J. Endocrinol. Invest. 18.' 723-728,1995

COMMENT

Temporal relationship between bone loss and increased bone turnover: A longitudinal study following natural menopause R. Rosso, S. Minisola, A. Scarda, M.T. Pacitti, V. Carnevale , E. Romagnoli, and G.F. Mazzuoli Istituto di II Clinica Medica, Cattedra di Medicina Interna, Universita degli Studi di Roma "La Sapienza", Roma, Italy

ABSTRACT. We report the results of a longitudinal study aimed at better defining concomitant changes of both bone mineral density (BMD) and of four in­dependent markers of bone turnover (serum osteo­calcin, serum alkaline phosphatase activity, fasting urinary hydroxyproline/creatinine and calcium/cre­atinine ratio) following natural menopause. The re­sults obtained indicate that, within a relatively short period of time since cessation of gonadal function, conventional markers of bone turnover behave dif­ferently. In fact, while the mean values of hydrox­yproline/creatinine ratio ( felt to be a marker of bone resorption) rise immediately at the first control (19.7±11.7 months), the bone formation markers gradually increase and, as far as serum osteocal­cin levels are concerned, this increment appears to be long-lasting. As a result of these changes, a neg­ative skeletal balance follows, which is documented by the prolonged reduction of bone mineral density

INTRODUCTION

A number of c linical (1-3) and experimental (4 , 5) studies have been carried out during the past decade in order to define the temporal relationship between inc reased skeletal turnover rate and bone loss which occurs following oestrogen deficiency; as a result of these studies , our knowledge of the pathophysiology of postmenopausal osteoporosis has been greatly enhanced. Research with a longitudinal design has been most­ly performed, both on experimental animals and on human beings , following surgical menopause. To

Key-words: Menopause. biomarkers. bone mineral density. bone turnover.

Correspondence: Dr. S. Minisola. Istituto di II Clinica Medica. Policlinico Umberto I. Via del Policlinico 155. 00161 Roma. Italy.

Accepted June 1. 1995.

723

during the entire observation period. Mean±SD % measured yearly bone loss was -2.83±2.6. There was a highly significant correlation between initial and final BMD values (r= 0.908, p<O.001; r2= 82.5) and a weak inverse correlation (r= -0.298, p<0.046) between initial serum alkaline phosphatase values and % yearly bone loss. In conclusion, measure­ment of the biological indices of bone remodelling following natural menopause indicate that the in­crease in osteogenesis is delayed compared to that of bone resorption; furthermore, in the immediate postmenopausal period , the actual bone mass should be considered the best predictor of future bone mass. The inverse correlation found between % yearly bone loss and serum alkaline phosphatase values seems to emphasize the importance of in­creased bone turnover as an independent predic­tor of bone loss.

the best of our knowledge , few investigations of this type have been conducted after natural meno­pause (6, 7). Furthermore, regarding the cross-sec­tional studies published so far, these have been fo­cused on the entire period ef estrogen deficiency (8 , 9) and not on the years immediately following menopause , when more deleterious effects on bone have been demonstrated (10-12) . In this paper, we report the results of a longitudinal study aimed at better defining, in a group of healthy women in the immediate postmenopausal period , concomitant changes of both bone miner­ai density and of four independent markers of bone turnover.

MATERIALS AND METHODS

Seventy-two healthy female subjects , retrospec­tively selected from the women attending our

