Effects of Thyrotropin-Releasing Hormone on Sleep and Sleep-Related Growth Hormone Release in Normal Subjects KAZUO CHIHARA, YUZURU KATO, KIYOSHI MAEDA,* HIROMI ABE, MASARU FURUMOTO, AND HIROO IMURA Third Division, Department of Medicine, and Department of Neuropsychiatry,* Kobe University School of Medicine, Kobe, Japan 1.3 ± 0.8%, P < 0.001). Plasma GH increased in close relationship to the initial appearance of slow wave sleep (SWS) within 40 min after sleep onset on both control nights in all four subjects. On nights of TRH administration, however, plasma GH levels during the initial 80 min of sleep were significantly lower (P < 0.005) than on control nights, whereas SWS was demonstrated before the interruption of sleep. On nights when sleep was interrupted by forced wakefulness 1 h after sleep onset, plasma GH rose to levels comparable to those on control nights during early sleep periods in all subjects examined. These results suggest that TRH inhibits sleep and sleep-related GH release in normal subjects. (J Clin Endocrinol Metab 44: 1094, 1977)
ABSTRACT. Effects of TRH on sleep and sleeprelated growth hormone (GH) release were examined in four normal volunteers. A bolus of 500 /xg of synthetic TRH was injected iv at the onset of sleep, followed by continuous iv infusion of 1000 fig of TRH dissolved in saline for 3 h on two nights. Saline alone was infused on two control nights in each of these subjects. Polygraphic sleep records showed that TRH transiently interrupted sleep on both nights in all of the four subjects. The arousal phenomenon was observed from 80 to 151 min after the start of TRH administration until 20 to 212 min after the end of TRH infusion. The mean (±SE) percentage of awakening on the nights of TRH administration was significantly larger than on the control nights (36.4 ± 1.9% us.
T
HYROTROPIN-RELEASING hormone (TRH), a potent stimulator of both thyrotropin (TSH) and prolactin release (1,2), has recently been documented to affect growth hormone (GH) secretion in man. The iv injection of TRH raises plasma GH in some patients with acromegaly (3,4), uremia (5), mental depression (6) and anorexia nervosa (7). In contrast, we have previously reported that GH release caused by L-Dopa, arginine or insulin-induced hypoglycemia is significantly inhibited by the concomitant administration of TRH in normal subjects (8,9). It is well known that plasma GH increases during the early period of nocturnal sleep in man (10-12). The nocturnal GH release is closely related to sleep, which is regulated by the central nervous system. Since there has been accumulating evidence suggesting that TRH influenced the central nervous system in man and in aniReceived September 26, 1976. Supported in part by grants from the Ministry of Education and the Ministry of Welfare, Japan.
mals, we examined its effect on sleep and sleep-related GH release in normal subjects.
Materials and Methods Four male volunteers, aged 19 to 21, were used throughout the experiment. They were non-obese (less than 10% above the ideal body weight) and apparently normal in endocrine and autonomic nervous function. None of them took any drugs regularly. After a light supper around 1700 h on the day of the experiment, they lay down in bed in our sleep laboratory, and were not allowed to take anything by mouth. Electroencephalogram (EEG), electrooculogram (EOG) and electromyogram (EMG) were recorded by standard techniques, and sleep was scored as stage 1, 2, 3, 4 and stage of rapid eye movement (REM) according to the criteria of Kales and Rechtschaffen (13). An indwelling needle (18 gauge) was inserted into an antecubital vein under local anesthesia and connected to an infusion apparatus. Another indwelling needle for blood withdrawal was inserted distally into another vein of the same arm, as previously described (14). Each of 4 subjects was examined on five
1094
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 17:55 For personal use only. No other uses without permission. . All rights reserved.
EFFECTS OF TRH ON SLEEP AND GH RELEASE
1095
TABLE 1. Effect of TRH administration on sleep stages Percentage of sleep stage in first 8 h following sleep onset
Control (4)f Night with TRH administered after sleep onset (4)
Awake
1
2
3
1.3 ± 0.8J
3.1 ± 1.7
53.9 ± 3.8
36.4 ± 1.9*
6.3 ± 1.4
32.3 ± 2.0**
4
REM
S.2 ± 1.3
J.I ± 1.6
25.5 ± 0.4
10.3 ± 1.6
8.2 ± 1.9
6.6 ± 0.6*
Statistical significance os. control night: * P < 0.001, ** P < 0.005. f Numbers of subjects are shown in parentheses. \ Values, shown as mean ± SE, were calculated using the average value of each subjects 2 nights. nights which were separated at an interval of 1 to 2 weeks. Saline solution was infused at a rate of 1 nil per min, which served as control periods. On the third and the fourth night, 500 /xg of synthetic TRH (Tanabe Co., Osaka, Japan) dissolved in 1 nil of saline was injected iv as a bolus through the needle at the onset of sleep, which was defined as the first entry into
stage 2 sleep (13), followed by constant iv infusion of 1000 fxg of TRH dissolved in 180 nil of saline over a period of 3 h. On the fifth night, subjects were forcedly awakened 1 h after the onset of sleep, and were not allowed to fall asleep for the next 3 h. Blood samples were withdrawn every 20 min through the latter needle into heparinized Subject # 2
FlC. 1A. Plasma GH concentrations after the onset of sleep (entry into stage 2) during control nights and nights of TRH administration in subjects #1 and #2. The sleep histogram is shown beneath each panel: open histograms (CD) show slow wave stages 3 and 4, closed histograms ( • • ) REM stage and solid line ( ) stages 1 and 2. Breaks in lines represent waking periods. Stippled areas indicate periods of TRH infusion. (A) and (B) show the first and second control nights, respectively. (C) and (D) show the first and second nights of TRH study.
