit of physical activity. The habit of physical activity was associated with a higher BMD only in the as genotype group. The similar interplay was observed in the VDR genotype at the Taql site. We thus suggest that physical activity and VDR genotypes affect BMD in independent mechanism to give a saturated level of BMD. Higher dietary calcium intake tended to be associated with higher BMD only in the Aa genotype, suggesting that the calcium intake and VDR genotype affect BMD in a synergistic mechanism. J Epidemiol, The data were analyzed using the SPSS software package.
Analysis of variance was used to examine the mean difference of BMD, the calcium intake and the habitual physical activity between VDR genotypes. The association of the BMD with physical activity and dietary calcium intake in each VDR genotype was examined by ANCOVA.

RESULTS
In this study, we analyzed the effects of habitual physical activity on BMD by indexing the choice of sports ("Yes") or cultural activities ("No") for the supplemental class during the participant's high school. Since BMD is a body size dependent measure, we first compared the height, the weight-and the lean body mass between "Yes" and "No" subgroups to find no significant difference (Table 1). Same was the case for "High" and "Low" subgroups of dietary calcium intake scored through the habit of eating calcium-containing foods (Table 1). These demonstrated that the habitual physical activity and dietary calcium intake on BMD (described below) were not influenced indirectly by difference of body size.
When the VDR genotype was not considered, habitual physical activity showed only a trend of association of higher BMD at distal radius ("Total" in Table 2). When the participants were grouped into the VDR genotypes of Aa and aa, the effect of their physical activity became statistically significant. As we have reported previously 13), the BMD at distal radius was associated with the Apal site polymorphism of the VDR gene in the opposite direction to that reported by Morrison et al. 4). The effect of physical activity was significant in the unfavorable VDR genotype of aa, but not in the favorable genotype of Aa (Table 2). This implies that physical activity affects BMD only in the population with an unfavorable VDR genotype. Another important observation was that the genotype Aa gave a higher BMD than as only in the population with no habitual physical activity. These results show that physical activity and VDR genotypes affect the BMD independently but BMD reaches to a saturated level by either factor. Since the haplotype frequency of t/T was 0.09/0.91 in the Japanese population 13), the Tag1 site polymorphism gave uneven sample numbers for the tt, Tt and TT genotypes and was not statistically very strong. Nevertheless, the BMD of the genotype Tt was significantly higher than that of TT only in the population with no habitual physical activity. Here again, the effect of physical activity was significant only in the unfavorable VDR genotype of TT.
The dietary calcium intake scored through the habit of eating calcium-containing foods suggested a cooperative interplay with the VDR genotype (Table  3). High scores for calcium intake indicate a trend towards higher BMD regardless of the genotype, and this trend was clearer in the Aa genotype than in the aa genotype.

DISCUSSION
In a survey of 470 healthy premenopausal women aged 44-50, Salamone et al. found that the VDR genotype modifies the association between physical activity and BMD at least at the femoral neck 10). Within the population having the bb genotype, which resulted in lower BMD, higher physical activity was associated with a higher femoral neck BMD. The effect of physical activity was less pronounced for the BB and Bb genotypes. Their pattern was consistent with our results for Apal and TaqI sites, suggesting together that the VDR genotype affected the BMD (at least at the femoral neck and distal radius) in an independent mechanism from that of physical activity. A clearer demonstration of the differential effects of years a progressive exercise three times a week for a period of 18 months, but they did not observe any modification of the association between VDR genotype and BMD through physical activity 12). Apparently different results suggest that "habit" (over a long period) and "loading" (within a limited period) of physical activity need to be distinguished to evaluate its effect on BMD.
If we can assume that the effect of VDR genotype on BMD is due to the role of VDR in calcium metabolism, the enhanced osteocyte proliferation demonstrated by Chambers et al. by mechanical stimulation in rat tail vertebrae may provide an explanation for the differential effects of VDR genotype and physical activity 15). They showed that mechanical loading induces bone formation through the expression of immediate early genes such as c-fos and insulin-like growth factor I in otseocytes 16). This animal model system suggests that physical activity enhances "osteogenesis" but not "calcification" in the subsequent step of bone formation. This may explain why the effect of physical activity on BMD was clear in the population with unfavorable VDR genotypes of aa, TT and bb. In the favorable VDR genotypes, reduced osteogenesis due to low physical activity may be overcome by the enhancement of calcium metabolism. In the study by Rauch et al. 11), the impact of the VDR genotype was not clear probably because all of their study population consisted of physically active women.