R. Rosso, S. Minisola, A Scarda , et al.

"Mineral Metabolism Service" because it had been three years, at most, since their last menstrual pe­riod, were initially invited to participate in this in­vestigation. Eleven of them refused, while another sixteen were excluded as they had developed con­comitant illness (5 subjects) or since they were be­ing treated pharmacologically for prevention or treatment of osteoporosis (the use of bone active drugs being a criterion of exclusion). All the re­maining 45 women gave informed consent to have repeated bone mineral density and biochemical pa­rameter measurements. Each subject underwent an initial comprehensive health survey, a physical examination and a multi­channel autoanalyzer chemical screening accord­ing to a procedure previously reported (13) Detailed information was obtained on menstrual periods; women who had had hysterectomy were initially ex­cluded because the operation made it impossible to date menopause accurately. Menopause was de­fined by the presence of elevated FSH and LH , low estradiol levels and an absence of menses for at least 12 months; no further period was reported dur­ing the follow-up of subjects investigated within 1 year of menses cessation. Metabolic study included a short urine collection (from 08:00 h to 11.00 h) after a 12-h overnight fast in order to determine creatinine (Cr), calcium (Ca) and hydroxyproline (OHPr). Halfway through this collection time (at 09:30) a blood sample was tak­en for the measurement of main parameters of cal­cium metabolism and in particular , for the purpos­es of this study, for quantification of serum total al­kaline phosphatase activity (ALP) and serum os­teocalcin (BGP). Finally, bone mineral density (BMD) at an ultradistal radial point was also evalu­ated. All these measurements were again repeated in 29 patients after a mean interval of 30±5.6 months since cessation of ovarian function and in all subjects at final evaluation (39.8± 7.7 months). Serum osteocalcin levels (initially measured in 35 patients) were assayed using a radioimmunoassay kit (Incstar Co., Stillwater, MN, USA) based on the method of Price and Nishimoto (14). Urinary hy­droxyproline was measured according to previously described methods (15). Serum ALP and creatinine in serum and urine were measured by autoanalyz­er (Technicon Autoanalyzer RA 1000, Tarrytown, NY, USA). Serum calcium was measured by atom­ic absorption spectrophotometry (Perkin Elmer, model 1100 B, Norwalk, CT, USA) . Bone mineral density (BMD) was measured at the ultradistal ra­dius of the non-dominant arm (1 /10 of the forearm length , approximately 2.5 cm from the styloid pro­cess) by a dual photon densitometer (NIM, Verona,

724

Italy) which utilizes two radionuclides 24 1Am and 1251,with energies of about 60 keV and 30 keV, re­spectively. Details of the method have been de­scribed elsewhere (16)

Statistical Methods

Unless otherwise specified all values are reported as mean±1SD. In order to compare the results among markers of bone turnover and BMD measurements, individual values were expressed as standard units (SO-unit or Z-transform) in respect to corresponding values found in a fertile age-matched population studied at the same period of time (17) This latter group was composed of 23 women , whose mean age, height and weight were 51.7 ±4.4 years, 164± 7.0 cm and 62.6±4.1 kg, respectively. Standard units (the dis­tribution of which is centered on a mean of 0 with a SO of 1) are used here as a means of simplifying the comparison of different variables. Mean values were used to test the difference between groups by Student's ttest for unpaired data. Finally, all cor­relations were tested by linear regression analysis. Significance level was set at p<0.05 for all tests. In all subiects , the rate of change in BMD was nor­malized for a 1-year period and expressed in g/cm2 and in percentage of baseline value . In su bjects measured more than twice, the annual change was calculated by the slope of linear regression versus time.

RESULTS

Basal values of demographic, anthropometric, den­sitometric and biochemical parameters of the pop­ulation studied are reported in Table 1. Figure 1

Tab le 1 - Initial demographic, anthropometric, densitometric and biochemical indices of postmenopausal women!.

AGE (yr) 52.4±4.1

HEIGHT (cm) 161.1±5.1

WEIGHT (kg) 65 .9±5.9

ISM (month s) 19. 7± 11.7

BMD (mg/cm2) 330±67

s. ALP (lUlL) 77 .0±19.2

s. BGP (ng/mL) 3.3± 1 0

f. u. OHPr/Cr (mol ratio) 28.2±9.6

f. u. CalCr (mol . rat io) O.38±O.16 ~~--- - --

1 All values are reported as mean± 1 SO. ISM= interval Since menopause ; s. ALP= serum total alka line phosphatase activity ; BGP= osteocalc in; f.u. OHPr/Cr= fasting urinary hyd roxyprol ine/creatinine ratio Ca/Cr= cal­c ium/ creatinine ratio.

