| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Clinical Psychobiology Branch (E.L., E.H.T., S.B.A., T.A.W.) and the Behavioral Endocrinology Branch (P.J.S., M.A.D., D.R.R.), National Institute of Mental Health, Bethesda, Maryland 20892
Address all correspondence and requests for reprints to: Dr. Ellen Leibenluft, Room 4S-239, 10 Center Drive, MSC 1390, Bethesda, Maryland 20892-1390.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Our second question was whether gonadal steroids modulate the sleep-activity cycle in men. Two studies indicate that estrogen replacement in postmenopausal women may increase rapid eye movement (REM) sleep (16, 17). The effects of gonadal steroids on the sleep-wake cycle in men have not been studied, although there is evidence that the androgen dehydroepiandrosterone may increase both REM sleep and sleep spindles (18). In addition to considering these direct effects of gonadal steroids on sleep, our third question was whether PRL and TSH could serve as links between the HPG axis and the sleep-activity cycle. PRL secretion is modified by gonadal steroid levels (19, 20) and may alter the microarchitecture of sleep (21, 22, 23). Similarly, the hypothalamic-pituitary-thyroid and HPG axes interact at multiple levels (24, 25), and variations in TSH secretion are associated with fluctuations in the sleep-wake cycle (26).
Although questions concerning the effects of gonadal steroids on circadian regulation or on the sleep-wake cycle can be addressed by studying patients with relevant clinical syndromes, the interpretation of such studies is limited by the inability to establish causal relationships, the variability in gonadal steroid levels, and the question of generalizability. A complementary research strategy involves the use of the GnRH agonist leuprolide acetate (Lupron), to reversibly induce hypogonadism in normal volunteers. When administered for more than 2–4 weeks, Lupron down-regulates GnRH receptors in the pituitary and inhibits gonadotropin secretion, thereby suppressing gonadal steroid secretion (27). As part of a larger study of the effects of gonadal steroids on brain physiology and behavior (Schmidt, P. J., unpublished data), we used Lupron administration and testosterone replacement to study the effects of hypogonadism and gonadal steroid replacement in men on the timing, duration, and microarchitecture of sleep and on the 24-h secretory profiles of melatonin, PRL, and TSH.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We recruited men between the ages of 18–45 yr with unremarkable physical and neurological examinations and normal values on screening laboratory tests (thyroid, liver, and hepatic functions; complete blood count; urinalysis; testosterone and gonadotropin levels; electrolytes; glucose; acid phosphatase; and prostate-specific antigen). The Structured Clinical Interview for DSM-III-R (28) was used to exclude subjects with a current or prior history of Axis I psychiatric disturbance. Individuals performing shift work or suffering from sleep disorders were also excluded. The study was approved by the investigational review board of the National Institute of Mental Health. Subjects gave informed consent for all procedures and were paid for their participation.
Hormonal treatments
Subjects were given Lupron (leuprolide acetate, TAPP Pharmaceuticals, Chicago, IL; 7.5 mg, im, monthly) for 3 months. After 1 month of Lupron alone, testosterone and placebo injections were administered for 1 month each in a double blind, randomized order, cross-over design. Subjects received two injections, 2 weeks apart, of either testosterone enanthate (Gynex Pharmaceuticals, Vernon Hills, IL; 200 mg, im) or sesame oil. Blood samples for gonadal steroid levels were drawn every 2 weeks.
Data collection
Early in the study, all subjects were admitted to an in-patient research unit for an adaptation night, during which they slept with sham electroencephalograph (EEG) leads and a rectal temperature monitor. Subjects were then readmitted for approximately 48 h during week 3 or 4 of each hormonal condition. On day 1 of each procedure, subjects entered the unit at approximately 1700 h and were allowed to choose their own bedtime and awakening time.
EEG recordings were obtained during sleep. Throughout the hospital stay, rectal temperature was measured by an indwelling probe (Yellow Springs Instrument Co., Yellow Springs, OH) connected to a portable electronic monitor that measured temperature every minute (Squirrel model 1206, Science Electronics, Dayton, OH; accuracy and resolution, 0.05 C; range, -50 to 150 C).
