| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Clinical Psychobiology Branch (D.G.B., T.A.W.), National Institute of Mental Health; and Laboratory of Developmental Neurobiology (T.C.F.) and Pediatric and Reproductive Endocrinology Branch (T.C.F., D.H., G.P.C.), National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland 20892; Sleep Disorders Unit, Department of Neurology, Fundacion Jiménez Díaz (D.G.B., O.L.), 28040 Madrid, Spain; Research and Epidemiology Unit (J.J.G.) and Division of Endocrinology, Department of Medicine, Cedars-Sinai Research Institute-University of California School of Medicine (T.C.F.), Los Angeles, California 90048; and Division of Endocrinology, Department of Medicine, Charles R. Drew University of Medicine and Sciences-University of California School of Medicine (T.C.F.), Los Angeles, California 90059
Address all correspondence and requests for reprints to: Theodore C. Friedman, M.D., Ph.D., Division of Endocrinology, Charles R. Drew University of Medicine and Sciences, 1721 East 120th Street, Los Angeles, California 90059. E-mail: friedmant{at}hotmail.com
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
Patients with primary adrenal insufficiency (Addisons disease), because of their lack of endogenous glucocorticoid production, offer researchers a unique opportunity to study the effects of manipulation of the HPA axis on sleep parameters. Prior studies administering exogenous glucocorticoids to normal volunteers were hindered by the difficulty in interpreting the effect of exogenous steroids vs. suppression of the endogenous HPA axis. Additionally, different exogenous steroids exert different effects on sleep (6). In this study we used two paradigms of HPA axis manipulation to study sleep parameters in patients with adrenal insufficiency. In one condition, glucocorticoid replacement was withheld for 1.5 days, thus providing undetectable cortisol, high plasma ACTH, and, presumably, high hypothalamic CRH levels. In the second condition, hydrocortisone replacement was given before bedtime, so that a state of relatively high (for that time of day) cortisol and lower ACTH and CRH was produced. Mineralocorticoid replacement was continued in both situations. A randomized, double blind, cross-over paradigm was used to compare the two situations to each other.
| Subjects and Methods |
|---|
|
|
|---|
The protocol was approved by the NIH clinical review board, and all subjects gave written informed consent before participating in the study. Ten patients [6 females and 4 males; 10 Caucasians; aged 34.8 ± 9.0 yr (mean ± SD); age range, 2349 yr] with primary adrenal insufficiency (Addisons disease) completed the sleep studies. Seven of those patients completed the hormonal sampling protocol. All patients had undetectable cortisol levels and were replaced with proper amounts of fludrocortisone and hydrocortisone. Daily fludrocortisone replacement varied from 0.050.5 mg daily and was monitored by an upright PRA measurement in the normal range. Daily hydrocortisone (HC) replacement dosage ranged between 2030 mg/day and was monitored by 24-h urinary free cortisol and 17-hydroxysteroid excretion within the normal range. Although most patients received their HC replacement in 2 divided doses, for the purposes of the study they received a single morning replacement dose comprised of their total daily dose. Five of 10 patients were receiving thyroid hormone replacement, which continued throughout the study. For comparison of cortisol and ACTH values, 7 age- and sex-matched normal volunteers underwent hormonal sampling every 60 min for the same time period. This control group was taken from our controls previously described (7) and age and sex matched to the current subjects.
Study design
After admission to an in-patient ward at the NIH Clinical Center, an adaptation night (night 0) to the sleep laboratory was carried out. During that night and all subsequent nights, lights were turned off at 2200 h and turned on at 0600 h. On the following day (day 1), medications were administered as usual [including the daily replacement dose of HC (2030 mg), fludrocortisone, and other medications], and a baseline sleep study was carried out, but was not analyzed, to acclimate the subjects to the procedure (adaptation night 1). So as not to exceed the permitted blood volume, no blood was drawn during that night. On day 2, the dose of HC was withheld, whereas fludrocortisone and other medications were administered without interruption. On night 2, at 2200 h, after randomization for the order of treatments, HC or placebo was administered under double blind conditions. After drug intake, lights were turned off, and a sleep study was carried out until 0600 h. On day 3 at 1100 h, administration of the alternative drug condition took place (i.e. the patients who had received HC at 2200 h received placebo at 1100 h, and vice versa). The arm of the study is thus labeled either HC deprived (placebo at bedtime) or HC at bedtime. Because of the short duration of HC deprivation, none of the HC-deprived patients exhibited any signs or symptoms of adrenal insufficiency. One week later, the same patients underwent the study protocol with a reversed order of treatments (i.e. those patients who received HC at 2200 h and placebo at 1100 h during the first week received placebo at 2200 h and HC at 1100 h during the second week).
