Lack of seasonal variation in bone mass and biochemical estimates of bone turnover

Bone, 9,285-288 (1988) Printed in the USA. All rights reserved.

8756-3282188 $3.00 + .OO Copyright 0 1988 Pergamon Press plc

Lack of Seasonal Variation in Bone Mass and Biochemical Estimates of Bone Turnover K. OVERGAARD, Department Address

L. NILAS, J. SIDENIUS

JOHANSEN

and C. CHRISTIANSEN

of Clinical Chemistry, Glostrup Hospital, University of Copenhagen,

for corresoondence

of Copenhagen,

and reorint reauests:

Glostrup, Denmark.

Kirsten Overeaard, Department of Clinical Chemistry, Glostrup Hospital, University

Di(-2600 Glostrup: Denmaik.

The aim of the present study was to examine the seasonal influence on both bone mineral content and biochemical parameters of bone turnover in a large group of healthy women.

Abstract Three previous studies have indicated a seasonal variation in bone mineral content, with values during the summer being 1.7% to 7.5% higher than during the winter. We have examined the seasonal influence on both hone mass, biochemical estimates of hone turnover and vitamin D metaholites in 86 healthy women, aged 29-53 years. All participants were followed up for 2 years with examinations every 6 weeks or 3 months. Bone mineral content in the proximal and distal part of the forearm (single photon absorptiometry) did not reveal any significant seasonal variation, whereas bone mineral density of the lumbar spine (dual photon absorptiometry) iudicated that the highest values occurred in winter. None of the biochemical parameters showed any statistically signifitaut cyclical changes. Serum concentrations of 25hydroxyvitamin D and 24,25-dihydroxyvitamin D, showed a highly siguificant seasonal variation, whereas the serum 1,25-dihydroxyvltamiu D concentration was virtually unchanged. We conclude that seasonal variation in bone mineral content and bone turnover should not be taken into account when interpreting data from longitudinal studies of healthy pre- and postmenopausal women on a sufficient vitamin D nutriture. Key Words: Seasonal Bone formation-Bone

variations-Bone resorption.

mineral

Materials and Methods Participants

Eighty-six healthy women were studied. Group I: 28 premenopausal women aged 29-45 years; group II: 14 premenopausal women aged 47-49 years (Nilas and Christiansen, 1988); and group III: 44 women aged 45-53 years, who’had passed a natural menopause within the preceding 6 months to 3 l/2 years (Christiansen et al., 1985). All were free of past or present diseases known to influence calcium metabolism, and none were receiving treatment with sex hormones, including oral contraceptives. The study period was 2 years for each participant. Bone

Mass

Measurements

The bone mineral content of the forearms (BMC) was measured by single photon absorptiometry using a rz51 source (100 mCi) with photopeak at 27 keV. The method determines BMC in a proximal (BMC,,) and a distal region (BMC& with an estimated trabecular bone content of 13% and 55%, respectively (Nilas et al., 1987). Fat correction was made by a method described earlier (Nilas et al., 1985). The long-term in vivo precision is 1.0% for BMC,, and 1.5% for BMC& (Nilas et al., 1985). BMC,,, and BMCdist were measured every 3 months in groups I and III, and every 6 weeks in group II. The bone mineral density (ratio of bone mineral content to area) of the lumbar spine (BMD, r,,e)was measured by dual photon absorptiometry using a e3Gd source (1Ci) with photopeaks at 44 and 100 keV (Lunar Radiation Corporation DP3 scanner). BMD,,, was calculated as the BMD values obtained in vertebrae L2, L3 and L4 including the intervertebral discs. The long-term in vivo precision for BMD,,i, is 3.% (Nilas et al., 1985). The results were corrected for source replacements (Nilas et al., 1988). BMD,,, was measured every 6 months in groups I and II. BMD,,, in group III was measured only every 12 months, and were, therefore, not included in the present study.