Sleep
Onset
Sleep Onset
Hours after Sleep Onset
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 17:55 For personal use only. No other uses without permission. . All rights reserved.
1096
CHIHARA ET AL.
JCE & M • 1977 Vol 44 • No 6
Subject
Subject # 3
FIG IB. Plasma GH concentrations during control nights (A, B) and nights of TRH administration (C, D) in subjects #3 and #4. Legends for sleep histogram and stippled areas are the same as in Fig. 1A.
Sleep Onset
Sleep Onset
Hours after Sleep Onset
syringes for 8 h after sleep onset on all five nights of the experiment. Plasma was promptly separated and frozen at - 2 0 C until assayed. Plasma GH was measured by the double antibody radioimmunoassay described previously (15) using highly purified human GH (NIH HS 1863) as the standard. Statistical * analysis was performed by Student's t test, by comparing the mean of each individual's average values on 2 control or experimental nights.
In contrast, on the nights of TRH study, all subjects awoke 80 to 151 min after sleep onset with a mean (±SE) of 110 ± 10 min (Figs. 1A and B). Nobody fell asleep again until 20 to 212 min after the end of TRH infusion, which corresponded to 106 to 203 min of sleep interruption with a mean (±SE) of 165 ± 1 1 min. Statistical analysis of polygraphic EEG sleep stages showed that the mean percentage of time awake during 8 h after sleep onset was signifiResults cantly (P < 0.001) greater on the study nights than on the control nights (Table 1). Effect of TRH on sleep EEG stages There was no significant difference beOn the control nights, all subjects slept tween the experimental and control nights in well with negligible periods of awakening, the content of slow wave sleep (SWS) during and the time spent in each sleep stage was the initial 3 h of sleep, which was deterwithin the normal limits of this laboratory, mined every 20 min following the first onas shown in Table 1. set of sleep, as shown in Fig. 2. The mean
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 17:55 For personal use only. No other uses without permission. . All rights reserved.
1097
EFFECTS OF TRH ON SLEEP AND GH RELEASE (±SE) time spent in SWS during the initial 3 h of sleep on the nights of TRH administration was not significantly different from that on the control nights (44 ± 7 min vs. 58 ± 7 min, respectively). On the nights of TRH study, SWS was greatly increased during the early period of the second sleep following a period of awakening. However, the mean percentage of SWS during 8 h after the first sleep onset was not significantly larger on the study nights than on the control nights (Table 1). Effect of TRH on sleep-related GH secretion Plasma GH levels were significantly increased during early sleep periods on both the first and the second control nights in all four subjects examined (Figs. 1A and B). The first elevation of plasma GH (more than 5.0 ng/ml) was noted 20 to 39 min after the onset of sleep, which was closely related to the first occurrence of SWS. The peak value of plasma GH varied from 15.4 to 36.0 ng/ml with a mean (±SE) of 24.0 ± 2.8 ng/ml on the control nights. The pattern of GH release during early sleep periods in each subject was very similar on both control nights. On both nights of TRH administration in all four subjects, no significant increase in plasma GH was demonstrated within 40 min after sleep onset, although it rose on the control nights. The first significant rise of plasma GH appeared 46 min after sleep onset in one case (Fig. 1A, subject #2, C) and as late as 80 to 120 min on three other determinations (Fig. IB, subject #3, C: subject #4, C and D). In the four other determinations, plasma GH levels remained low throughout the period of TRH infusion. Mean plasma GH levels during the first 60 min period on the study nights were significantly lower (P < 0.025) than on the control nights (Fig. 2). However, when subjects fell asleep again following the waking period induced by TRH, a definite elevation of plasma GH was observed within 40 min after the second onset of sleep on both nights. The peak
t
2
3 - 1
A
Second Sleep Onset
Sleep Onset Time
in
Hours
FIG 2. Effects of TRH on sleep-induced GH rises, slow wave sleep (stage 3 + 4) time and waking time in 4 subjects. The solid line (O O) and dark column show plasma GH values and time spent in slow wave sleep and waking during control nights, respectively. The broken line (O O) and white column show plasma GH values and time spent in slow wave sleep and waking during nights when TRH was infused after sleep onset. Values, shown as mean ± SE, were calculated using the average value of each subject's 2 night. The number of subjects is indicated in parentheses. Statistical differences between these two groups are shown by asterisks: * P < 0.005, **P
ABSTRACT. Effects of TRH on sleep and sleeprelated growth hormone (GH) release were examined in four normal volunteers. A bolus of 500 /xg of synthetic TRH was injected iv at the onset of sleep, followed by continuous iv infusion of 1000 fig of TRH dissolved in saline for 3 h on two nights. Saline alone was infused on two control nights in each of these subjects. Polygraphic sleep records showed that TRH transiently interrupted sleep on both nights in all of the four subjects. The arousal phenomenon was observed from 80 to 151 min after the start of TRH administration until 20 to 212 min after the end of TRH infusion. The mean (±SE) percentage of awakening on the nights of TRH administration was significantly larger than on the control nights (36.4 ± 1.9% us.