Sleep stages were identified by means of a continuous recording of the electroencephalogram, vertical and horizontal electrooculogram, and electromyogram. The records were scored off-line according to the criteria of Rechtschaffen and Kales (8). Sleep stages were scored under conditions blind to the order of medication.
Sleep parameters were defined as following: total sleep time, total time asleep; sleep latency, time between lights off and sleep onset (first epoch of stage 2, 3, 4, or REM); wakefulness after sleep onset, time of wakefulness between sleep onset and sleep offset (last epoch of sleep); early morning awakening, time between sleep offset and lights on; sleep efficiency, total sleep time x 100/total recording period; and total recording period, sum of sleep latency, wakefulness after sleep onset, early morning awakening, movement time, disconnecting time, and total sleep time. In addition, REM sleep latency was defined as the time between sleep onset and first epoch of REM sleep, and REM density was defined as the number of rapid eye movements per min of REM sleep.
Hormonal analyses
On night 2, blood was drawn every 30 min between 1700 h (day 2) and 1300 h (day 3) from a separate room by means of an iv line extended through the wall. Blood was transferred to a prechilled ethylenediamine tetraacetate tube and refrigerated until centrifugation. The supernatant was frozen at -20 C until it was assayed for delta sleep-inducing peptide-like immunoreactivity (DSIP-LI), GH, cortisol, and ACTH. DSIP-LI was measured by RIA as previously described (9), using antiserum K-7914, which recognizes the nonapeptide and its phosphorylated and precursor forms. Cortisol (10) and GH (11) were measured by RIA, and ACTH was determined by immunoradiometric assay (12) as previously described. Hormone levels are expressed as the mean ± SD.
Statistical analyses
Sleep parameters for both arms of the study were compared using paired t tests. As hormone levels were not normally distributed, differences in hormone levels between HC at bedtime and HC-deprived conditions were statistically assessed by the Wilcoxon signed ranked nonparametric test for comparison of paired samples.
| Results |
|---|
|
|
|---|
|
|
|
| Discussion |
|---|
|
|
|---|
The major difference between the two experimental conditions (HC
deprived and HC at bedtime) was the amount of circulating nighttime
cortisol. Bedtime administration of HC also reduced plasma ACTH levels
and, most likely, CRH levels. However, after acute administration of
one dose of HC, plasma ACTH levels were still elevated compared with
those in age- and sex-matched normal volunteers (Fig. 1B
). Thus, either
an absolute level of cortisol is permissive for the initiation and
maintenance of REM sleep or, alternatively, in a nonmutually exclusive
manner, declining levels of CRH or ACTH are permissive for REM sleep.
We do not attribute the changes in sleep parameters to a nonspecific
effect of glucocorticoid withdrawal, as the HC-deprived (for <2 days)
patients did not exhibit any symptoms or signs of adrenal
insufficiency.
We also found that plasma levels of two putative sleep hormones, GH and DSIP, were similar throughout the night in the HC-deprived and HC at bedtime conditions, suggesting that the change in sleep parameters was not mediated by either of these two hormones.
