| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Endocrine Research Unit (H.M.H., B.L.R., C.A.M., S.K.), the Departments of Laboratory Medicine (M.F.B.) and Biostatistics (P.C.W.), Mayo Clinic and Mayo Foundation, Rochester, Minnesota 55905
Address all correspondence and requests for reprints to: Sundeep Khosla, M.D., Mayo Clinic, 200 First Street SW, 5–164 West Joseph, Rochester, Minnesota 55905.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
During study A, serum osteocalcin had a circadian pattern, with a peak at 0400 h and a nadir at 1400 h. During study B, however, the afternoon nadir of serum osteocalcin was eliminated (P < 0.001 and P < 0.005 for the difference in the patterns of peak and nadir, respectively, on the 2 study days). In contrast, the circadian patterns of serum PICP and urinary N-telopeptide of type I collagen and free deoxypyridinoline were virtually identical during the two studies. Urinary calcium excretion declined after the cortisol peak, without differences between the 2 study days in phosphorus or sodium excretion or in serum PTH. We conclude that the circadian variation in serum cortisol is responsible for the circadian pattern of serum osteocalcin, but not that of PICP or bone resorption markers. The physiological variation in serum cortisol may also reduce urinary calcium excretion.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Several studies have attempted to define the cause of the circadian pattern of bone turnover. Posture is an obvious candidate, as prolonged bed rest leads to an increase in bone turnover (8). However, Schlemmer et al. (8) found that the circadian variation in the urinary excretion of the bone resorption markers, total pyridinoline (Pyd) and Dpd, remained unchanged after 5 days of total bed rest. Several hormones with effects on bone turnover, including PTH, GH, and cortisol, also exhibit circadian rhythms and could be candidates for mediating the circadian changes in bone turnover. In previous studies, however, Ledger et al. (9) found that abolishing the circadian variation in serum PTH by a continuous iv infusion of calcium had no effect on the circadian variation in the urinary excretion of the bone resorption marker, the cross-linked N-telopeptide of bone type I collagen (NTx). Similarly, the nocturnal increase in serum osteocalcin was not affected by somatostatin-induced inhibition of the nocturnal rise in GH (10).
Several studies have examined the role of cortisol in mediating the circadian rhythm of bone turnover, with somewhat conflicting results (11, 12, 13, 14). Studies using pharmacological agents such as prednisone (11) or metyrapone (12) have found that these agents did alter the circadian pattern of serum osteocalcin. In addition, Kendler et al. (13) found that dexamethasone administration altered the circadian pattern of urinary total Dpd excretion. In contrast, Schlemmer et al. (14) reported that hydrocortisone administered orally in divided doses to hypoadrenal subjects did not prevent the nocturnal increase in bone resorption.
In addition to having effects on bone turnover, cortisol may have significant effects on overall calcium homeostasis, including PTH secretion (15) and renal calcium handling (16, 17). It remains unclear, however, whether these are pharmacological effects or whether the normal circadian variation in serum cortisol also alters PTH secretion or renal calcium handling.
The aim of this study was to use the most rigorous design possible in normal subjects to test for a role for the circadian variation in serum cortisol in determining the circadian variation in bone turnover. We also sought to define the effects of the circadian variation in serum cortisol on overall calcium homeostasis. Thus, we inhibited endogenous cortisol synthesis using metyrapone and infused cortisol at either a variable rate (to mimic the physiological circadian variation in serum cortisol) or at a constant rate (to eliminate the cortisol rhythm) and assessed the circadian variation in bone formation and bone resorption as well as in serum calcium, PTH, and renal calcium handling under the two conditions.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
After approval of the protocol by the Mayo Institutional Review Board, 10 untreated normal postmenopausal women, aged 63–75 yr (mean, 69 yr), were studied. The mean duration of menopause was 19 yr (range, 11–29 yr), and the mean body mass index was 27.7 (range, 21.7–36.1). All subjects gave informed consent. Based on medical history, physical examination, and hematological and biochemical tests, subjects with significant medical diseases such as osteoporosis, renal failure (creatinine, >1.5 mg/dL), adrenal dysfunction, malabsorption, active malignancy, congestive heart failure, hypotension, and psychiatric disorders were excluded from the study. We also excluded nightshift workers and subjects who recently traveled across several time zones. No subject was taking any medication known to affect bone metabolism or adrenal function.
