Response of several markers of bone collagen degradation to calcium supplementation in postmenopausal women with low calcium intake

Clinical Chemistry 44:7 1437 1442 (1998) Enzymes and Protein Markers Response of several markers of bone collagen degradation to calcium supplementation in postmenopausal women with low calcium intake Saïd Kamel, 1,2* Patrice Fardellone, 2 Boumedienne Meddah, 1 Florence Lorget-Gondelmann, 1 Jean Luc Sebert, 2 and Michel Brazier 1,2 We investigated the response of bone-specific resorption markers in fasting urine samples from postmenopausal women with low daily dietary calcium (Ca) intake (<800 mg/day) who received either Ca supplementation (1200 mg/day, n 18) or placebo (n 14) for 2 months. We measured urinary hydroxyproline, total pyridinoline, and deoxypyridinoline by HPLC, and free deoxypyridinoline (i-f-dpd) and N- and C-telopeptide fragments of type I collagen (NTX and CTX) by immunoassays. Before supplementation, the urine concentrations of bone resorption markers in the 32 subjects were not statistically different from those measured in 21 subjects with daily dietary Ca intake >800 mg/day. In contrast to the placebo group, Ca supplementation decreased all collagen-related degradation markers except i-f-dpd as early as the first month. The magnitude of response after 2 months of Ca supplementation, expressed as mean percentage of decrease from baseline values or as individual Z scores, was greatest for the telopeptide assays. Furthermore, the percentage of change assessed at 2 months was greater than the within-person biological variability (CV) assessed in the placebo-treated women for NTX and CTX, whereas for the other markers the percentage of change was very close of the within-person CVs. We conclude that crosslinked telopeptide fragments of type I collagen most sensitively reflect the change in bone resorption after Ca supplementation. 1 Clinical Pharmacy Laboratory, Faculty of Pharmacy, 1 Rue des Louvels, 80037 Amiens, France. 2 Department of Biochemistry and of Rheumatology, Centre Hospitalier Régional d, 80054 Amiens, France. *Author for correspondence. Fax 03-22-82-74 69; e-mail Said.Kamel@sa.upicardie.fr. This study is dedicated to the memory of Jean Luc Sebert who initiated this work and whose untimely death prevented him from following it to its conclusion. Received November 12, 1997; revision accepted April 21, 1998. During recent years, a great deal of attention has been focused to the urinary excretion of pyridinoline (Pyd) 3 cross-links [Pyd and deoxypyridinoline (Dpd)] as biochemical markers of bone resorption (1, 2). These compounds seem to offer much greater specificity and sensitivity than the traditional urinary hydroxyproline (Hyp) determination (3, 4). For their measurements, several assays have been developed. These include an HPLC method that uses the natural fluorescence of Pyd and Dpd present in acid hydrolysate of urine (5, 6). Because this method is rather cumbersome, ELISA immunoassays, which allow the urinary determination of either free Dpd (Pyrilinks-D TM ) (7) or peptide-bound cross-links including urine type I collagen cross-linked N-telopeptide (NTX, Osteometer) (8) and urine type I collagen cross-linked C-telopeptide (CTX, CrossLaps TM, Osteomark ) (9) have also been developed. Numerous clinical data have shown that these markers provide a sensitive and specific index of bone resorption process particularly useful to monitor antiresorptive therapy (10 14). Among treatments designed to decrease the rate of bone loss, the usefulness of bone resorption markers has been evaluated essentially with bisphosphonate and estrogen replacement therapy. Despite the convincing evidence that calcium (Ca) supplementation can retard postmenopausal bone loss (15, 16), very few studies have reported its effects on the more bone-specific markers of collagen degradation that have been developed recently. Therefore, we investigated the response of these markers to Ca supplementation (1200 mg/day) during a course of 2 months. The study was conducted in a group of postmenopausal women without vitamin D insufficiency but with a low daily 3 Nonstandard abbreviations: Pyd, pyridinoline; Dpd, deoxypyridinoline; Hyp, hydroxyproline; NTX, N-telopeptide fragment of type I collagen; CTX, C-telopeptide fragment of type I collagen; T-Pyd, total pyridinoline; T-Dpd, total deoxypyridinoline; and i-f-dpd, free deoxypyridinoline. 1437