(f) UJ

2

::J 1 ...J « > //-~ O~~~--------~~-----------­o o (f)

N -1

12 24 36 48

MONTHS SINCE MENOPAUSE

"ALP +BGP *OHPr/Cr ~Ca/Cr ~BMD

Fig. 1 - Longitudinal changes of mean±SE Z score values of four independent markers of bone remodelling and of bone min­erai density, following natural menopause. * significantly differ­ent in respect to mean values of both serum ALP (p<D.042) and CaiCr ratio (p<D.033);** significantly different in respect to mean values of other biomarkers, *** significantly different in respect to other biomarkers. See also the text for explanation.

shows longitudinal changes of mean±SE Z score values of parameters of bone remodelling taken in­to consideration along with concomitant changes of BMD values. Mean absolute initial values of each biomarker considered were significantly higher in respect to the fertile age-matched population (ALP= 63.8± 16.5 IU/L, p<0.004; BGP= 2.5±0.9 ng/mL; p<0.002; OHPr/Cr= 21.2±4.5 mol/mol, p<0.001; Ca/Cr= 0.27 ±0.14, p<0.002). A statisti­cally significant decline of BMD values (p<0.05) in respect to the values found in the whole fertile pop­ulation recruited in our laboratory (347 ±40 mg/cm2) was also observed. As far as each observation pe­riod was concerned, initially there was a significant difference between mean Z score values of the OHPr/Cr ratio (1.55±2.14) in respect to corre­sponding values of both serum ALP (O.80± 1.16; p<0.042) and urinary Ca/Cr ratio (0.77 ± 1.11; p<0.033); at the second visit, mean Z score values of Ca/Cr ratio (0.38±0.95) were significantly reduced (p between <0.015 and 0.001) in respect to other biomarkers. At the final visit, mean Z score values of serum BGP (1.99±1.64) were significantly high­er in comparison to corresponding values of other bone remodelling parameters (OHPr/Cr= 0.96±2.78, p<0.035; ALP= 1.25± 1.28, p<O.019; Ca/Cr= 0.40± 1.07, p<0.001). Finally, there were significant differences between initial and final Z score units only as far as serum BGP (p<0.001) and BMD (p<0.038) values were concerned. Mean±SD % measured yearly bone loss was

725

Bone loss and bone turnover

500

r=O.908 •• 450 p<O.OO1 :/. '" E 400

.!:! • C) • • 5 350 • • c • ::iE 300 , .. co ...J 250 .& •• <C Z

~ • u::: 200

150 200 250 300 350 400 450 500

INITIAL BMD (mg/cm2)

Fig. 2 - The correlation between initial and final bone mineral density values in subjects longitudinally followed following nat­ural menopause.

-2.83±2.6. Figure 2 shows the correlation found between initial and final BMD values (r= 0.908, p<0.001; r2= 82.5); no correlation was found in this series of patients between individual rates of bone loss and initial mass (r= -0.11, p= NS). Finally, there was a weak inverse correlation (r= -0.298, p<0.046) between initial serum ALP values and % yearly bone loss (Fig. 3).

DISCUSSION

This longitudinal study indicates that, within a rel­atively short period of time since natural menopause, conventional markers of bone

140

:J'130 • ...... ~120 • a.. 110 r=-0.29 ..J <C 100 • •• p< 0.04

::2 90 • ::J • a: 80 • W C/) 70

..J 60 • <C •• • • E 50 • • • • Z 40 •• • 30

-9 -8 -7 -6 ·5 -4 -3 -2 ·1 2

% YEARLY BONE LOSS

Fig. 3 - The correlation between initial serum ALP values and percent yearly bone loss.