content-

Introduction Seasonal variation in serum vitamin D metabolites (Juttmann et al., 1981; Tjellesen and Christiansen, 1983) and urinary calcium excretion (Morgan et al., 1972; Robertson et al., 1974) is well established, with high summer and low winter values. Three previous studies have indicated an associated seasonal variation in bone mass, with maximal values during the summer. The differences between the lowest and highest values ranged from 1.7% to 7.5% (Aitken et al., 1973; Krolner, 1983; Hyldstrup et al., 1986). A seasonal variation of that magnitude would erroneously affect the interpretation of bone mass measurements in longitudinal studies. It is, therefore, of great importance to establish whether bone mass varies cyclically during the year. 285

286

Biochemical

K. Overgaard et al.: Lack of seasonal variation in bone mass transfer

Measurements

of Bone Turnover

Blood samples were taken and urine collected in the morning after an overnight fast. All blood and urine samples were stored at - 20°C until analysed. Serum alkaline phosphatase (SAP) was measured enzymatically according to Scandinavian recommendations (The Committee on Enzymes of the Scandinavian Society for Clinical Chemistry and Clinical Physiology, 1974). Plasma BGP (pBGP) was determined by radioimmunoassay (Johansen et al., 1987). Antiserum was raised in rabbits immunized with purified intact calf BGP, and homogeneous calf BGP was used for standard and tracer. The sensitivity of the assay was 0.8 @ml. The in&a- and interassay variations were ~7% and 12%. Fasting urinary hydroxyproline was measured by a spectrophotometric method (Pgdenphant et al., 1984), and corrected for creatinine excretion (FU Hpr/Cr). To eliminate the interassay variation, samples from each woman were determined in the same assay. Fasting urinary calcium and creatinine were measured on a SMA 6160 autoanalyser, and the urinary calcium was corrected for creatinine excretion (FU G&r). SAP, FU Hpr/Cr, and FU Ca/Cr were measured every 3 months in groups I and III, and every 6 weeks in group II. pBGP was measured every 6 months in groups I and III, and every 6 weeks in group II. Vitamin D Measurements Vitamin D metabolites were determined by methods involving specific extraction procedures, followed by chromatography on a Sephadex LH 20 and high pressure liquid chromatography. The serum concentrations of 25-hydroxyvitamin D (25(OH)D) and 1,25_dihydroxyvitamin D (1,25(OH),D) were measured by competitive protein binding assays, based on vitamin D binding protein from rat serum (Hummer et al., 1984) and calf thymus receptor (Hartwell and Christiansen, 1987), respectively. The serum concentration of 24,25_dihydroxyvitamin D3 (24,25(OH),D,) was measured by radioimmunoassay (Hummer and Christiansen, 1984). The intra- and interassay variations were 8/11%, 9/14%, and 8/13%, respectively. All samples from each woman were measured in the same assay. The vitamin D metabolites were determined every 6 months during the last year of the observational period in a coincidental subgroup of group I (n = 20). Statistical Analysis For each participant the mean value of each parameter was calculated and the individual measurements were expressed in per cent hereof. The percentage values were grouped in 6-week intervals according to the sampling time of year. For vitamin D metabolites 2-month intervals were used instead of 6-week intervals. For each time interval the average v and the standard deviation SD were calculated. A cyclic regression curve y = Bl + B2 SIN ((B3 + t) r/6) were fitted for each parameter by the least square method with weighting for group size, n, and SD: Weight = SD*In (Duggleby, 1981). B2 is the amplitude about Bl, B3 the phase, and t the time in months. IT/~ is used to give a 12month cycle. Cyclical seasonal variation was tested as t test for the hypothesis: B, = 0. Compared to the younger women, bone mass in the postmenopausal women (group III) declined during the 2 years. Since the mean change per year was 2.19% in BMC,,, and 2.86% in BM&*, the individual data were corrected for this mean change.