1438 Kamel et al.: Bone collagen and dietary calcium dietary Ca intake ( 800 mg/day). At baseline, the concentrations of bone resorption markers were compared with a group of postmenopausal women with adequate Ca intake ( 800 mg/day). Changes of each marker with time were assessed by comparison with a group of postmenopausal women with low Ca intake who had received a placebo during the same period. Furthermore, because the major drawback of urinary markers measurement is their biological variability (17), we assessed the response in account of this variability. Materials and Methods subjects and sample collection Thirty-two women 75 years of age (mean age, 64 5.1 years) who had passed menopause more than 5 years previously were studied. They had low daily dietary Ca intake (503 159 mg/day), as assessed by a food frequency questionnaire (18) before the beginning of the treatment, and vitamin D status within reference values, i.e., a value of 25[OH]D 10 g/l (mean serum 25[OH]D concentration, 14.7 3.5 g/l). None of them had received any treatment for bone disorder with Ca, phosphate, vitamin D, fluoride, or bisphosphonates during the previous 6 months. They presented no serious progressive disease and had normal renal function (serum creatinine 120 mol/l). These subjects were randomized to receive either Ca supplementation (1200 mg/day, n 18) as Ca carbonate (Caltrate, Lederle Laboratoire) or placebo (n 14). The active treatment and placebo consisted of two doses to be taken every day with the morning and evening meals. A group of 21 women, ages 75 years (mean age, 65 6 years), with adequate daily dietary Ca intake (1130 352 mg/day) and vitamin D status within reference values (mean serum 25[OH]D concentration, 15.3 3.9 g/l) was studied simultaneously as a control group. They had not received any treatment acting on bone and Ca metabolism during the previous 6 months. Fasting urinary samples were obtained at baseline (M0) in the Ca- and placebo-treated groups as well as in the control group, and at 1 (M1) and 2 months (M2) during Ca or placebo administration. Urine samples were taken in the morning without any predetermined diet and stored at 20 C until use. The subjects of the present study were taken from a larger cohort enrolled to assess the biological effects of Ca supplementation. The study protocol was approved by the local ethics committee. Written informed consent was obtained from all patients, and the trial was conducted in accordance with good clinical practices. urinary assays HPLC analyses of total Pyds (T-Pyd and T-Dpd) were performed on hydrolyzed urinary sample essentially as described previously (19). ELISA immunoassay of free Dpd (i-f-dpd) was performed using a kit (Pyrilinks-D, Metra biosystems ). NTX and CTX were measured respectively with the NTX ELISA immunoassay (Osteomark, Ostex Inc.) and with the CrossLaps ELISA immunoassay (CTX) (Osteometer A/S ). Total Hyp determination was performed by HPLC after acid hydrolysis of the samples (20). All the urinary samples were run in duplicate in the same assay, and all the results were corrected to creatinine excretion. For all these urinary assays, the interassay CVs were 10% (Table 1). statistical analysis Comparisons between the placebo- and Ca-treated groups were performed by using two-way ANOVA for repeated measures. In case of significant interaction between time and treatment effects, a complementary analysis was assessed, comparing the two groups at each time according to Winer s procedure (21). Moreover, Dunnett s tests were performed within groups to assess changes from baseline. The critical difference (CD) for the 95% range of differences was calculated as: CD 1.96 2 CV where CV denotes the within-person biological variability assessed by estimating the CV for repeated measurements performed at baseline and at 1 and 2 months in the 14 placebo-treated women. Results The individual baseline values of each biochemical marker in the group of postmenopausal women with low Ca intake before any supplementation, compared to sexand age-matched controls (expressed as Z scores) are shown in Fig. 1. None of the markers showed any significant differences between the two groups. In all cases, 90% of the values were between the range of 2 and 2 Z scores, except for NTX, which presented slightly more values ( 20%) above 2 Z scores. Reproducibility of the urinary markers was assessed by estimating the CV for repeated measurements performed at baseline and at 1 and 2 months in the 14 placebo-treated women. As can be seen in Table 1, the mean within-subject CV ranged from 13.1% to 21.8%, the greatest reproducibility and stability being obtained for NTX and the smallest for T-Dpd. Table 1. Interassay CV, biological variability, and critical differences for bone resorption markers. Imprecision, Interassay CV, % Within-person biological variability, % Critical difference, % T-Pyd (HPLC) 5.4 16.4 8.3 45 T-Dpd (HPLC) 8.0 21.8 10.5 60 i-f-dpd (ELISA Pyrilinks-D) 6.9 17.0 8.7 47 Hyp (HPLC) 3.9 14.9 6.3 41 NTX (ELISA Osteometer) 7.0 13.1 9.2 36 CTX (CrossLaps, ELISA Osteomark) 4.9 18.3 12.4 51

Clinical Chemistry 44, No. 7, 1998 1439 Fig. 1. Individual baseline values of bone resorption markers (expressed as Z scores) in the group of postmenopausal women (n 32) with low Ca intake. (Z scores are the number of SDs from the mean of sex- and age-matched controls) The time course of the response of each marker of bone resorption to Ca or placebo supplementation is shown in Table 2. Comparisons between the placebo- and Catreated groups yielded significant differences only for T-Dpd at 1 month, and for Hyp and NTX at 2 months. However, differences over time compared with baseline within each group showed that, in the Ca-treated group, all the markers were significantly decreased as early as the first month except for Hyp, whose decrease reached a significant level only at 2 months, and for i-f-dpd, whose changes did not reach a significant level at either 1 or 2 months. In the placebo group, no significant changes with time were seen for any of the markers. The magnitude of response expressed as the mean percentage of change from baseline values after 2 months of Ca supplementation was quite different. As shown in Fig. 2, the greatest response was obtained for the telopeptide assays NTX and CTX ( 26% and 31%, respectively). Figure 3 represents the individual changes after 2 months of Ca supplementation expressed as Z scores and also shows that the most significant responses were obtained with telopeptide measurements. In the placebo group, the mean percentage of change was greater than the within-subject CV only for telopeptides. In addition, 50% of the patients receiving Ca had a change in NTX and CTX concentrations that exceeded the critical difference (i.e., the least significant change at P 0. 05), whereas for the other markers this percentage was 20%. Discussion It is generally agreed that Ca supplementation can prevent postmenopausal bone loss and can reduce bone resorption, as reflected by a reduction in fasting urinary excretion of Hyp (22, 23). We recently studied (24), in a Table 2. Mean values of bone resorption markers at baseline (M0) and at 1 (M1) and 2 (M2) months after Ca supplementation or placebo administration in postmenopausal women with low Ca intake. Indices Placebo group (n 14) Ca-treated group (n 18) a M0 M1 M2 M0 M1 M2 T-Pyd, nmol/mmol Cr b 49.3 14.0 55 18.4 49.5 13.1 54 12.9 46.7 12.3 c,d 44.3 14.7 c,d T-Dpd, nmol/mmol Cr 8.6 3.6 9.2 3.7 8.0 3.3 9.0 2.4 7.0 2.2 c,e 6.4 2.4 c,f i-f-dpd, nmol/mmol Cr 7.2 2.3 7.9 2.2 7.8 2.8 6.4 1.4 6.0 1.9 5.6 1.9 Hyp, mol/mmol Cr 18.5 4.7 17.3 5.0 19.8 6.8 17.0 3.7 15.7 2.9 13.7 3.3 c,e,f CTX, g/mmol Cr 258 124 262 135 261 136 248 90 176 53 c,f 171 62 c,f NTX, nmol BCE/mmol Cr 80 45 78 41 86 45 76 41 55 22 c,d 54 23 c e a Significance effects of time and Ca supplementation and interaction were calculated using a two-way ANOVA with repeated measures. In case of significant interaction between treatment and time, differences between Ca and placebo groups were assessed for each time using a Winer s analysis; P 0.05. b Cr, creatinine; BCE, bone collagen equivalent. c Dunnett s multiple comparison test was used to identify differences over time compared with baseline. d P 0.05. e Winer s analysis, P 0.05. f P 0.01.