R. Rosso, S. Miniso/a, A Scarda, et al.

turnover behave differently. In fact, the mean val­ues of OHPrlCr ratio (felt to be a marker of bone resorption) rise immediately at the initial visit, per­formed within a mean interval of 19 months since menopause. On the contrary, the mean values of bone formation markers gradually increase and, as far as serum osteocalcin levels are concerned, this increment appears to be long-lasting. Finally, the mean CalCr values, even though initially sig­nificantly different in respect to those found in premenopausal women, increase very little. As a result of these changes, a negative skeletal bal­ance follows, which is documented by the pro­longed reduction of bone mineral density during the entire observation period. In some aspects, these results are similar to those obtained in longitudinal investigations carried out on ovariectomized women (1, 3). In both condi­tions, measurement of the biological indices of bone remodelling indicate that the increase in os­teogenesis is delayed compared to that of bone resorption. Following natural oestrogen deficien­cy, and as far as our observation period is con­cerned, we can assume that the observed nega­tive skeletal balance is due to a relatively higher bone resorption rate, not parallel led by a corre­sponding increase in bone formation rate. Histological findings lend support to this hypoth­esis (18,19). Unfortunately, we do not have, ow­ing to the inclusion criteria adopted (see Material and Methods section), early biochemical and densitometric data, i.e. those of the first twelve months after last menstrual period, when a more remarkable uncoupling of the two processes of bone remodelling cannot be excluded. This notwithstanding a comparison between physio­logical and sUl'gical menopause seems to indi­cate that changes that follow the latter are more marked in respect to those found in our investi­gation carried out after physiological gonadal function cessation. From the anatomical point of view, the observation that the increase of bone formation markers lag behind that of bone resorption markers, might be ascribed to the normal delay between resorption and formation during the completion of each re­modelling cycle (20). This simple mechanism, more accurately described as postponement of re­modelling (21), might be responsible for the tran­sient increase in bone mass we have observed fol­lowing the initial 3-year period of oestrogen defi­ciency (22). An important observation arising from this study is the very high direct correlation found between ini­tial and final BMD values in our series of subjects.

726

Considering the (2 value, this means that about 85% of the variability of the final BMD can be ex­plained by variation in initial BMD results, sug­gesting that only 15% of the initial mass may be modified by other variables. This emphasizes the role of peak bone mass, as a protective factor against the loss of skeletal mass (23-25). The inverse correlation found between initial serum ALP values and % yearly bone loss, seems to un­derline, amongst other variables, the role of in­creased bone turnover as a risk factor for osteo­porosis as already observed by other authors (10, 26,27) and also in other skeletal disorders (16,28, 29). This latter finding makes one consider that in­formation derived from one biochemical marker might substitute the clinical use of more than one in combination for screening programs (30); however, markers with greater sensitivity and specificity may prove to be more useful (31,32). In this context, it is important to emphasize that the lack of signifi­cant correlation between serum osteocalcin values and % yearly bone loss, might be partly ascribed to the relatively small number of subjects studied, or, alternatively, to the technical performance of the assay employed (33). Finally, the absence of such a correlation with advancing years after cessation of gonadal function, seems to suggest that the pos­sible predictive relationship between a biochemi­cal measurement and the rate of bone loss could vary with time since menopause (34). Finally, the lack of a significant correlation between rate of loss and initial skeletal mass seems to indi­cate that the latter may not have a strong effect on the immediate rate of loss even though this slight­ly negative correlation may persist throughout the postmenopausal years. From a practical point of view, our results first of all emphasize the importance of early interventions aimed at increasing peak bone mass; secondly, in the first few years following oestrogen deficiency, owing to the physiological delay between bone re­sorption and formation phases, only anti resorptive agents should be used.

ACKNOWLEDGMENTS

ThiS work was supported by grants from the Italian National Research Council (Progetto Finalizzato Invecchiamento, Code 95-3-623) and the Ministry of Scientific and Technological Research. We wish to thank Ms Gloria Costa for technical as­sistance and Ms Lee OHara for editorial assistance. Portions of the findings were presented at the 15th Annual Meetlllg of the American Society for Bone and Mineral Research (September 18-22,1993; Tampa, FL) and at the 10th International Bone Densitometry Workshop (April 24-29, 1994; Venezia Lido, Italy).

