Introduction: This study aimed to establish sex- and age-specific reference curves enabling the calculation of z-scores and to examine correlations between bone markers and anthropometric data.

Methods: Morning blood samples were obtained from 572 healthy children and adolescents (300 boys) aged 2 months to 18 yr. Height, weight and pubertal stage were recorded. Serum osteocalcin (OC), bone-specific alkaline phosphatase (BALP), type-1 collagen degradation markers (ICTP, CTX) and tartrate-resistant acid phosphatase (TRAP5b) were measured. Cross-sectional centile charts were created for the 3^rd, 50^th and 97^th centiles.

Results: Apart from TRAP5b all bone markers were non-normally distributed, requiring ln (BALP, OC, ICTP) or sqrt (CTX) transformation. Back-transformed centile curves for age and sex are presented for practical use. All bone markers varied with age and pubertal stage (P < 0.001). Significant correlations were found between SDS for bone formation markers BALP and OC (r=0.13; P = 0.004), SDS for collagen degradation markers ICTP and CTX (r=0.14; P = 0.002) and SDS for the phosphatases (r=0.34, P < 0.001). Height and weight SDS correlated weakly with some bone marker SDS, particularly with lnBALP SDS (r=0.20 and 0.24, respectively, both P < 0.001).

Conclusion: This study provides reference curves for OC, BALP, CTX, ICTP and TRAP5b in healthy children. Taller and heavier individuals for age had greater bone marker concentrations, likely reflecting greater growth velocity. SDS for markers of bone formation, collagen degradation and phosphatases were each independently correlated, suggesting they derive from the same biological processes. The possibility of calculating SD-scores will facilitate monitoring of antiresorptive therapy or disease progression in children with metabolic bone disease.