T
HYROTROPIN-RELEASING hormone (TRH), a potent stimulator of both thyrotropin (TSH) and prolactin release (1,2), has recently been documented to affect growth hormone (GH) secretion in man. The iv injection of TRH raises plasma GH in some patients with acromegaly (3,4), uremia (5), mental depression (6) and anorexia nervosa (7). In contrast, we have previously reported that GH release caused by L-Dopa, arginine or insulin-induced hypoglycemia is significantly inhibited by the concomitant administration of TRH in normal subjects (8,9). It is well known that plasma GH increases during the early period of nocturnal sleep in man (10-12). The nocturnal GH release is closely related to sleep, which is regulated by the central nervous system. Since there has been accumulating evidence suggesting that TRH influenced the central nervous system in man and in aniReceived September 26, 1976. Supported in part by grants from the Ministry of Education and the Ministry of Welfare, Japan.
mals, we examined its effect on sleep and sleep-related GH release in normal subjects.
Materials and Methods Four male volunteers, aged 19 to 21, were used throughout the experiment. They were non-obese (less than 10% above the ideal body weight) and apparently normal in endocrine and autonomic nervous function. None of them took any drugs regularly. After a light supper around 1700 h on the day of the experiment, they lay down in bed in our sleep laboratory, and were not allowed to take anything by mouth. Electroencephalogram (EEG), electrooculogram (EOG) and electromyogram (EMG) were recorded by standard techniques, and sleep was scored as stage 1, 2, 3, 4 and stage of rapid eye movement (REM) according to the criteria of Kales and Rechtschaffen (13). An indwelling needle (18 gauge) was inserted into an antecubital vein under local anesthesia and connected to an infusion apparatus. Another indwelling needle for blood withdrawal was inserted distally into another vein of the same arm, as previously described (14). Each of 4 subjects was examined on five
1094
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 17:55 For personal use only. No other uses without permission. . All rights reserved.
EFFECTS OF TRH ON SLEEP AND GH RELEASE
1095
TABLE 1. Effect of TRH administration on sleep stages Percentage of sleep stage in first 8 h following sleep onset
Control (4)f Night with TRH administered after sleep onset (4)
Awake
1
2
3
1.3 ± 0.8J
3.1 ± 1.7
53.9 ± 3.8
36.4 ± 1.9*
6.3 ± 1.4
32.3 ± 2.0**
4
REM
S.2 ± 1.3
J.I ± 1.6
25.5 ± 0.4
10.3 ± 1.6
8.2 ± 1.9
6.6 ± 0.6*
Statistical significance os. control night: * P < 0.001, ** P < 0.005. f Numbers of subjects are shown in parentheses. \ Values, shown as mean ± SE, were calculated using the average value of each subjects 2 nights. nights which were separated at an interval of 1 to 2 weeks. Saline solution was infused at a rate of 1 nil per min, which served as control periods. On the third and the fourth night, 500 /xg of synthetic TRH (Tanabe Co., Osaka, Japan) dissolved in 1 nil of saline was injected iv as a bolus through the needle at the onset of sleep, which was defined as the first entry into
stage 2 sleep (13), followed by constant iv infusion of 1000 fxg of TRH dissolved in 180 nil of saline over a period of 3 h. On the fifth night, subjects were forcedly awakened 1 h after the onset of sleep, and were not allowed to fall asleep for the next 3 h. Blood samples were withdrawn every 20 min through the latter needle into heparinized Subject # 2
FlC. 1A. Plasma GH concentrations after the onset of sleep (entry into stage 2) during control nights and nights of TRH administration in subjects #1 and #2. The sleep histogram is shown beneath each panel: open histograms (CD) show slow wave stages 3 and 4, closed histograms ( • • ) REM stage and solid line ( ) stages 1 and 2. Breaks in lines represent waking periods. Stippled areas indicate periods of TRH infusion. (A) and (B) show the first and second control nights, respectively. (C) and (D) show the first and second nights of TRH study.