Table I. Seasonal changes in bone mass measurements, estimates of bone turnover and vitamin D metabolites expressed as estimated amplitudes in a sine curve. Estimated amplitudes 2 SEE (%) 0.08 0.27 1.15 -1.2 -4.8 2.8 -4.9 0.7 7.4 1.1

BMC,, BMC,xt

BMQ,in,

SAP pBGP FU Hpr/Cr FU C&r 25(OH)D 24,25(OH),D,

1.25(OH),D

P

>o.10 >0.05 co.01 >O.lO >0.05 >O.lO >0.05 o. 10

i 0.26 -c 0.28 2 0.80 * 1.4 -+ 5.3 + 4.8 2 5.6 t 0.4 f 4.0 2 3.2

Results Table I shows the seasonal changes expressed as estimated amplitudes in a sine curve. To visualize the magnitude of detectable seasonal variations the amplitudes are given +2 SEE (corresponding to 95% confidence intervals). The variation of BMC,,,, BMCdist, and BMD,,,, is shown in Figure 1. During the first 1 l/2 years of observation, the values of BMC,,,, and BMCdist tended to be highest in summer, the maximal difference being 1.5% and 1.2%, respectively; but this was not reproduced during the last year, and the cyclical changes were not statistically significant. On the other hand, BMD,Pi”, increased in winter and decreased in summer and showed a significant cyclical variation. Figure 2 shows the changes in the biochemical estimates of bone turnover. None of the biochemical variables showed any significant cyclical change.

%

BMCdlst

102 -I

100

JAY

J;”

JiY

Ja’”

July

JC%l

Fig. 1. Bone mass measurements during the study period. and BMC denote bone mineral content in the proximal ~~~~~a1 forearm? and BMD bone mineral density in the lumbar spine. The pkrcentage v~~I?:s are given as mean 2 2 SEM at 6-week intervals.

287

K. Overgaard et al.: Lack of seasonal variation in bone mass transfer

Vitamin D metabolites revealed significant seasonal variations in 24,25(OH)zD, and 25(OH)D, with maximum in June-July and August-September, respectively. 1,25(OH),D revealed no seasonal variation (Fig. 3). Table II demonstrates the relation between vitamin D metabolites and bone mass. All 20 participants had serum vitamin D metabolites determined 3 times. The “low group” consisted of all 20 participants at that time when their vitamin D metabolites were lowest, and the “high group” consisted of all 20 participants at that time when their vitamin D metabolites were highest. The BMC values were averaged at the corresponding times. The groups of low versus high serum vitamin D concentrations had virtually the same bone mass in all three bone compartments.

24.25 (0H)q 03

t-g/ml

5 4 3 2 1 i ng/ml

25 (0H)D

Discussion In the present study we have examined a large, representative group of healthy women. We used a number of classical techniques for determining bone mass and the currently available biochemical estimates of bone turnover. All blood and urine samples from each participant were measured in the same assays. Measurements of BMC,, and BMCdist are well-documented methods with a precision that can be kept below l- 1.5% (Nilas et al., 1985). Measurements of BMD,,,, by

1.25 (OH)2 D

T

30-’

()

2Q-

_

I

T

I

i

T

loQ/o

Jan.

SAP

FlJCdCr

,

lzOf Ia-

J&.

24,25(OH),D, denotes 24,25-dihydroxy-vitamin D,; 25(OH)D denotes 25-hydroxyvitamin D; and 1,25-(OH),Ddenotes I ,25-dihydroxyvitamin D.

Q/o

100

J:lv

Fig. 3. Vitamin D metabolites.

110 -I

0

I

-

, July

, Jan

I

Juhl

,

J?.ll

1

July

I

Jan.

Fig. 2. Biochemical estimates of calcium metabolism during the study period. Indices of bone formation: Serum alkaline phosphatase (SAP), and plasma bone Gla protein (pBGP). Indices of bone resorption: Fasting urinary hydroxyprolinelcreatinine (FU Hprl Cr), and calcium/creatinine (FU Ca/Cr). The percentage values are given as mean * 2 SEM at 6-week intervals.