REFERENCES 1. Stepan J.J., Pospichal J., Presl J., Pacovsky V.

Bone loss and biochemical indices of bone remod­eling. Bone 8: 279, 1987.

2. Stepan J.J., Pospichal J., Schreiber V., Kanka J., Mensik J., Presl J., Pacovsky V. The application of plasma tartrate-resistant acid phosphatase to assess changes in bone resorption in response to artificial menopause and its treatment with estrogen or norethisterone. Calcif. Tissue Int. 45: 273, 1989.

3. Mazzuoli G.F., Tabolli S, Bigi F., Valtorta C, Minisola S, Diacinti D., Scarnecchia L., Bianchi G., Piolini M., Dell'Acqua S. Effects of salmon calcitonin on the bone loss induced by ovariectomy. Calc if. Tissue Int. 47: 209, 1990.

4. Wronski T.J., Walsh C.C, Ignaszewski LA Histologic evidence for osteopenia and increased bone turnover in ovariectomized rats. Bone 7: 119, 1986.

5. Wronski T.J., Cintron M., Dann L.M. Temporal relationship between bone loss and in­creased bone turnover in ovariectomized rats. Calcif. Tissue Int. 43 179,1988.

6. Slemenda C., Hui S.L., Longcope C, Johnston C.C. Sex steroids and bone mass. A study of changes about the time of menopause. J. Clin. Invest. 80: 1261, 1987.

7. Falch JA, Gautvik K.M. A longitudinal study of pre- and postmenopausal changes in calcium metabolism. Bone 9 15,1988.

8. Nilas E., Christiansen C Bone mass and its relationship to age and the menopause. J. Clin. Endocrinol. Metab. 65 697, 1987.

9. Kelly P.J., Pocock NA, Sambrook PH, Eisman JA Age and menopause-related changes in indices of bone turnover. J. Clin. Endocrinol. Metab. 69: 1160, 1989.

10. Mazzuoli G.F., Minisola S., Valtorta C. Antonelli R., Tabolli S., Bigi F. Changes in mineral content and biochemical bone markers at the menopause. Isr. J. Med. Sci. 21: 875,1985.

11. Dempster OW., Lindsay R. Pathogenesis of osteoporosis. Lancet 341: 797,1993.

12. Pouilles J.M., Tremollieres F., Ribot C. The effects of menopause on longitudinal bone loss from the spine. Calc if. Tissue Int. 52: 340, 1993.

13. Minisola S, Pacitti MT, Scarda A., Rosso R., Romagnoli E., Carnevale V., Scarnecchia L., Mazzuoli G.F.

727

Bone loss and bone turnover

Serum ionized calcium, parathyroid hormone and re­lated variables: effect of age and sex. Bone Miner. 23: 183, 1993.

14. Price P.A., Nishimoto S.K. Radioimmunoassay for the vitamin K-dependent pro­tein of bone and its discovery in plasma. Proc. Natl. Acad. Sci. USA 77: 2234, 1980.

15. Minisola S., Antonelli R., Mazzuoli G. Clinical significance of free plasma hydroxyproline measurement in metabolic bone disease. J. Clin. Chem. Clin. Biochem. 23: 515,1986.

16. Minisola S., Rosso R., Romagnoli E., Pacitti MT, Scarnecchia L., Carnevale V., Mazzuoli G.F. Trabecular bone mineral density in primary hyper­parathyroidism; relationship to clinical presentation and biomarkers of skeletal turnover. Bone Miner. 20: 113, 1993.

17. Scarnecchia L., Minisola S., Pacitti M.T., Carnevale V., Romagnoli E., Rosso R., Acca M., Scarda A., D'Erasmo E., Mazzuoli G.F. Bone remodeling behaviour in early menopause. Bone Miner. 17 (Suppl.): 170, 1992 (Abstract) .