1440 Kamel et al.: Bone collagen and dietary calcium Fig. 2. Response of bone resorption markers in postmenopausal women with low Ca intake ( 800 mg/day), after 2 months of Ca supplementation. Results are expressed as the mean ( SE) percentage of decrease from baseline values. larger cohort of postmenopausal women who were not yet 75 years old and with different Ca intake, the effects of Ca supplementation on markers of bone turnover and of Ca homeostasis. We showed that the greatest effects of the supplement were obtained in women with the lowest Ca intake. In the present study, we continued our investigations by studying more particularly the effects of shortterm intervention with supplemental Ca on more specific bone resorption markers in postmenopausal women with low dietary Ca intake assessed before the beginning of the Ca supplementation. Before supplementation, the comparison with a control group whose Ca intake was considered sufficient ( 800 mg/day) has shown that the level of bone resorption assessed by biochemical markers is the same in both groups. This result suggests that neither of these indices was sensitive enough to discriminate between the low and adequate dietary Ca intake groups. However, the number of patient subjects in this study is rather too low to draw firm conclusions. Ca supplementation led to a significant decrease in collagen-related degradation markers, which was already evident after 1 month. At 2 months, the statistical difference in the Ca-treated group reach a high level for all the markers, except i-f-dpd. These results confirm, with more bone-specific markers, the early reduction of bone Fig. 3. Individual values of bone resorption markers (expressed as Z scores) at 2 months in the group of postmenopausal women (n 18) with low Ca intake receiving a Ca supplementation for 2 months. Significant differences vs baseline values by ANOVA: * P 0.05; ** P 0.01; *** P 0.001.

Clinical Chemistry 44, No. 7, 1998 1441 resorption after Ca supplementation that have been previously reported using Hyp as marker (22, 23). Among the indices that have been measured, only i-f-dpd did not change significantly during Ca supplementation. Several recent reports have shown the lack of significant response of i-f-dpd to other antiresorptive therapy such bisphosphonate (25, 26), estrogen (27), and vitamin D Ca supplementation (28). Whether this is because of reduced specificity for i-f-dpd of the osteoclastic resorption process compared with the conjugated form (29) or because of different renal clearance of free and conjugated cross-links, as has been recently proposed (30) is not yet clearly known. The magnitude of the response expressed as the percentage of decrease from baseline values and the individual changes at 2 months expressed as Z scores was greater for the telopeptide assays than the other markers. However, the NTX and CTX percentages of decrease obtained with Ca supplementation in this study are rather weak compared with those obtained with other antiresorptive agents, which ranged from 30% to 80% of decrease (25 28), suggesting that the effects of Ca supplementation on bone resorption are less than the effects of other antiresorptive therapies. Because in our study the decrease in bone resorption markers never exceeded 30%, we compared the magnitude of the marker decrease with the reproducibility of each marker during the 2-month period. It is well known (17) that variability represents one of the main disadvantages of urinary bone resorption markers, and this fact needs to be taken into account when assessing the effects of bone-active drugs. The mean within-subject CV, which describes the variability, was 20% in all cases except for T-Dpd, indicating that, at least during the time intervals assessed, the subjects had relatively stable marker concentrations. However, we measured only short-term reproducibility, and greater variability has been reported, particularly for Hyp and Pyd cross-links measured by HPLC over longer periods (31). Herein, the variability of the markers expressed by the Cvs was less than the percentage of decrease, especially for