dual photon absorptiometry involves major precision problems, which are described in detail elsewhere (Nilas et al., 1988). The larger precision error of the spinal measurments is illustrated by the larger variation in the percentage values during the study. We were unable to find any significant seasonal variation in the forearm measurements, but BMD,,, showed the highest values in winter. Some of the data look, as if they do show cyclical changes, but the period is not the same for all measurements. The changes are of such a magnitude that it could be explained by the expected variations in measurement procedures. An overlooked seasonal change in the forearm measurements is less than 1% (Table I), equalling a positive/negative calcium balance of 3.7 g within 6 months. Our results agree with those obtained by Tothill et al. (1984) who found no seasonal variation in total body calcium in 154 patients with various rheumatic diseases. On the other hand, three other studies have indicated a maximal bone mass during summer, 7.5% higher than during winter in oophorectomized women (Aitken et al., 1973), and 2% in healthy women (Krolner, 1983) and men (Hyldstrup et al., 1986). Bone mass was measured in the metacarpal bones, the lumbar spine, and the forearms, respectively. Only the latter investigation included measurements of bone turnover. SAP and whole body retention of diphosphonate were significantly elevated in the first quarter. The increases could not be reproduced, and it might be due to random change (Hyldstrup et al., 1986). Seasonal changes in bone mineral content of 1.7-7.5% within three months reflect a calcium balance of 140-625 mg per day (if mean total body calcium equals 750 g and

288

K. Overgaard et al.: Lack of seasonal

Table II. Relation between vitamin D metabolites and bone mass measurements. serum levels. Values are percentages of mean.

Low

BMC,i,, (%) Mean + 2 SEM

99.6 f 0.5

BMD,,,, (%) Mean k 2 SEM

99.7 2 0.8

loo.4 f 2.0

High

100.0 t

0.3

99.8 2 0.4

99.6 2 2.0

24,25(OH),D,

Low High

99.6 2 0.5 loo.0 * 0.3

99.6 2 0.8 99.5 2 0.6

100.1 2 2.0 99.5 t 2.0

1,25(OH),D

Low High

100.1 ? 0.5 99.9 ? 0.3

100.2 2 0.8 99.7 k 0.6

100.3 * 1.8 100.3 * 2.0

the local bone mass measurements are representative of total body calcium). It is notable that the mean change in calcium balance from the pre- to postmenopausal status is about 50 mg per day, or 2% per year (Heaney et al., 1978). In addition, the daily mean positive calcium balance in successful treatment of osteomalacia never exceeds 300 mg. There are profound changes in the biochemical estimates of bone turnover when passing from the pre- to the postmenopausal status or when osteomalacia is treated. Some change in bone turnover would, therefore, have been expected if there were a significant seasonal variation in bone mass. In other studies seasonal variation in bone mineral content was suggested, because of the well-known cyclical changes in vitamin D metabolites, but these metabolites were not measured (Aitken et al., 1973; Krolner, 1983; Hyldstrup et al., 1986). Our data clearly confirm that 25(OH)D and 24,25(OH),D, vary seasonally, whereas 1,25(OH)zD is virtually constant. As I ,25(OH),D is regarded as the most biologically active vitamin D metabolite, it is most physiological if the levels do not show cyclical changes. Some investigators, however, have found a significant seasonal variation in 1,25(OH),D (Meller et al., 1986; Bouillon et al., 1987). This has been explained in the light of the overall state of vitamin D nutriture. If the vitamin D supply is adequate throughout the year, cyclical changes are less likely to occur. In our part of the world there is a high level of vitamin D nutriture. In communities with an insufftcient vitamin D supply, 1,25(OH)2D may go down during the winter, accompanied by increases in parathyroid hormone and decreases in bone mass. In the present study there was, however, no relationship between any of the vitamin D concentrations and bone measurements . We conclude that seasonal variation in bone mineral content and bone turnover should not be taken into account when interpreting data from longitudinal studies of healthy pre- and postmenopausal women on a sufficient vitamin D nutriture.

Heaney R.P., Reeker R.R. and Saville P.D.: Menopausal changes in calcium balance performance. J. Lab. Clin. Med. 92:953-%3, 1978. Hummer L. and Christiansen C.: A sensitive and selective radioimmunoassay for serum 24,25dihydroxycholecalciferol in man. Clin. Endocrinol. 21:71-79, 1984. Hummer L., Tjellesen L., Rickers H. and Christiansen C.: Measurement of 25.hydroxyvitamin D, and 25hydroxyvitamin D, in clinical settings. &and.

J. Clin. Lab. Invest.

44:595-601,

1984.

Hyldstrup L., McNair P., Jensen GE and Transbol I.: Seasonal variations in indices of bone formation precede appropriate bone mineral changes in normal men. Bone 7:167- 170, 1986. Johansen IS.. Melholm Hansen I.E. and Christiansen C.: A radioimmunoassay for bone Gla protein (BGP) in human plasma. Acta Endocrinol. 114:410-416,

1987.