Study protocol
Throughout the study period, subjects were maintained on their
habitual calcium intake, which was assessed by a trained dietitian
using a food frequency questionnaire (18). Each subject was studied
before (study A) and after (study B) elimination of the morning peak of
cortisol as an in-patient at the General Clinical Research Center. For
each study (A or B), subjects were admitted to the General Clinical
Research Center at 1500 h on day 1 and were dismissed at 0900
h on day 3 (see Fig. 1
, A and B, for outline of study protocol). Meals
were served at 0800, 1200, and 1800 h and were consumed within 30
min. Subjects were ambulant (sitting or walking) from 0700–2300 h and
were recumbent from 2300–0700 h (sleeping hours), except to urinate.
During study A, metyrapone (750 mg, orally, every 4 h for 24
h) was administered to block endogenous synthesis of cortisol, and a
variable infusion of hydrocortisone (0.1–1.1 µg/kg·min) was used
to reproduce the normal circadian pattern of cortisol. During study B,
metyrapone, as described above, plus a constant, low dose infusion of
hydrocortisone (0.2 µg/kg·min), was given to eliminate the
endogenous circadian rhythm of cortisol (Fig. 1) (19). Blood was sampled through an
indwelling catheter every 2 h for measurement of serum cortisol,
ionized calcium, PTH, osteocalcin, and carboxyl-terminal propeptide of
type I collagen (PICP). Timed 4-h urine samples were collected for
measurement of calcium, phosphorus, sodium, potassium, NTx, and free
Dpd (f-Dpd).
|
Serum cortisol was measured by an immunochemiluminometric assay [Sanofi, Chaska, MN; inter- and intraassay coefficients of variation (CVs), 6.8% and 5.2%, respectively], serum osteocalcin was measured by an enzyme-linked immunosorbent assay (ELISA; CIS US, Bedford, MA; inter- and intraassay CVs, 13.6% and 3.9%, respectively), serum PICP was determined by ELISA (Metra Biosystems, Mountain View, CA; inter- and intraassay CVs, 9.9% and 4.7%, respectively), urinary NTx was determined by ELISA (Ostex International, Seattle, WA; inter- and intraassay CVs, 13.1% and 7.6%, respectively), and urinary f-Dpd was measured by ELISA (Metra Biosystems; inter- and intraassay CVs, 14.0% and 5.4%, respectively). Serum intact PTH was measured by an immunochemiluminometric assay (20) (inter- and intraassay CVs, 8.0% and 6.0%, respectively). The serum ionized calcium concentration was measured with a Radiometer ICA 2 Analyzer (Radiometer America, Westlake, OH; inter- and intraassay CVs, 1.6% and 0.8%, respectively) (21), and these values are reported corrected to a pH of 7.40. Urinary phosphorus, sodium, potassium, and serum and urinary creatinine were measured by routine automated methods (Hitachi 911 Analyzer, Boehringer Mannheim, Indianapolis, IN). Urinary calcium was measured by an atomic absorption spectrophotometry. The glomerular filtration rate was assessed by measuring creatinine clearance, and the urinary parameters were normalized per dL glomerular filtrate (GF).
Statistical analysis
All data are reported as the mean ± SEM. A variety of statistical analyses have been used to identify or test for circadian rhythm, including trigonometric regression, power spectrum analysis, Kalman filters, nonparametric regression, and deconvolution (22, 23, 24). All of these methods, however, have difficulties handling data consisting of a single cycle of an unspecified functional form. We have chose to employ a method, O’Brien’s extended t test (25), which does not test for any specific form of cycle, but, rather, tests for the difference in pattern between the location of the maximum and the location of the minimum concentration within each subject’s responses. A significant effect (P < 0.05) under this test can be interpreted as evidence of a cycle of some form, consistent across subjects.
O’Brien’s test was used to compare within-subject time of peak concentration to the time of minimum concentration, within studies A and B separately, to test for circadian rhythm. Also, the same test was used to compare the time of peak and the time of nadir between study A and study B to test for an effect of cortisol suppression. In addition, for each parameter, the magnitude of the variable at a given time point was compared between study A and study B using the Wilcoxon rank sum test.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Figure 2 shows the serum cortisol
levels in the study subjects during either the variable cortisol
infusion or the continuous infusion days. During the variable infusion,
serum cortisol levels peaked in all 10 subjects at 0800 h, whereas
on the continuous infusion day, cortisol levels did not have an early
morning peak, but, rather, there was a gradual increase over the course
of the day (P < 0.0001 and P < 0.02
for the differences in the times of peak and nadir on the 2 days,
respectively).