Sleep
Onset
Sleep Onset
Hours after Sleep Onset
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 17:55 For personal use only. No other uses without permission. . All rights reserved.
1096
CHIHARA ET AL.
JCE & M • 1977 Vol 44 • No 6
Subject
Subject # 3
FIG IB. Plasma GH concentrations during control nights (A, B) and nights of TRH administration (C, D) in subjects #3 and #4. Legends for sleep histogram and stippled areas are the same as in Fig. 1A.
Sleep Onset
Sleep Onset
Hours after Sleep Onset
syringes for 8 h after sleep onset on all five nights of the experiment. Plasma was promptly separated and frozen at - 2 0 C until assayed. Plasma GH was measured by the double antibody radioimmunoassay described previously (15) using highly purified human GH (NIH HS 1863) as the standard. Statistical * analysis was performed by Student's t test, by comparing the mean of each individual's average values on 2 control or experimental nights.
In contrast, on the nights of TRH study, all subjects awoke 80 to 151 min after sleep onset with a mean (±SE) of 110 ± 10 min (Figs. 1A and B). Nobody fell asleep again until 20 to 212 min after the end of TRH infusion, which corresponded to 106 to 203 min of sleep interruption with a mean (±SE) of 165 ± 1 1 min. Statistical analysis of polygraphic EEG sleep stages showed that the mean percentage of time awake during 8 h after sleep onset was signifiResults cantly (P < 0.001) greater on the study nights than on the control nights (Table 1). Effect of TRH on sleep EEG stages There was no significant difference beOn the control nights, all subjects slept tween the experimental and control nights in well with negligible periods of awakening, the content of slow wave sleep (SWS) during and the time spent in each sleep stage was the initial 3 h of sleep, which was deterwithin the normal limits of this laboratory, mined every 20 min following the first onas shown in Table 1. set of sleep, as shown in Fig. 2. The mean
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 08 July 2015. at 17:55 For personal use only. No other uses without permission. . All rights reserved.
1097
EFFECTS OF TRH ON SLEEP AND GH RELEASE (±SE) time spent in SWS during the initial 3 h of sleep on the nights of TRH administration was not significantly different from that on the control nights (44 ± 7 min vs. 58 ± 7 min, respectively). On the nights of TRH study, SWS was greatly increased during the early period of the second sleep following a period of awakening. However, the mean percentage of SWS during 8 h after the first sleep onset was not significantly larger on the study nights than on the control nights (Table 1). Effect of TRH on sleep-related GH secretion Plasma GH levels were significantly increased during early sleep periods on both the first and the second control nights in all four subjects examined (Figs. 1A and B). The first elevation of plasma GH (more than 5.0 ng/ml) was noted 20 to 39 min after the onset of sleep, which was closely related to the first occurrence of SWS. The peak value of plasma GH varied from 15.4 to 36.0 ng/ml with a mean (±SE) of 24.0 ± 2.8 ng/ml on the control nights. The pattern of GH release during early sleep periods in each subject was very similar on both control nights. On both nights of TRH administration in all four subjects, no significant increase in plasma GH was demonstrated within 40 min after sleep onset, although it rose on the control nights. The first significant rise of plasma GH appeared 46 min after sleep onset in one case (Fig. 1A, subject #2, C) and as late as 80 to 120 min on three other determinations (Fig. IB, subject #3, C: subject #4, C and D). In the four other determinations, plasma GH levels remained low throughout the period of TRH infusion. Mean plasma GH levels during the first 60 min period on the study nights were significantly lower (P < 0.025) than on the control nights (Fig. 2). However, when subjects fell asleep again following the waking period induced by TRH, a definite elevation of plasma GH was observed within 40 min after the second onset of sleep on both nights. The peak
t
2
3 - 1
A
Second Sleep Onset
Sleep Onset Time
in
Hours
FIG 2. Effects of TRH on sleep-induced GH rises, slow wave sleep (stage 3 + 4) time and waking time in 4 subjects. The solid line (O O) and dark column show plasma GH values and time spent in slow wave sleep and waking during control nights, respectively. The broken line (O O) and white column show plasma GH values and time spent in slow wave sleep and waking during nights when TRH was infused after sleep onset. Values, shown as mean ± SE, were calculated using the average value of each subject's 2 night. The number of subjects is indicated in parentheses. Statistical differences between these two groups are shown by asterisks: * P < 0.005, **P
