| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Clinical Research Centers |
Division of Endocrinology, Diabetes, and Clinical Nutrition, Oregon Health Sciences University, Portland, Oregon 97201
Address all correspondence and requests for reprints to: Dr. M. H. Samuels, Division of Endocrinology, Oregon Health Sciences University, 3181 SW Sam Jackson Park Road, Portland, Oregon 97201.
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
To answer this question, we designed a hydrocortisone (HC) infusion protocol that reproduced serum cortisol levels measured during fasting in healthy subjects. In each subject, TSH levels were measured in the fed state, during fasting, and during the HC infusions. This allowed us to isolate one variable, serum cortisol levels, from the many variables that change during fasting and investigate whether this variable was sufficient to account for the observed TSH suppression. We measured serum TSH levels over 24 h during these studies and analyzed dynamic TSH secretion, because single or random TSH measurements may underestimate the severity of TSH suppression during fasting or glucocorticoid administration.
| Subjects and Methods |
|---|
|
|
|---|
Eight healthy nonobese subjects (four men and four women, aged 2043 yr) were recruited for the study. Each subject underwent a history, physical examination, and laboratory testing (complete blood count, SMA-20, free T4, and TSH) to exclude chronic illness, medication use, or thyroid or adrenal disease. Each subject signed a consent form approved by the Oregon Health Sciences University institutional review board before starting the study.
Studies
Each subject underwent three in-patient studies conducted at the Oregon Health Sciences University General Clinical Research Center, performed at least 1 month apart and in random order.
Baseline. The subject was admitted at 0800 h on day 1, and iv saline was administered at 75 cc/h for the next 48 h. The subject was allowed ad libitum access to food and beverages. Blood samples (3 cc) were withdrawn every 15 min over 24 h from an iv catheter starting at 0800 h on day 2. At 0800 h on day 3, 250 µg TRH were administered iv, and blood samples were withdrawn 20, 30, and 60 min later.
Fasting. The subject was instructed not to consume any food or beverages starting at midnight the night before admission. He or she was admitted at 0800 h on day 1, and saline was administered as described above. The subject was allowed unlimited access to water, but was not allowed to consume any calorie-containing food or beverages for the next 48 h, for a total fasting time of 56 h. Blood samples were withdrawn as described above starting at 0800 h on day 2 (i.e. during the final 24 h of the 56-h fasting period), and a TRH test was performed at 0800 h on day 3. This protocol is identical to previous protocols used by the investigators to study TSH suppression during fasting (1).
HC. The subject was admitted at 0800 h on day 1, and an iv infusion of HC in saline was administered for the next 48 h in a dose designed to mimic the subjects fasting-induced cortisol elevations (see below). The subject was allowed unlimited access to food and beverages. Blood samples were withdrawn as in the other studies starting at 0800 h on day 2, and a TRH test was performed at 0800 h on day 3.
HC doses
In a previous study, investigators at the NIH developed an iv HC infusion that reproduces normal serum cortisol levels in subjects with Addisons disease (8). This protocol includes a variable basal HC infusion rate with superimposed HC pulses to mimic the normal circadian and pulsatile patterns of cortisol secretion. The total amount of HC infused over 24 h is 19 mg. The protocol was obtained from the investigators (J. Zawadzki, personal communication) and was verified to reproduce normal 24-h serum cortisol levels in seven subjects with Addisons disease (data not shown).
For the current study, this HC infusion protocol was initially
increased by 40% (24-h dose = 27 mg) based on our previous study
in fasted subjects that found that mean serum cortisol levels increased
40% over those in fed studies (1). The resulting serum cortisol levels
in the first two subjects were compared in fasting and HC studies by
visual inspection. In both cases, the HC infusion rate was too low to
mimic fasting-induced cortisol elevations. Further modifications in the
HC infusion rates were made until each subjects 24-h cortisol
profiles during fasting and HC studies were comparable. The 24-h HC
infusion protocol was given from 0800 h on day 1 until 0800 h
on day 2 and was then repeated until 0800 h on day 3. Blood
samples were withdrawn during the second 24-h infusion. Subjects
required a mean HC dose of 47 mg/24 h (24 mg/m2·24 h) to
achieve adequate cortisol levels (dose range, 3293 mg/24 h or 1941
mg/m2·24 h). Five subjects required a single HC study,
one subject required two HC studies, and two subjects required three HC
studies to optimize doses. In each case, the optimal infusion profile
was chosen without knowledge of serum TSH levels to avoid potential
bias in determining HC doses. Figure 1
illustrates a typical 24-h HC infusion protocol for a subject.
|
Serum samples were assayed for cortisol and TSH by chemiluminescent assays (Nichols Institute, San Juan Capistrano, CA). All samples from a single study were run in the same assay. Assay sensitivity levels were 0.8 µg/dL for cortisol and 0.01 mU/L for TSH. Mean intraassay coefficients of variation were 4% for cortisol at serum levels measured in the current study, 5% for serum TSH levels above 0.5 mU/L, and 13% for serum TSH levels below 0.5 mU/L. The mean interassay coefficients of variation were 11% for cortisol, 7% for TSH levels above 0.5 mU/L, and 20% for TSH levels below 0.5 mU/L. Serum total T3 levels were measured by RIA, and serum free T4 levels were measured by equilibrium dialysis (Nichols Institute) at 0800 h on day 3 of each study.
