Laboratory screening of body fluids for products of bone turnover aids prediction and therapy of osteoporosis. New recent findings pointed to a close correlation between the content of proline-hydroxyproline dipeptide (PHP) in non-hydrolyzed urine and that of hydroxyproline (HP) after urine hydrolysis.The objective of our study was to evaluate the use of PHP as a new biomarker in osteopathies. Current urinary resorption markers (NTx, DPD, HP) were assayed together with PHP in both morning and all-day urine samples of 40 postmenopausal female subjects in ten conse-cutive days. Three serum markers were also measured. DXA-densitometry was used to earmark the osteoporotic subjects. Spearman rank correlation, factor analysis and receiver operating characteristics served as statistical tools. NTx assayed in all-day urine plus PHP and HP in the morning urine gave the most significant correlations. The highest paired coefficients were found between NTx and HP (0.730; 0.689) and between PHP and HP (0.636; 0.648) in both kinds of urine. Lower correlation coefficients were found in DPD and the serum markers. Factor analysis confirmed the preceding findings by the highest communalities for NTx and HP in the all-day and PHP in the morning urine. ROC-analysis afforded AUC values > 0.7 for NTx in both kinds of urine, PHP in the morning urine and osteocalcin. These markers showed also significant concentration differences between osteopenic and osteoporotic subjects. PHP and NTx gave highest values of specificity (0.917), whereas osteocalcin afforded the highest sensitivity (0.857). Due to the easy of processing of non-hydrolyzed urine, speed of chromatographic analysis and the low costs, the PHP assay offers new possibilities in diagnostics of osteopathies.

        Key words: osteoporosis, bone turnover markers, proline-hydroxyproline peptide, Spearman rank correlation, factor and ROC-analysis.