BACKGROUND: Bone remodeling takes place in the bone marrow environment. We investigated if levels of bone turnover markers (BTMs) differ between bone marrow and peripheral blood, if the bone marrow is an independent compartment, and how well the measurements in bone marrow correlate with bone mineral density.

METHODS: Sixty-six men participated in a placebo controlled study designed to evaluate the effect of 16 weeks supplementation with resveratrol on bone mineral density and BTM. Bone marrow aspirates and blood samples were drawn at baseline and at week 16. Procollagen I N-terminal propeptide (PINP), C-terminal telopeptide of type I collagen (CTx), osteocalcin, bone specific alkaline phosphatase (BAP), and osteoprogeterin (OPG) were analyzed. Bland-Altman analysis was used to compare measurements across compartment to detect possible systematic or proportional differences. Paired t-test was performed if no proportional difference was revealed at the difference vs concentration plot.

RESULTS: Measurements of PINP, CTx, and BAP differed proportionally between compartments depending on concentration; at low concentrations absolute values were only slightly different, while at higher average concentrations the levels were much higher in bone marrow than blood. Osteocalcin measures in bone marrow were systematically and significantly lower than in blood (mean ± SD; 14.4 μg/L ± 5.3 μg/L versus 21.7 μg/L ± 6.0 μg/L respectively, p < 0.001). OPG measures were comparable between compartments (p = 0.69). CTx and OPG measured in blood were negatively associated with lumbar spine BMD (β = -0.22, p = 0.05 and β = -0.02, p = 0.02, respectively), whereas both markers measured in bone marrow were not (p = 0.60 and p = 0.50 respectively). None of the BTMs, bone marrow or blood, were associated with total hip BMD.

DISCUSSION: The levels of most BTMs differed significantly between bone marrow and peripheral blood, while OPG was comparable. Levels of PINP, CTx, and BAP differed between compartments depending on concentration, suggesting bone marrow to represent a compartment separate from the general circulation. Unexpectedly, osteocalcin was lower in the marrow, a gradient that was independent of concentration. BTMs measured in bone marrow did not show any association with bone mineral density. Although further studies are needed to investigate potential explanatory causes of the differences, BTMs in bone marrow do not seem to contribute further to fracture risk assessment.