Data analysis
TSH pulses were located by cluster analysis, using dose-dependent coefficients of variation calculated from sample replicates in each subjects hormone series. Cluster parameters were two points for test nadirs and one point for test peaks. The t statistics were 2.0 for both up- and down-strokes. These parameters yield false positive and negative peak detection rates less than 5%, as determined by analysis of pooled serum samples and simulated hormone series (1).
To compare serum cortisol levels at different times, 24-h cortisol profiles for each study were divided into four 6-h frames (08001345, 14001945, 20000145, and 02000745 h). To compare daytime and nocturnal TSH levels, 24-h TSH levels were divided into two 12-h frames (daytime, 08001945 h; nocturnal, 20000745 h). All 24, 12, and 6 h hormone levels, TSH pulse parameters, TSH responses to TRH (peak minus baseline), and serum T3 and free T4 levels were compared among the studies by repeated measures ANOVA with Bonferroni t tests. Daytime and nocturnal TSH levels were compared within a study by paired t tests. Results are presented as the mean ± SEM.
| Results |
|---|
|
|
|---|
Twenty-four-hour mean serum cortisol levels increased 32% over baseline during fasting (from 6.84 ± 0.66 to 9.04 ± 0.54 µg/dL; P < 0.03) and 37% during HC infusions (to 9.39 ± 0.79 µg/dL; P < 0.03). Six-hour mean cortisol levels did not differ among the studies from 08001345 h or from 02000745 h. Between 14001945 h, serum cortisol levels increased 42% during fasting (from 5.98 ± 0.58 to 8.52 ± 0.82 µg/dL; P < 0.01) and 58% during HC infusions (to 9.43 ± 0.84 µg/dL; P < 0.01). Between 20000145 h, serum cortisol levels increased 128% during fasting (from 2.50 ± 0.44 to 5.70 ± 1.06 µg/dL; P < 0.003) and 170% during HC infusions (to 6.74 ± 0.72 µg/dL; P < 0.003). Cortisol levels in fasting and HC studies were not statistically different during any time period.
|
|
Twenty-four-hour mean serum TSH levels decreased 65% from baseline during fasting (from 2.33 ± 0.32 to 0.82 ± 0.07 mU/L; P < 0.001) and 51% during HC infusions (to 1.14 ± 0.19 mU/L; P < 0.001). Daytime (08001945 h) TSH levels decreased 55% during fasting (from 1.80 ± 0.25 to 0.81 ± 0.06 mU/L; P < 0.001) and 53% during HC infusions (to 0.84 ± 0.12 mU/L; P < 0.001). Nocturnal (20000745 h) TSH levels decreased 71% during fasting (from 2.85 ± 0.44 to 0.83 ± 0.11 mU/L; P < 0.001) and 49% during HC infusions (to 1.45 ± 0.27 mU/L; P < 0.001). Twenty-four-hour TSH pulse frequency did not change, whereas TSH pulse amplitude was suppressed to a similar degree as mean TSH levels by fasting or HC infusions.
|
|
The normal nocturnal surges in TSH levels and TSH pulse amplitude were evident at baseline, with a 58% increase in mean TSH levels at night (20000800 h, 2.85 ± 0.44; 08001945 h, 1.80 ± 0.25 mU/L; P < 0.001). The nocturnal TSH surge was abolished by fasting (0.83 ± 0.11 vs. 0.81 ± 0.06 mU/L), but not by HC infusions (1.45 ± 0.27 vs. 0.84 ± 0.12 mU/L; a 73% increase; P < 0.01).
Mean TSH responses to TRH were decreased 45% by fasting (from 14.03 ± 1.67 to 7.74 ± 0.75 mU/L; P < 0.01) and 32% by HC infusions (to 9.56 ± 1.53 mU/L; P < 0.01).
