Biochemical markers of bone formation reflect endosteal bone loss in elderly men--MINOS study.

In the skeleton of elderly men, two opposite activities occur: bone loss at the endosteal envelope, which increases bone fragility, and periosteal apposition, which improves bending strength of bone. Both may contribute to serum bone formation markers although they have an opposite effect on bone fragility. The aim of this study was to determine if circulating bone formation markers reflect periosteal bone formation and endosteal bone remodelling in 640 men aged 55-85 years belonging to the MINOS cohort. We measured biochemical markers of bone formation (osteocalcin, bone alkaline phosphatase, N-terminal extension propeptide of type I collagen) and bone resorption (urinary and serum beta-isomerised C-terminal telopeptide of collagen type I, total and free deoxypyridinoline). Parameters of bone size (cross-sectional surface of third lumbar vertebral body measured by X-ray, projected areas of total hip, femoral neck, radius and ulna measured by dual-energy X-ray absorptiometry) increased with age (r = 0.20-0.32, P < 0.0001). In contrast, parameters related to bone loss (areal bone mineral density [aBMD], volumetric bone mineral density [vBMD] and cortical thickness) and determined mainly by bone resorption, decreased with ageing (r = -0.14 to -0.23, P < 0.005-0.0001). Men in the highest quartile of bone resorption markers had lower aBMD (3.8-10.2%, P < 0.05-0.0001), lower vBMD (3.9-13.0%, P < 0.05-0.0001), and lower cortical thickness (1.5-9.6%, P < 0.05-0.0001) than men in the lowest quartile. Markers of bone resorption were not significantly associated with estimates of bone size at any skeletal site. Markers of bone formation were not associated with estimates of periosteal formation after adjustment for covariates. In contrast, men in the highest quartile of the bone formation markers had significantly lower aBMD (4.0-11.7%, P < 0.05-0.0001), lower vBMD (4.2-16.3, P < 0.05-0.0001) and lower cortical thickness (4.0-7.4%, P < 0.05-0.0001) than men in the lowest quartile. In summary, serum levels of bone formation markers are negatively correlated with the estimates of endosteal bone loss. In contrast, they disclose no association with parameters reflecting periosteal apposition. Thus, in elderly men, bone formation markers reflect endosteal bone remodelling, probably because of the coupling between resorption and formation activities. In contrast, they do not reflect the periosteal bone formation, probably because the periosteal surface is smaller and has a slower remodelling rate than the endosteal surface.