The relative contribution of vitamin D receptor (VDR), collagen type 1, α-1 (COL1A1), tumor necrosis factor receptor 2 (TNFR2), polymorphisms, physical activity and bone mineral-free lean mass to bone parameters in children

• Main Author: Taylor, Ian Wesley
• Language: English
• Published: 2009
• Online Access: http://hdl.handle.net/2429/13993

Background: This study sought to investigate the relationship of physical activity (PA), dietary calcium and 3 candidate gene (VDR, COL1A1 and TNFR2) genotypes on bone mass in children (n = 327, age 10.33 ± 0.65). The study also sought to investigate the effect of PA and genotype on bone mineral-free lean mass (BMFL). Finally, the relationships between bone mass and BMFL and PA, BMFL and genotype and PA and genotype interactions were investigated. Methods: Anthropometric data (height, sitting height and weight) was determined using standard techniques. Bone mass, and BMFL were assed using total body DXA scan using a Hologic QDR 4500. Dietary calcium, PA and maturity were assessed using previously validated questionnaires. VDR FokI and VDR BsmI genotypes were determined by standard restriction fragment length polymorphism techniques. COL1A1 genotype was determined by a novel TaqMan technique and TNFR2 genotypes and haplotypes were determined by a novel automated sequenceing protocol. Associations between PA or candidate gene genotype and either BMFL or bone mass was first controlled for inter-subject differences in body size and maturity. Results: PA was significantly associated with BMFL in boys (p = 0.038). PA score was associated with a 3-5% difference in proximal femur BMC, femoral neck BMC and femoral neck aBMD but not lumbar spine BMC in boys as well as a 4% difference in femoral neck aBMD in girls. Average dietary calcium intake was not associated with differences in bone mass in children. VDR Bsml and VDR Fokl genotype did not have a relationship to bone mass or BMFL in children. COL1A1 Ss or ss genotype is associated with 4.8% higher femoral neck BMC in boys but not BMFL in either sex. TNFR2 A593G gg genotype was associated with a 3.8% higher BMFL in boys (p =0.038) and a 3.4% higher femoral neck BMC in boys (p = 0.045). Girls with a TNFR2 T598G tg genotype had a 3.3% higher femoral neck BMC (p = 0.029). TNFR2 G593-G598/G593-T598 haplotype was associated with a 10% higher femoral neck BMC in girls. For boys, when the BMFL by PA interaction term was added to the model it explained significantly more (2.5-3.9%, p = 0.004) of the variance in femoral neck BMC, femoral neck aBMD and lumbar spine aBMD than the main effect for PA alone. When the BMFL by VDR Fokl genotype interaction was added to the model the main effect of the VDR Fokl genotype became significant where boys with the FF genotype had a 1.4% greater femoral neck aBMD than boys with the Ff or ff genotype. For girls, significant interactions between TNFR2 haplotype and BMFL changed the model such that girls with the G593-G598/G593-T598 haplotype had a 10-11% greater femoral neck BMC or aBMD than girls with other TNFR2 haplotypes. Girls with the Ss or ss genotype had a 4% greater femoral neck aBMD after a significant interaction between COL1A1 genotype and PA was accounted for. Conclusions: High levels of PA are associated with increased BMFL and bone mass. TNFR2 genotypes are associated with both lean mass and bone mass in a complex fashion suggesting that the TNFR2 genotypes and interactions between BMFL and TNFR2 genotypes affect and moderate a combined lean mass/bone mass effect. COL1A1 Ss and ss genotypes are associated with high bone mass particularly in girls with high PA.