Thyroid hormone levels (Table 3
)
Serum total T3 levels were decreased 21% at the end of the fasting studies compared to those at the end of the baseline studies (66.8 ± 7.8 vs. 84.3 ± 5.2 ng/dL; P < 0.05) and were decreased 19% at the end of the HC studies (68.0 ± 8.9 ng/dL; P < 0.05). There were no significant changes in serum free T4 levels at the end of any of the studies.
|
| Discussion |
|---|
|
|
|---|
Using this protocol, we found a 51% decrease in 24-h mean serum TSH
levels, which is close to the 65% decrease observed in the same
subjects during fasting. Decreases in daytime mean TSH levels and TSH
pulse amplitude were identical between the fasting and HC studies (55%
vs. 53%), suggesting that cortisol can completely account
for fasting-induced TSH suppression during the day. This finding is
especially interesting because serum cortisol levels did not increase
between 08001400 h and only increased 24 µg/dL over baseline
between 14002000 h during either study. As evident in Fig. 3
, TSH
suppression was already apparent before these changes in serum cortisol
levels became significant. There are two possible explanations for
this. 1) Figure 2
suggests that there were slight increases in cortisol
levels between 08001400 h during fasting or HC infusions that were
not significant due to small sample size. Such minor cortisol
elevations may be sufficient to suppress TSH secretion. 2) As Fig. 2
shows cortisol elevations were more pronounced at night during either
fasting or HC infusions. Similar nocturnal cortisol elevations probably
occurred during the night before blood drawing started (i.e.
at the end of the first 24 h of HC infusions or fasting) and could
have caused TSH suppression that lasted into the next day (when blood
drawing began).
In contrast to daytime TSH levels, nocturnal TSH levels during HC infusions fell between levels in baseline (fed) and fasting studies. These data suggest that cortisol can account for over 50% of fasting-induced TSH suppression at night, but that other factors must also contribute to TSH suppression during this time. Possible other mediators of TSH suppression include the inflammatory cytokines, dopamine, or somatostatin (9, 10, 11, 12). In the case of dopamine, studies using the dopamine antagonist metoclopramide suggest that dopaminergic tone is not increased during fasting (13). In the case of the cytokines, it appears that these compounds stimulate the hypothalamic-pituitary-adrenal axis (14, 15). Therefore, the changes in thyroid hormone levels observed during increased levels of cytokines may still occur via increased cortisol levels.
In addition to fasting-induced TSH suppression, our study has broader implications for glucocorticoid-induced changes in serum TSH levels. Previous studies reported decreased TSH levels with acute glucocorticoid administration, but at much higher glucocorticoid doses than in the current study (5, 6). Studies in patients with Addisons disease or in healthy subjects given metyrapone suggest that more physiological doses of glucocorticoids decrease TSH levels within the normal range, but these data may be confounded by coexisting thyroid disease in patients with Addisons disease (16, 17, 18, 19, 20, 21, 22, 23). Hangaard et al. performed the most detailed study to date of serum TSH levels during variable infusions of HC in patients with Addisons disease (24). They measured serum TSH levels in subjects given no HC, an infusion of HC that led to normal serum cortisol levels, and an infusion of HC that led to high serum cortisol levels. Serum TSH levels showed a progressive decline with increasing doses of HC in the same range as the results in our study. Together with Hangaards study, our results suggest that low doses of glucocorticoids, which cause either no or mild elevations in serum cortisol levels, are sufficient to suppress TSH secretion. These findings have implications for the interpretation of TSH levels in patients who receive near-physiological doses of glucocorticoids as antiinflammatory treatment or for Addisons disease.
In our study, both fasting and HC infusions blunted TSH responses to exogenous TRH, suggesting that the observed TSH suppression occurs directly at the pituitary level. Previous studies in humans support these findings (1, 5, 6, 25, 26, 27), although other studies suggest that fasting- or glucocorticoid-induced TSH suppression may occur via hypothalamic TRH suppression in animals (28, 29, 30).
The changes in serum T3 levels we measured during fasting or HC infusions are also consistent with published reports, although the current doses of HC are much lower than glucocorticoid doses previously shown to decrease serum T3 levels (1, 2, 5, 31, 32, 33, 34). This suggests that the mild cortisol elevations induced by fasting may mediate the observed decreases in T3 levels. However, it should be noted that alterations in peripheral T4 metabolism induced by fasting may differ from those induced by glucocorticoids. For example, fasting causes decreased rT3 clearance without changes in rT3 production (35), whereas dexamethasone administration causes increased rT3 production without changes in rT3 clearance (36). Thus, the same changes in peripheral thyroid hormone levels may be mediated via different mechanisms in states of fasting or hypercortisolemia.