NTX and CTX. In addition, using the concept of least significant change by calculating the CD, which represents a cutoff point such that a change greater than this value constitutes a true change, it appeared, when bone resorption was measured with such telopeptide assays as NTX and CTX, that a nonnegligible proportion ( 50%) of patients receiving Ca had a change greater than CD, whereas this was not the case with other markers. In conclusion, our results demonstrate and confirm the early decrease of bone resorption after Ca supplementation, using specific markers of collagen degradation. Among these, those that allow the determination of peptide-bound cross-links seem to be more sensitive and could be useful for monitoring the change in bone turnover after Ca supplementation. We are indebted to to A. Morel and C. Diot for technical assistance. We thank Behring Diagnostic France, Cis Bio International France, the Lederle Laboratory, and Ortho Clinical Diagnostics France for supplying immunoassays. References 1. Demers LM, Kleerekoper M. Recent advances in biochemical markers of bone turnover [Editorial]. Clin Chem 1994;40:1994 5. 2. Delmas PD. Biochemical markers of bone turnover. In: Riggs BL, Melton LJ, eds. Osteoporosis: etiology, diagnosis and management, 2nd ed. 1995:319 33. 3. Uebelhardt D, Gineyts EC, Chapuy MC, Delmas PD. Urinary excretion of pyridinium crosslinks: a new marker of bone resorption in metabolic disease. Bone Miner 1990;8:87 96. 4. Bettica P, Moro L, Robins SP, Taylor AK, Talbot J, Singer FR, Baylink DJ. Bone resorption markers galactosyl hydroxylysine, pyridinium crosslinks and hydroxyproline compared. Clin Chem 1992;11:2313 7. 5. Eyre DR, Koob TJ, Van New KP. Quantitation of hydroxypyridinium cross links in collagen by high performance liquid chromatography. Anal Biochem 1984;137:380 8. 6. Black D, Duncan A, Robins SP. Quantitative analysis of pyridinium crosslinks of collagen in urine using ion-paired reversed-phase high performance chromatography. Anal Biochem 1989;169:197 203. 7. Robins SP, Woitge H, Hesley R, Ju J, Seyedin S, Seibel MJ. Direct, enzyme linked immunoassay for urinary deoxypyridinoline as a specific marker for measuring bone resorption. J Bone Miner Res 1994;9:1643 9. 8. Hanson DA, Weis MA, Bollen AM, Maslan SL, Singer FR, Eyre DR. A specific immunoassay for monitoring human bone resorption: quantitation of type I collagen cross-linked N-telopeptides in urine. J Bone Miner Res 1992;11:1251 8. 9. Bonde M, Qvist P, Fledelius C, Riis BJ, Christiansen C. Immunoassay for quantifying type I collagen degradation products in urine evaluated. Clin Chem 1994;40:2022 5. 10. Uebelhardt D, Schlemmer A, Johansen JS, Gyneyts EC, Christiansen C, Delmas PD. Effect of menopause and estrogen treatment on the urinary excretion of pyridinium crosslinks. J Clin Endocrinol Metab 1991;72: 367 73. 11. Garnero P, Shih WJ, Gineyts E, Karpf DB, Delmas PD. Comparison of new biochemical markers of bone turnover in late postmenopausal osteoporotic women in response to alendronate treatment. J Clin Endocrinol metab, 1994;79:1693 700. 12. Rosen HN, Dresner-Pollak R, Moses AC, Zeind AJ, Clemens JD, Greespan SL. Specificity of urinary excretion of crosslinked N- telopeptides of type I collagen as a marker of bone turnover. Calcif Tissue Int 1994;54:26 9. 13. Bonde M, Qvist P, Fledelius C, Riis BJ, Christiansen C. Application of an enzyme immunoassay for a new marker of bone resorption (CrossLaps TM ): follow up on hormone replacement therapy and osteoporosis risk assessment. J Clin Endocr Metab 1995;80: 864 8. 14. Brazier M, Kamel S, Maamer M, Agbomson F, Garabedian M, Chauvenet M, et al. Bone remodeling markers in elderly subjects: effects of vitamin D insufficiency and of it s correction. J Bone Miner Res 1995;10:1753 61. 15. Reid IR, Ames RW, Evans MC, Gamble GD, Sharpe SJ. Effects of calcium supplementation on bone loss in postmenopausal women. N Engl J Med 1993;328:460 4. 16. Recker R, Hinders S, Davies M, Heaney R, Stegman MR, Lappe JM, Kimmel D. Correcting calcium nutritional deficiency prevent spine fractures in elderly women. J Bone Miner Res 1996;11: 1961 6.