|
Figure 3 shows the serum osteocalcin
and PICP levels in the study subjects during either the variable
cortisol infusion or the continuous infusion days. The cortisol peak at
0800 h resulted in a daytime nadir in serum osteocalcin levels,
which was absent on the continuous infusion day (Fig. 3A;
P < 0.001 and P < 0.005 for the
differences in the times of peak and nadir on the 2 days,
respectively). In contrast, serum PICP levels showed a persistent
daytime nadir on both the variable and continuous cortisol infusion
days (Fig. 3B). Toward the end of the study period, however, serum PICP
levels tended to be higher on the continuous cortisol infusion day than
those on the variable cortisol infusion day, although the differences
in the magnitude of PICP levels at the individual time points were not
statistically significant.
|
, the
circadian variation in urinary NTx excretion was virtually identical
during the variable cortisol infusion and the continuous cortisol
infusion days. For f-Dpd, the circadian pattern was present during both
study days (Fig. 4B), although there was a broader peak with a smaller
magnitude (P = 0.08) during the continuous infusion
vs. the variable infusion days.
|
Serum ionized calcium and PTH values were virtually identical
during the variable and continuous cortisol infusion days (Fig. 5, A and B). The pattern over time in
urinary calcium excretion was, however, significantly different between
the 2 study days; the morning peak of cortisol was followed by a
decrease in urinary calcium between 1200–2000 h, whereas there was a
late afternoon rise in urinary calcium excretion on the continuous
infusion day (Fig. 6A; P
< 0.05 for the timing of the peak in urinary calcium excretion on the
2 study days). Urinary phosphorus excretion, however, was similar on
the 2 days (Fig. 6B). The differences in renal calcium handling were
not due to differences in sodium handling, as sodium excretion was also
not different between the 2 study days (Fig. 7A). Potassium excretion, however, was
significantly different between the 2 study days; the cortisol peak was
followed by a peak in potassium excretion, which was not present during
the continuous infusion (Fig. 7B; P < 0.0001 for the
timing of the peak in potassium excretion on the 2 study days),
although potassium excretion tended to increase over the course of the
continuous infusion day, perhaps reflecting the gradually rising
cortisol levels (Fig. 2
). Neither the pattern nor the magnitude of the
glomerular filtration rate was significantly different between the 2
study days (data not shown).
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In contrast to osteocalcin, serum PICP levels showed a persistent circadian variation in the presence or absence of the cortisol peak. The discrepancy between the changes in serum osteocalcin and PICP are not surprising, as bone formation involves multiple steps, and the two serum markers represent different aspects of osteoblast function (26). Thus, these data would be consistent with a cortisol-induced inhibition of osteocalcin synthesis, without effects of the physiological variation in cortisol on type I collagen synthesis by bone cells. Previous studies have shown that serum osteocalcin levels are the most sensitive marker of glucocorticoid inhibition of osteoblast function compared to either PICP or bone alkaline phosphatase levels (26, 27). However, as glucocorticoids also inhibit osteocalcin gene transcription (28, 29), it is possible that at least part of the decline in circulating osteocalcin levels after the cortisol peak is related to a direct inhibition of osteocalcin gene expression by cortisol.
Our results also show that the circadian variation of cortisol is not responsible for the circadian variation in bone resorption, as assessed by two different bone resorption markers. Thus, elimination of the morning peak of cortisol had no effect on the circadian variation in urinary NTx excretion. The circadian pattern in f-Dpd excretion was also present on both days, although the peak was somewhat broader and of a smaller magnitude when the cortisol peak was absent. Although not affecting the circadian pattern of f-Dpd excretion, cortisol may alter the relative amount of f-Dpd released during collagen breakdown and/or alter the metabolism or clearance of f-Dpd.
Our findings regarding the effects of cortisol on the circadian variation of bone resorption are in contrast to the results of Kendler et al. (13), who administered a pharmacological dose of dexamethasone and found that this eliminated the circadian variation in urinary total Dpd excretion. Similarly, Lakatos et al. (30) reported that the circadian rhythm of in vitro bone-resorbing activity in human serum was altered by the glucocorticoid antagonist, RU486. In contrast, Schlemmer et al. (14) found, in agreement with our results, that oral hydrocortisone administered in divided doses to hypoadrenal subjects did not prevent the circadian variation in urinary total Pyd excretion. Again, by first reproducing and then eliminating the cortisol peak, our study provides the most definitive test of the hypothesis that cortisol is responsible for the circadian variation in bone resorption. As noted earlier, previous studies have excluded posture (8) or PTH (9) as mediators of this circadian pattern in bone resorption. As cortisol does not mediate this effect either, the cause of the circadian variation in bone resorption remains unclear at present. Studies in rats have shown that the frequency of feeding may influence the circadian changes in bone resorption (31); whether this is also true in humans is unknown and requires further study.