A potential limitation of our study was that subjects received the same HC dose on both days of the infusion, whereas their cortisol levels were only documented to be in the same range on the second day of the fasting period. Therefore, subjects may have had slightly higher cortisol levels during the first day of HC infusion compared to those on the first day of fasting. Practical difficulties in measuring cortisol levels over 48 h and matching each subjects cortisol levels precluded us from changing the HC dose between the first and second days. However, the overall changes in serum cortisol levels were so mild and the resulting changes in serum TSH levels so marked that we expect this limitation did not markedly affect our results.
In conclusion, by employing a variable infusion of HC tailored to each individual, we were able to mimic the mild increases in serum cortisol levels seen during short term fasting in healthy subjects. These HC infusions led to decreases in 24-h serum TSH levels, TSH responses to TRH, and serum T3 levels similar to those seen during fasting. These results suggest that cortisol may at least in part mediate fasting-induced changes in thyroid hormone levels. In addition, these results prove that marked decreases in serum TSH levels can be induced by near-physiological doses of HC, with resulting changes in serum cortisol levels within the normal range.
| Acknowledgments |
|---|
| Footnotes |
|---|
Received April 3, 1997.
Revised June 26, 1997.
Accepted July 9, 1997.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
W. M. Golden, K. B. Weber, T. L. Hernandez, S. I. Sherman, W. W. Woodmansee, and B. R. Haugen Single-Dose Rexinoid Rapidly and Specifically Suppresses Serum Thyrotropin in Normal Subjects J. Clin. Endocrinol. Metab., January 1, 2007; 92(1): 124 - 130. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Darmon, F. Dadoun, S. Boullu-Ciocca, M. Grino, M.-C. Alessi, and A. Dutour Insulin resistance induced by hydrocortisone is increased in patients with abdominal obesity Am J Physiol Endocrinol Metab, November 1, 2006; 291(5): E995 - E1002. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. M Oltmanns, B. Dodt, B. Schultes, H. H Raspe, U. Schweiger, J. Born, H. L Fehm, and A. Peters Cortisol correlates with metabolic disturbances in a population study of type 2 diabetic patients Eur. J. Endocrinol., February 1, 2006; 154(2): 325 - 331. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Fekete, D. L. Marks, S. Sarkar, C. H. Emerson, W. M. Rand, R. D. Cone, and R. M. Lechan Effect of Agouti-Related Protein in Regulation of the Hypothalamic-Pituitary-Thyroid Axis in the Melanocortin 4 Receptor Knockout Mouse Endocrinology, November 1, 2004; 145(11): 4816 - 4821. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. S. Nylen and B. Muller Endocrine Changes in Critical Illness J Intensive Care Med, March 1, 2004; 19(2): 67 - 82. [Abstract] [PDF] |
||||
![]() |
R. M. Ortiz, D. P. Noren, B. Litz, and C. L. Ortiz A new perspective on adiposity in a naturally obese mammal Am J Physiol Endocrinol Metab, December 1, 2001; 281(6): E1347 - E1351. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. M. Ortiz, C. E. Wade, and C. L. Ortiz Effects of prolonged fasting on plasma cortisol and TH in postweaned northern elephant seal pups Am J Physiol Regulatory Integrative Comp Physiol, March 1, 2001; 280(3): R790 - R795. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Fekete, E. Mihaly, L.-G. Luo, J. Kelly, J. T. Clausen, Q. Mao, W. M. Rand, L. G. Moss, M. Kuhar, C. H. Emerson, et al. Association of Cocaine- and Amphetamine-Regulated Transcript-Immunoreactive Elements with Thyrotropin-Releasing Hormone-Synthesizing Neurons in the Hypothalamic Paraventricular Nucleus and Its Role in the Regulation of the Hypothalamic-Pituitary-Thyroid Axis during Fasting J. Neurosci., December 15, 2000; 20(24): 9224 - 9234. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. H. Samuels Effects of Metyrapone Administration on Thyrotropin Secretion in Healthy Subjects-A Clinical Research Center Study J. Clin. Endocrinol. Metab., September 1, 2000; 85(9): 3049 - 3052. [Abstract] [Full Text] |
||||
![]() |
M. H. Samuels Effects of Variations in Physiological Cortisol Levels on Thyrotropin Secretion in Subjects with Adrenal Insufficiency: A Clinical Research Center Study J. Clin. Endocrinol. Metab., April 1, 2000; 85(4): 1388 - 1393. [Abstract] [Full Text] |
||||
![]() |
L. Plat, R. Leproult, M. LHermite-Baleriaux, F. Fery, J. Mockel, K. S. Polonsky, and E. Van Cauter Metabolic Effects of Short-Term Elevations of Plasma Cortisol Are More Pronounced in the Evening Than in the Morning J. Clin. Endocrinol. Metab., September 1, 1999; 84(9): 3082 - 3092. [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 |