On day 2, an iv line was inserted, and subjects were placed in a dim room (<10 lux) at 1100 h. Constant routine conditions (29, 30) were maintained to eliminate the masking effects of sleep, light, and meals on circadian rhythms. In these conditions, subjects remained awake, sitting in a reclining chair, and were fed bland, isocaloric meals every 2 h. Subjects completed the Stanford Sleepiness Scale hourly throughout the procedure (31). Beginning at 1300 h, blood samples to measure melatonin were drawn every 30 min, and blood samples to measure PRL and TSH were drawn every 60 min. Blood for free T4, free T3, total T3, and total T4 was drawn at 1400 h. Samples were stored on ice and centrifuged within 3 h. The plasma was then extracted and frozen at -30 C. In addition to these in-patient procedures, subjects completed a sleep log with 15-min resolution for 2 weeks before the in-patient procedure.
Hormonal assays
Plasma melatonin was measured by RIA at Stockgrand Laboratories (Surrey, UK). The detection limit of the assay was 5 pg/mL. The interassay coefficient of variation (CV) was 9.3% at 11.0 pg/mL, 7.8% at 21.0 pg/mL, 5.3% at 59.9 pg/mL, and 3.1% at 252.0 pg/mL. The intraassay CV was 8.0% at 10.0 pg/mL, 5.5% at 22.0 pg/mL, 4.9% at 57.0 pg/mL, and 2.2% at 246.0 pg/mL (n = 20 for all quality control values). Two subjects had unusually high daytime melatonin levels, and their samples were reassayed using an extraction method.
Plasma PRL, TSH, T3, T4, free T3, and free T4 were measured at Covance Laboratories (Vienna, VA). Detection limits were 0.056 ng/mL, 0.08 µU/mL, 20 ng/dL, 1.0 µg/dL, 0.5 pg/mL, and 0.05 ng/dL, respectively. PRL, TSH, and free T4 were measured by immunoenzymometric assay using reagents supplied by TOSOH Corp. (Tokyo, Japan). For PRL, interassay CVs were 4.6% at 9.2 ng/mL, 6.0% at 18.3 ng/mL, and 9.0% at 39.4 ng/mL. Intraassay CVs were 9.2% at 9.90 ng/mL, 4.8% at 18.7 ng/mL, and 4.2% at 39.6 ng/mL. For TSH, interassay CVs were 6.1% at 1.5 µU/mL, 2.2% at 8.7 µU/mL, and 5.2% at 29.8 µU/mL. Intraassay CVs were 8.4% at 1.53 µU/mL, 2.0% at 8.98 µU/mL, and 2.8% at 30.9 µU/mL. For free T4, interassay CVs were 10.9% at 0.43 ng/dL, 7.7% at 1.62 ng/dL, and 3.9% at 6.15 ng/dL. Intraassay CVs were 10.1% at 0.436 ng/dL, 3.9% at 1.63 ng/dL, and 3.7% at 6.19 ng/dL. T3, T4, and free T3 were measured by RIA using a commercial kit supplied by Coat-A-Count. For T3, interassay CVs were 5% at 75.50 ng/dL, 2% at 148.73 ng/dL, and 4% at 321.42 ng/dL. Intraassay CVs were 8.9% at 56 ng/dL, 5.9% at 166 ng/dL, and 3.1% at 398 ng/dL. For T4, interassay CVs were 9% at 2.64 µg/dL, 5% at 8.06 µg/dL, and 5% at 13.85 µg/dL. Intraassay CVs were 3.8% at 2.4 µg/dL, 2.7% at 7.4 µg/dL, and 2.8% at 13.8 µg/dL. For free T3, interassay CVs were 12% at 2.24 pg/mL, 10% at 4.56 pg/mL, and 4% at 11.28 pg/mL.
Testosterone levels were measured by RIA at Mayo Laboratories (Rochester, NY). Interassay CVs were 16% at 33 ng/dL and 8% at 1003 ng/dL. Intraassay CVs were 10% at 80 ng/dL and 6% at 1257 ng/dL.
Data analysis
The times of melatonin onset and offset were defined as the times of appearance and disappearance (respectively) of sustained detectable melatonin levels, with the duration being the interval between them. The peak value of melatonin was used to determine amplitude (as all subjects had undetectable daytime melatonin levels). Paired t tests were used to compare these variables and to compare levels of free T3, free T4, total T3, and total T4 in the two hormonal conditions. PRL and TSH profiles were analyzed using repeated measures ANOVA with Greenhouse-Geisser corrections.
EEG records were scored for wakefulness and sleep stages in 30-s epochs using conventional criteria by technicians blind to hormonal condition. Sleep logs for the 2 weeks before each in-patient procedure were analyzed by calculating the time of wake onset (the time at the start of the first run of three consecutive 15-min awake intervals from 0300–1200 h), time of sleep onset (the time at the start of the first run of three consecutive 15-min sleep episodes after 1900 h), and sleep duration (number of hours between sleep onset and wake onset). Paired t tests were used to compare sleep data from the two hormonal conditions.
Temperature data were smoothed using a five-point moving average, and the minimum, maximum, and amplitude were determined. The duration of the nocturnal temperature trough was defined as the interval between the downward and upward midrange crossings of the smoothed curve. Data were analyzed using paired t tests.
All means are presented ±SD, and the significance level was set at P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
24-h hormonal profiles
There was no difference between hormonal conditions in the times
of melatonin onset (2133 h ± 1:34 during Lupron plus placebo
vs. 2154 h ± 1:17 during Lupron plus testosterone) or
offset (0833 h ± 0:57 during Lupron plus placebo vs.
0845 h ± 1:30 during Lupron plus testosterone), or in melatonin
amplitude (49.8 ± 20.8 pg/mL during Lupron plus placebo
vs. 51.2 ± 19.5 pg/mL during Lupron plus testosterone;
Fig. 1
). For the PRL analysis, analyzable
data were available from eight subjects (two subjects had large numbers
of missing samples in one of the hormonal conditions). There was a
significant main effect for condition, with the mean value being higher
in the Lupron plus testosterone condition compared with Lupron plus
placebo (mean, 13.4 ± 9.9 vs. 7.8 ± 5.6 ng/mL;
F = 9.6; df = 7; P = 0.02; Fig. 2). Neither the main effect for time nor
the interaction between time and condition was significant. Total
T3 and total T4 were both significantly
increased in the Lupron plus placebo condition compared with those in
the Lupron plus testosterone condition [total T3,
77.1 ± 12.1 vs. 63.0 ± 20.8 ng/dL
(t = 2.36; df = 9; P = 0.04);
total T4, 6.5 ± 1.0 vs. 5.8
± 1.0 mg/dL (t = 2.58; df = 9; P
= 0.03)]. However, there were no significant differences between the
two conditions in levels of free T3 or free
T4 or in TSH secretion.
|
|
EEG data are shown in Table 1.
Because a paper jam occurred during one recording, data are available
from nine subjects. There was a significant increase in stage 4 sleep
(in terms of both minutes and percentage) during the Lupron plus
testosterone condition compared with that during the Lupron plus
placebo condition, with a trend in the opposite direction for stage 3
sleep. There was also a trend for REM latency to lengthen during Lupron
plus testosterone treatment compared with that during Lupron plus
placebo. In the self-reported sleep log data, there was no difference
between conditions in the time of sleep or wake onset or in sleep
duration. There was also no difference between conditions in the
ratings on the Stanford Sleepiness Scale obtained during the constant
routines.
|
Because of probe slippage in two subjects, temperature data are available from eight subjects. There was no difference between conditions in the minimum, maximum, amplitude, or time of the temperature fall. However, there was a trend for the time of the upward midrange crossing to be later in the Lupron plus placebo condition than in the Lupron plus testosterone condition (mean, 0940 h ± 2:25 vs. 0753 h ± 2:12; t = -2.05; df = 7; P = 0.08; 90% confidence interval for the mean difference, 0:08–3:24). Similarly, there was a trend for the duration of the nocturnal temperature trough to be prolonged in the Lupron plus placebo condition compared with that in the Lupron plus testosterone condition (1102 h ± 2:05 vs. 0900 h ± 1:45; t = -2.15; df = 7; P = 0.07; 90% confidence interval for mean difference, 0:14–3:51).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In our study, PRL levels were lower during Lupron plus placebo treatment compared with those during Lupron plus testosterone at all times of the day. The effect of testosterone levels on PRL secretion has received little study, although androgens have been shown to induce PRL production by human endometrial cells in vitro (19). Estrogen appears to act at a number of physiological levels to facilitate PRL secretion (20), so a possible mechanism for our finding could involve the aromatization of testosterone to estradiol. The only study comparable to ours is that by Stoffel-Wagner et al. (32), in which PRL levels were measured in male control subjects before, during, and after treatment with a GnRH agonist. In that study, PRL levels were elevated during the posttreatment period at a time when testosterone levels were high normal. Thus, the data reported by both Stoffel-Wagner et al. (32) and our group are consistent with the interpretation that testosterone and/or its metabolites facilitate PRL secretion.