To the extent that changes in serum osteocalcin reflect changes in bone formation, the demonstration that the morning peak of cortisol depresses osteocalcin production during the day has potential implications for diseases associated with even mild derangements in the hypothalamic-pituitary-adrenal axis. Recent studies indicate, for example, that depression is associated with reduced bone mineral density (32) as well as with alterations in 24-h cortisol secretion rates and the circadian variation in serum cortisol (33). Similarly, anorexia nervosa is also associated with osteopenia (34) and with alterations in the circadian pattern of cortisol secretion (35). Our data suggest that these alterations in cortisol rhythmicity by themselves may significantly impair osteoblast function.
Our data also indicate that the circadian variation in serum cortisol has significant effects on urinary calcium excretion. Thus, calcium excretion declined after the cortisol peak compared to the day without the cortisol peak, when there was a rise in calcium excretion. These results are somewhat surprising, as pharmacological glucocorticoid therapy in humans is generally associated with hypercalciuria (16, 17). Our findings, however, suggest that the physiological variation in serum cortisol results in renal calcium conservation. This does not appear to be due to increased PTH secretion, because PTH levels were similar on the 2 study days. The effect appears to be relatively specific for renal calcium handling, as phosphorus excretion was not different on the 2 study days. The decrease in calcium excretion is also not due to alterations in renal sodium handling, as that was also similar on the 2 days, although potassium excretion was different, perhaps reflecting a small mineralocorticoid effect of the variable cortisol infusion. Taken together, therefore, these data indicate that the physiological variation in serum cortisol may have an independent effect on renal tubular calcium reabsorption that requires further study.
In summary, our data show that the circadian variation in serum cortisol is responsible for the circadian pattern of serum osteocalcin, but not that of PICP or bone resorption markers. The etiology of the circadian variation in bone resorption remains unclear at present and requires further investigation. These studies also define the effects of the circadian variation in serum cortisol on overall calcium homeostasis, including possible direct effects of cortisol on renal calcium handling.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received August 27, 1997.
Revised November 17, 1997.
Accepted November 25, 1997.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. Debono, R. J Ross, and J. Newell-Price Inadequacies of glucocorticoid replacement and improvements by physiological circadian therapy Eur. J. Endocrinol., May 1, 2009; 160(5): 719 - 729. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Gafni, A.A. Ptitsyn, Y. Zilberman, G. Pelled, J.M. Gimble, and D. Gazit Circadian Rhythm of Osteocalcin in the Maxillomandibular Complex Journal of Dental Research, January 1, 2009; 88(1): 45 - 50. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. B. Henriksen The Gut Feeling of Bone Remodeling IBMS BoneKEy, November 1, 2005; 2(11): 16 - 23. [Full Text] [PDF]
|
|
|
|
 |
|
 B. F. Jackson, A. Blumsohn, A. E. Goodship, A. M. Wilson, and J. S. Price Circadian variation in biochemical markers of bone cell activity and insulin-like growth factor-I in two-year-old horses J Anim Sci, November 1, 2003; 81(11): 2804 - 2810. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Clowes, H. C. Allen, D. M. Prentis, R. Eastell, and A. Blumsohn Octreotide Abolishes the Acute Decrease in Bone Turnover in Response to Oral Glucose J. Clin. Endocrinol. Metab., October 1, 2003; 88(10): 4867 - 4873. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Clowes, R. T. Robinson, S. R. Heller, R. Eastell, and A. Blumsohn Acute Changes of Bone Turnover and PTH Induced by Insulin and Glucose: Euglycemic and Hypoglycemic Hyperinsulinemic Clamp Studies J. Clin. Endocrinol. Metab., July 1, 2002; 87(7): 3324 - 3329. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Caillot-Augusseau, L. Vico, M. Heer, D. Voroviev, J.-C. Souberbielle, A. Zitterman, C. Alexandre, and M.-H. Lafage-Proust Space Flight Is Associated with Rapid Decreases of Undercarboxylated Osteocalcin and Increases of Markers of Bone Resorption without Changes in Their Circadian Variation: Observations in Two Cosmonauts Clin. Chem., August 1, 2000; 46(8): 1136 - 1143. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Dennison, P. Hindmarsh, C. Fall, S. Kellingray, D. Barker, D. Phillips, and C. Cooper Profiles of Endogenous Circulating Cortisol and Bone Mineral Density in Healthy Elderly Men J. Clin. Endocrinol. Metab., September 1, 1999; 84(9): 3058 - 3063. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