Sleep patterns of Addisons patients on and off replacement steroids were previously studied by Gillin et al. (14). In their study of four patients (only three with primary Addisons disease), delta sleep increased during the replacement period. REM parameters, including REM latency, were not different between the two groups. In fact, the REM latency both on replacement (71 min) and off replacement (70 min) was very similar to the REM latency that we observed in HC at bedtime patients (71 min), whereas our sleep latency of 145 min in patients with HC deprivation was quite different from that reported by Gillin et al. (14). The discrepancy in the results of the two studies may be due to differences in the dose, type, and time of replacement therapy or to the small number of patients examined in the previous study.
In normal individuals, exogenous glucocorticoids have been found to reduce REM sleep, whereas continuous administration of iv hydrocortisone may increase non-REM sleep (6, 14, 15, 16, 17). However, there is no unanimity for the latter finding, as clinical and experimental studies have indicated that glucocorticoids are associated with sleep disturbance (18, 19, 20) and non-REM sleep drives the inhibition of pituitary-adrenal secretion in humans (21, 22). We interpret the inhibitory role of glucocorticoids on REM sleep in normal subjects along with the permissive role in Addisons patients as demonstrating that some cortisol is needed for REM sleep, with excess cortisol inhibiting REM sleep. An alternative interpretation is that the decrease in REM sleep in normal subjects receiving exogenous glucocorticoids is caused by CRH and/or ACTH suppression. In Addisons patients, on the other hand, high ACTH and CRH levels are still present despite HC administration, so that CRH and/or ACTH may still signal the entry into REM sleep. Although one study found that iv administration of CRH reduced REM sleep (23), this is probably an indirect effect of CRH, as iv administration of this hormone is unlikely to cross the blood-brain barrier.
Some earlier studies have reported decreasing levels in plasma cortisol during episodes of REM sleep, whereas increasing levels of cortisol were found during slow wave sleep (24, 25). The researchers concluded that REM sleep had an inhibitory effect on adrenal secretion. It is important to realize, however, that cortisol secretion is delayed compared with CRH and ACTH by approximately 30 and 15 min, respectively. Thus, a decline in cortisol during REM sleep may represent a peak in ACTH or CRH secretion 1530 min previously. Born and colleagues showed an increase in plasma cortisol during slow wave sleep before entry into REM sleep (15). This increase in cortisol might represent a preparatory function of each cortisol peak before entry into REM sleep and could be accompanied by declining plasma levels of cortisol once REM sleep started.
In our study hydrocortisone-deprived patients had excess sleep fragmentation, which was reduced by administration of HC at bedtime. Similarly, administration of the glucocorticoid antagonist, mifepristone, increased sleep fragmentation (26). On the other extreme, excessive activity of the HPA axis in conditions such as insomnia (27) or depression (28) is associated with sleep fragmentation. Therefore, it is likely that both extremes of HPA activity might lead to sleep fragmentation.
Finally, untreated or undertreated Addisons patients have excessive daytime fatigue (29). Our results suggest that the decrease in REM sleep and increased awake time after sleep onset in the HC-deprived Addisons patient might contribute to the daytime fatigue of these patients.
| Acknowledgments |
|---|
| Footnotes |
|---|
2 Supported by Center of Clinical Research Excellence Grant
U54-RR-1461601( to Charles R. Drew University of Medicine and
Sciences) and a Culpeper Fellow. ![]()
Received April 3, 2000.
Revised July 18, 2000.
Accepted August 4, 2000.
| References |
|---|
|
|
|---|
sleep-inducing peptide in
humans: evidence for positive correlation with body temperature and
negative correlation with REM and slow wave sleep. J Clin
Endocrinol Metab. 78:10851089.[Abstract]
This article has been cited by other articles:
![]() |
T. M. Buckley and A. F. Schatzberg On the Interactions of the Hypothalamic-Pituitary-Adrenal (HPA) Axis and Sleep: Normal HPA Axis Activity and Circadian Rhythm, Exemplary Sleep Disorders J. Clin. Endocrinol. Metab., May 1, 2005; 90(5): 3106 - 3114. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Baschetti, P. De Becker, and K. De Meirleir Chronic Fatigue Syndrome, Decreased Exercise Capacity, and Adrenal Insufficiency Arch Intern Med, June 25, 2001; 161(12): 1558 - 1559. [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |