Urinary collagen cross-links as biochemical markers of growth: an evaluation of biological variables.

Aim: To investigate the validity of urinary pyridinium cross-links (pyridinoline and deoxypyridinoline) as markers of growth in healthy children.

Methods: Three pilot studies (P1-P3) were conducted to investigate the time of day, the minimal duration within a day, and how many times per week urine samples needed to be collected to obtain representative values of cross-link excretion in normal children 3-5 years of age. The results were used to design a 4-month longitudinal protocol to evaluate whether pyridinium cross-links could be used as markers of growth velocity.

Results: Mean differences from 24-hour values were only between 1 and 4% for urinary cross-links (nmol/h) in overnight 12-hour collections. Three consecutive collections were required for weekly output estimates with a maximum error of 10% in >90% of the children. During the 4-month longitudinal study, the regression equation of height velocity on pyridinoline and deoxypyridinoline excretion explained approximately 60% of the variance in the subgroup of subjects who provided three complete urinary collections per observation period. No relationship was observed when the cases with fewer or incomplete collections were included in the analysis. Cross-link values collected at baseline were of no use to predict height velocity at 4 months.

Conclusions: Urinary pyridinium cross-links correlate with the growth velocity in healthy children when using an appropriate urinary collection protocol. However, their predictive value in this population is negligible.