1442 Kamel et al.: Bone collagen and dietary calcium 17. Blumsohn A, Hannon RA, Al Dehaimi AW, Eastell R. Short term intraindividual variability of markers of bone turnover in healthy adults [Abstract]. J Bone Miner Res 1994;9:S153. 18. Fardellone P, Sebert JL, Bouraya M, Bonidan O, Leclercq G, Doutrellot C, et al. Evaluation de la teneur en calcium du régime alimentaire par autoquestionnaire fréquentiel. Rev Rhum 1991; 58:99 103. 19. Kamel S, Brazier M, Desmet G, Picard C, Mennecier I, Sebert JL. High performance liquid chromatographic determination of 3-hydroxypyridinium derivatives as new markers of bone resorption. J Chromat 1992;574:255 60. 20. Castelain S, Kamel S, Picard C, Desmet G, Sebert JL, Brazier M. Simple and automated HPLC method for determination of total hydroxyproline in urine. comparison with pyridinolines excretion. Clin Chim Acta 1995;235:81 90. 21. Winer BJ. Statistical principles in experimental design. New York: McGraw-Hill, 1971:380 pp. 22. Horowitz M, Need AG, Philcox JC, Nordin BEC. Effect of calcium supplementation on urinary hydroxyproline in osteoporotic postmenopausal women. Am J Clin Nutr 1984;39:857 9. 23. Horowitz M, Need AG, Morris HA, Wishart J, Nordin BEC. Biochemical effects of calcium supplementation in postmenopausal osteoporosis. Eur J Clin Nutr 1988;42:775 8. 24. Fardellone P, Brazier M, Kamel S, Guéris J, Graulet AM, Sebert JL. Biochemical effects of calcium supplementation in postmenopausal women: influence of dietary calcium intake. Am J Clin Nutr 1998; in press. 25. Garnero P, Gineyts E, Arbault P, Christiansen C, Delmas PD. Different effects of bisphosphonate and estrogen therapy on free and peptide-bound crosslinks excretion. J Bone Miner Res 1995; 10:641 9. 26. Blumsohn A, Naylor KE, Assiri A, Eastell R. Different responses of biochemical markers of bone resorption to bisphosphonate therapy in Paget disease. Clin Chem 1995;41:1592 8. 27. Ebeling PR, Atley LM, Guthrie JR, Burger HG, Dennerstein L, Hopper JL, Wark JD. Bone turnover markers and bone density across the menopausal transition. J Clin Endocrinol Metab 1996; 81:3366 71. 28. Kamel S, Brazier M, Rogez JC, Vincent O, Desmet G, Sebert JL. Different responses of free and peptide-bound crosslinks to vitamin D and calcium supplementation in elderly women with vitamin D insufficiency. J Clin Endocrinol Metab 1996;81:3717 21. 29. Eyre DR. The specificity of collagen crosslinks as markers of bone and connective tissue degradation. Acta Orthop Scand 1995; 66(Suppl 266):166 70. 30. Cowell A, Eastell R. The renal clearance of free and conjugated pyridinium crosslinks of collagen. J Bone Miner Res 1996;11: 1976 80. 31. Gertz BJ, Shao P, Hanson H, Quan H, Harris ST, Genant HK, et al. Monitoring bone resorption in early postmenopausal women by an immunoassay for crosslinked collagen peptides in urines. J Bone Miner Res 1994;9:135 41.