Increased PRL secretion in the Lupron plus testosterone condition may
be physiologically linked with our second finding, a significant
increase in the amount of stage 4 sleep in the testosterone-replaced
condition. This finding, combined with the trend we observed toward a
decrease in stage 3 sleep during Lupron plus testosterone, indicates
that although the total amount of slow wave (i.e. stages 3
and 4) sleep is unchanged between the two hormonal conditions, 
counts per min are probably higher in the testosterone-replaced state.
A recent study using spectral analysis of EEG recordings found a
positive correlation between PRL secretory rates (calculated by a
deconvolution procedure) and relative power in the band in men
(23). Because the interval between PRL and -wave pulses in this
study (23) ranged from 0–20 min (with PRL secretion lagging behind
sleep), these researchers suggest that common regulatory mechanisms are
involved in the two phenomena. Our results, in which increased PRL and
increased stage 4 sleep are both associated with the
testosterone-replaced state, indicate that testosterone levels may
modulate the set-point of this common regulatory mechanism.
Additionally, one can speculate that age-related declines in
testosterone levels in normal men may contribute to the decline in slow
wave sleep (33) that is seen with age.
Our study revealed no differences in melatonin and TSH secretion between the two hormonal conditions. As free T3 and free T4 levels did not differ with hormonal state, the differences we detected in total T3 and total T4 levels are probably secondary to alterations in binding protein levels. With respect to melatonin, the lack of a baseline melatonin profile means that suppression of endogenous gonadotropin secretion by Lupron may have caused melatonin secretion to increase (relative to baseline) in both conditions. However, the amplitude of the melatonin profiles that we obtained is well within the range of values commonly seen in control subjects (34), so this possibility seems unlikely.
Therefore, the most probable conclusion from our results is that gonadal steroid manipulations in postpubertal males do not affect the profile of nocturnal melatonin secretion. What is the relevance of this negative finding to the clinical literature, which shows a possible association between CPP and low melatonin levels and between high melatonin levels and hypogonadism (see introduction)? Our experimental design cannot speak to one possible explanation for these clinical associations; namely, that melatonin has antigonadal effects in humans. Alternatively, these clinical associations might indicate that gonadal steroid manipulations can alter melatonin secretion, a possibility that is addressed both by treatment studies in patients and by our study. The former have produced inconsistent results. For example, although Luboshitzky et al. (35) reported a decline in melatonin levels after testosterone treatment in men with idiopathic hypogonadotropic hypogonadism, these findings were not replicated by Ozata et al. (9). However, the latter study relied on a single daytime measurement of plasma melatonin (36, 37). More germane to our results (because they involve pharmacological attempts to reverse puberty) are studies in which GnRH agonists were used to treat patients (mostly girls) with CPP. In these studies, GnRH agonists clearly caused pituitary-gonadal suppression, but melatonin levels [which in at least some studies were reported to be abnormally low at baseline (10, 11, 12)] either did not change (38) or declined further (10). One possible explanation for these seemingly disparate results is that maturity in the hypothalamic-pineal system is a one-way street; although treatment with gonadal steroids might lower melatonin levels in patients who have never undergone puberty (35), neither suppression of precocious puberty (10, 38) nor pharmacologically induced hypogonadism (as in our study) restores melatonin to its high prepubertal levels.
In summary, we found that testosterone replacement had relatively discrete effects on sleep and hormonal rhythms in hypogonadal men. Testosterone replacement caused robust elevations in PRL levels as well as increases in the time and percentage of stage 4 sleep; as discussed above, these two findings may be physiologically linked. However, pharmacologically induced hypogonadism and testosterone replacement did not have significant effects on melatonin or TSH secretion or on the timing or duration of sleep.
Received March 24, 1997.
Revised June 12, 1997.
Accepted June 23, 1997.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  P. Y. Liu, B. Yee, S. M. Wishart, M. Jimenez, D. G. Jung, R. R. Grunstein, and D. J. Handelsman The Short-Term Effects of High-Dose Testosterone on Sleep, Breathing, and Function in Older Men J. Clin. Endocrinol. Metab., August 1, 2003; 88(8): 3605 - 3613. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Matsumoto Andropause: Clinical Implications of the Decline in Serum Testosterone Levels With Aging in Men J. Gerontol. A Biol. Sci. Med. Sci., February 1, 2002; 57(2): M76 - 99. [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 |