Juttmann J.R., Visser T.J., Buurman C., De Kam E. and Birkenhager J.C.: Seasonal fluctuations in serum concentrations of vitamin D metabolites in normal subjects. Br. Med. J. 282:1349-1352, 1981. Krolner B.: Seasonal variation of lumbar spine bone mineral content in normal women. Calcif. Tiss. Int. 35: 145-147, 1983. Meller Y.. Kestenbaum R.S.. Galinsky D. and Shany S.: Seasonal variation in serum levels of vitamin D metabolites and parathormone in geriatric patients with fractures in southern Israel. Isr. J. Med. Sci. 22:8-l I, 1986.

Morgan D.B., Rivlin R.S. and Davis R.H.: Seasonal changes in the urinary excretion of calcium. Am. J. Clin. N&r. 25:652-654, 1972. Nilas L.. Borg J.. Gotfredsen A. and Christiansen C.: Comparison of singleand dual-photon absorptiometry in postmenopausal bone mineral loss. J. Nucl. Med. 26: 1257- 1262, 1985. Nilas L. and Christiansen C.: Rates of bone loss in normal women-evidence of accelerated trabecular bone loss after the menopause. Eur. J. C/in. Invest., resubmitted, 1988. Nilas L.. Hassager C. and Christiansen C.: Long-term precision of dual photon absorptiometry in the lumbar spine in clinical settings. Bone and Mineral 3:305-315. 1988. Nilas L.. Nergaard H., Podenphant J., Gotfredsen A. and Christiansen C.: Bone composition in the distal forearm. Stand. J. Clin. Lab. Invest. 47:41-46,

1987.

Pedenphant J., Larsen N.-E. and Christiansen C.: An easy and reliable method for determination of urinary hydroxyproline. C/in. Chim. Acta. 142: 145- 148, 1984. Robertson W.G., Gallagher J.C.. Marshall D.H., Peacock M. and Nordin B.E.C.: Seasonal variations in urinary excretion of calcium. Br. Med. J. 4:436-437.

1974.

Tjellesen L. and Christiansen C.: Vitamin D metabolites in normal subjects during one year. A longitudinal study. &and. J. Clin. Lab. Invest.

References Aitken J.M., Anderson J.B. and Horton P.W.: Seasonal variations in bone mineral content after the menopause. Nature 241:59-60, 1973. Bouillon R.A., Auwerx J.H., Lissens W.D. and Pelemans W.K.: Vitamin D status in the elderly: seasonal substrate deficiency causes 1.25-dihydroxycholecalciferol deficiency. Am. J. C/in. Nutr. 45:755-763, 1987. Chnsttansen C., Riis B.J., Nilas L., Rodbro P. and Deftos L.: Uncouphng of bone formation and resorption by combined estrogen and progestagen therapy in postmenopausal osteoporosis. Loncet ii:SOO-801. 1985. Duggleby R.G.: A nonlinear regression program for small computers. Anal. Eiochem.

in bone mass transfer

Vitamin D metabolites are stratified into low and high

BMC,,, (a) Mean t 2 SEM ZS(OH)D

variation

110:9-18,

43:85-89,

1983.

The Commtttee on Enzymes of the Scandinavian Society for Clinical Chemistry and Clinical Physiology. Recommended methods for the determination of four enzymes in blood. Sand. J. C/in. Lab. Invest. 33:291306. 1974.

Tothill P.. Nicoll J., Kennedy N.S.J.. Smith M.A., Reid D.M. and Nuki G.: The lack of seasonal variation of total body calcium. In: Osteoporosis. The Proceedings of the Copenhagen International Symposium on Osteoporosis. Christiansen et al.. eds., Copenhagen, pp. 815-817, 1984.

1981.

Hartwell D. and Christiansen C.: Comparisons between two receptor assays for 1,25-dihydroxyvitamin D. Stand. J. Clin. Lab. Inwst. 48: 10% 114, 1988.

Received: March 4. 1988 Revised: July 18, 1988 Accepted: July 21, 1988