Influence of serum 25-hydroxyvitamin D levels, fat-free mass, and fat mass on bone density, geometry and strength, in healthy young and elderly adults.

Abstract

The association between serum 25-hydroxyvitamin D (25-OHD) levels and cortical/trabecular bone parameters has been explored in the elderly, but less so in younger adults; body composition may also influence bone parameters across the life span. We aimed to investigate, with peripheral quantitative computerized tomography (pQCT), the relationship between serum 25-OHD levels and bone geometry and strength and, at the same time, to explore the influence of fat mass and fat-free mass on bone parameters, for the tibia and radius, in healthy young and elderly adults. The study involved 149 healthy adults grouped by age: 65 were under 65 years old, and 84 were older. All participants were assessed in terms of: clinical history; serum 25-OHD levels; fat-free mass (FFM) and fat mass (FM), measured with DXA; total and cortical bone cross-sectional area (CSA, CSAc), and trabecular and cortical bone mineral density (BMDt, BMDc); and fracture load x and y for the tibia and radius, measured with pQCT. In the younger group, the association between 25-OHD levels and bone parameters did not remain as significant for any parameters after multivariate adjustment. In the elderly, 25-OHD correlated with CSAc (partial R2 = 0.33), fracture load x (partial R2 = 0.54), and fracture load y (partial R2 = 0.46) for the radius, and marginally with BMDt (partial R2 = 0.09; B-H adjusted p < 0.05 for all) for the tibia. FFM correlated with all bone parameters in both age groups. In the elderly group alone, FM correlated with BMDt at the tibia (r = 0.25, p < 0.05), with CSA at both sites (radius r = -0.25, p < 0.05; tibia r = -0.32, p < 0.001), and with fracture load y on the radius (r = -0.22, p < 0.05). While serum 25-OHD levels correlated only weakly with bone parameters in younger adults, a significant relationship was observed for elderly people. Fat-free mass showed positive simple correlation with pQCT-derived bone parameters in both age groups except with BMDc in the younger group. Further longitudinal studies are needed to clarify these relationships.