Urinary bone resorption markers in patients with metabolic bone disorders.

Recently, urinary pyridinoline and deoxypyridinoline have been commonly employed as bone resorption markers. We studied these markers in 17 patients with hyperthyroidism, 15 undergoing long-term anticonvulsant drug therapy, and 28 with postmenopausal osteoporosis. Both markers had significantly higher levels than those in age-matched control groups. Values of urinary pyridinoline and deoxypyridinoline correlated well with urinary hydroxyproline levels in patients with hyperthyroidism (r = 0.856, p < 0.001 for pyridinoline and hydroxyproline; r = 0.919, p < 0.001 for deoxypyridinoline and hydroxyproline); however, poor correlations were observed, especially between urinary deoxypyridinoline and urinary hydroxyproline (r = 0.357, NS) in patients with postmenopausal osteoporosis. To compare the discriminatory ability of urinary pyridinoline and deoxypyridinoline, receiver operating characteristic (ROC) curves were generated for each of these patient groups using data from age-matched healthy females as the control group. The areas under the curves for both markers were 100.0% in hyperthyroidism. The areas under the curves for pyridinoline in patients undergoing long-term anticonvulsant drug therapy (mean +/- SE; 98.1 +/- 2.8%) and postmenopausal osteoporosis (77.9 +/- 5.7%) were significantly higher than those for deoxypyridinoline in anticonvulsant drug therapy (92.4 +/- 3.3%) and in osteoporosis (64.9 +/- 4.3%). Using data from premenopausal healthy females as the control group, areas under ROC curves for urinary pyridinoline (100.0%) and deoxypyridinoline (94.8 +/- 5.9%) were significantly higher than those for urinary hydroxyproline (73.8 +/- 9.4%) in patients undergoing long-term anticonvulsant drug therapy. In patients with postmenopausal osteoporosis, those for urinary pyridinoline (97.0 +/- 2.8%) were also significantly higher than those for urinary hydroxyproline (74.0 +/- 6.4%).(ABSTRACT TRUNCATED AT 250 WORDS)