18. Parfitt A.M. Implications of architecture for the pathogenesis and prevention of vertebral fracture. Bone 13 S41, 1992.

19. Boyce RW., Mosekilde L., Weisbrode S.E., Safron JA, Stills Jr H.G., Jankowsky M.L., Franks A.F., Ebert D.C., Paddock C.L., Gundersen H.J.G. Calcium-restricted ovariectomized Sinclair S-1 minipigs: a model of osteopenia and trabecular plate perforation. J. Bone Min. Res. 6 (Suppl. 1) S218, 1991 (Abstract).

20. Parfitt A.M. Pharmacologic manipulation of bone remodelling and calcium homeostasis. In: Kanis JA (Ed.), Calcium metabolism. Progresses in basic and clinical pharmacology. Karger, Basel, 1990, vol 4, p. 1.

21. Parfitt A.M. Use of bisphosphonates in the prevention of bone loss and fractures. Am. J. Med. 91 (Suppl. 5B): 42S, 1991.

22. Mazzuoli G.F., Bianchi G., Rosso R., Valtorta C. Menopause and bone loss. Calcif. Tissue Int. 48 (Suppl.): A72, 1991 (Abstract).

23. Hansen MA, Overgaard K., Riis B.J., Christiansen C. Role of peak bone mass and bone loss in post­menopausal osteoporosis. Br. Med. J. 303961,1991.

24. Heaney R.P. Bone mass, nutrition, and other lifestyle factors. Am. J. Med. 95 (Suppl. 5A) S29, 1993.

25. Hui S.H., Slemenda CW, Johnston Jr CC. The contribution of bone loss to postmenopausal os­teoporosis. Osteoporos. Int. 1: 30, 1990.

R. Rosso, S. Minisola, A. Scarda, et al.

26. Christiansen C., Riis B.J., Rodbro P. Prediction of rapid bone loss in postmenopausal women. Lancet 1: 1105, 1987.

27. Christiansen C, Riis B.J., Rodbro P Screening procedure for women at risk of develop­ing postmenopausal osteoporosis. Osteoporos. Int. 1: 35, 1990.

28. Scarnecchia L., Minisola S., Pacitti M.T., Carnevale V., Romagnoli E., Rosso R., Mazzuoli G.F. Clinical usefulness of serum tartrate-resistant acid phosphatase activity to evaluate bone turnover. Scand. J. Clin. Lab. Invest. 51517,1991.

29. Riggs B.L. Toward optimal therapy of established osteoporosis: evidence that anti resorptive and formation-stimulat­ing regimens reduce vertebral fractures by inde­pendent mechanisms. Fourth International Symposium on Osteoporosis. Hong Kong, 27 March - 2 April, 1993 (Astract).

30. Blumsohn A., Eastell R.

728

Prediction of bone loss in postmenopausal women. Eur. J. Clin. Invest. 22 764,1992.

31. Mole PA, Walkinshaw MH, Robins sr., Paterson CR. Can urinary pyridinium crosslinks and urinary estro­gens predict bone mass and rate of bone loss after the menopause? Eur. J. Clin. Invest. 22 767,1992.

32. Uebelhart D., Schlemmer A., Johansen J.S., Gineyts E., Christiansen C., Delmas p.o. Effect of menopause and hormone replacement ther­apy on the urinary excretion of pyridinium cross-links. J. Clin. Endocrinol. Metab. 72367,1991.

33. Minisola S., Carnevale V, Pacitti M.T., Romagnoli E., Scarnecchia L. , Rosso R., Minisola G., Mazzuoli G.F. Serum osteocalcin in metabolic bone disease what is its real significance? J. Endocrinol. Invest. 16 277,1993.

34. Minisola S., Rosso R., Carnevale V, Romagnoli E, Scarnecchia L., Pacitti M.T., Scarda A., Mazzuoli GF Peak bone mass and rate of loss in determining post­menopausal skeletal mass. J. Bone Min. Res. 8 (Suppl. 1): S252; 1993 (Abstract).