0021-972X/91/7201-0090$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1991 by The Endocrine Society
Vol. 72, No. 1 Printed in U.S.A.
Dose-Dependent Effects of Scopolamine on Nocturnal Growth Hormone Secretion in Normal Adult Men: Relation to 5-Sleep Changes* JAMES T. McCRACKEN, RUSSELL E. POLAND, ROBERT T. RUBIN, AND LEONARDO TONDO Department of Psychiatry, University of California-Los Angeles School of Medicine, Harbor-University of California-Los Angeles Medical Center (J.T.M., R.E.P., R.T.R.), Torrance, California 90509; and the Department of Psychiatry, University of Cagliari (L. T.), Cagliari, Italy
a significantly greater delay observed after the larger dose. Similarly, a significant delay in the time of the GH rise was produced by SCOP. In contrast, the effects of both doses of SCOP on 5-sleep or sleep onset were small. These data confirm earlier reports demonstrating that cholinergic muscarinic input represents a potentially important source of regulation of nocturnal GH release and suggest that the magnitude of the reduction in GH and the extent of delay in the GH rise time may reflect quantitative differences in the degree of cholinergic blockade. These data are in agreement with recent studies suggesting that the timing of GH release need not be associated with 6sleep per se. (J Clin Endocrinol Metab 72: 90-95, 1990)
ABSTRACT. To explore the sensitivity of nocturnal GH secretion to different degrees of cholinergic blockade, we investigated the effects of two doses of the muscarinic receptor antagonist scopolamine (SCOP; 3.0 and 6.0 Mg/kg, im) and placebo, administered in a randomized fashion at 2300 h on three nights to eight normal male volunteers. Both doses of SCOP produced significant reductions in mean nocturnal GH concentration compared to the effects of the placebo; the higher dose of SCOP reduced GH to a greater degree than the lower dose, but this difference was not statistically significant (mean, 2.3 Mg/L after 6 Mg/kg vs. 3.0 jug/L after 3 Mg/kg). Both SCOP doses significantly shifted GH secretion into later portions of the night, with
T
ies have shown that nocturnal GH secretion can be abolished by the muscarinic antagonists methscopolamine (2), atropine (10), and pirenzepine (11), but whether the reduction is a consequence of suppression or delay of release, or both, is unknown. Since only one dose of a cholinergic antagonist was administered in these three studies, and GH secretion was completely abolished, it is unclear whether the degree of reduction of nocturnal GH secretion could reflect the degree of cholinergic blockade. Because of our interest in clinical studies of possible cholinergic abnormalities in severe depression, we elected to investigate whether nocturnal GH release would be sensitive to varying levels of muscarinic blockade by the muscarinic cholinergic antagonist scopolamine (SCOP). Our study also enabled us to examine the relationship between slow wave (5) sleep and GH secretion. This was of interest in view of reports suggesting that the apparent linkage of 5-sleep and GH release (12, 13) might, in fact, be fortuitous (14-16). We have previously described the effects of SCOP on allnight sleep architecture and nocturnal cortisol secretion in these subjects (17).
HE INFLUENCE of the cholinergic system on GH secretion has been appreciated for some time, beginning with early reports demonstrating daytime GH release after cholinergic agonist administration and reduction of sleep-related GH secretion after administration of medication with potent anticholinergic properties (1). Other studies have shown with more specificity that cholinergic muscarinic receptor antagonists, such as methylscopolamine, atropine, and pirenzepine, block the GH rise due to insulin-induced hypoglycemia (2, 3), clonidine, exercise, arginine (4), the enkephalin analog FK 33-824 (5), glucagon, L-dopa, apomorphine (6), and GH-releasing hormone (7-9). However, it has been suggested that the neurochemical regulation of nocturnal GH secretion may differ from that of GH release during wakefulness induced by pharmacological or physiological stimulation (1). Three stud-
Received December 5, 1989. Address requests for reprints to: Dr. James T. McCracken, Department of Psychiatry, Harbor-University of California-Los Angeles Medical Center, 1000 West Carson Street, Building F-5, Torrance, California 90509. * This work was supported by NIMH Grants MH-34471 (to R.E.P.) and MH-28380 (to R.T.R.), Physician Scientist Award MH-00722 (to J.T.M.), Research Scientist Development Award MH-00534 (to R.E.P.), Research Scientist Award MH-47363 (to R.T.R.), and NIH Clinical Research Center Grant RR-00425.
Subjects and Methods Eight normal male volunteers, aged 19-45 yr, gave their written informed consent to participate in the study. All sub90
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 02:48 For personal use only. No other uses without permission. . All rights reserved.
EFFECTS OF SCOP ON GH SECRETION jects were admitted three times, for 2 consecutive nights each time, to the Clinical Research Center at Harbor-University of California-Los Angeles Medical Center at approximately weekly intervals. All subjects were previously judged to be medically and psychologically healthy, as determined by a physical examination, electrocardiogram, full blood chemistry panel, thyroid function tests, urine drug screen, and clinical psychiatric interview. No subject had a past or present psychiatric illness or a history of major psychiatric disorder in any first degree relative. For each admission, the subjects spent 2 consecutive nights in the hospital, with the first night considered an adaptation night. On all nights, physiological sleep data (electroencephalogram, electrooculogram, and electromyogram) were recorded using Nihon Kohden polygraphs. Conventional electroencephalogram electrodes were attached at 1900 h, and recordings were made from 2300-0700 h. Lights out was from 2300-0700 h. On the second night of each 2-night pair, an iv catheter was inserted into the medial cubital vein at 1800 h and kept patent by periodic flushing with heparinized saline. A 10-ft extension allowed blood sampling to be performed from an adjacent room. Blood samples (5.0 mL) were taken every 15 min beginning at 2230 h and continuing until 0700 h. Also, on the second night of each admission at 2300 h just before lights out, the subjects were given an im injection of saline or SCOP (3.0 or 6.0 ng/ kg). The three doses were administered in a double blind, randomized fashion. Coded sleep records were blindly scored according to standard criteria (18). Sleep onset was defined as the first minute of stage 2 sleep, followed by at least 9 min of stage 2 sleep not interrupted by more than 1 min of waking or stage 1. 5-Sleep (stages 3 and 4) minutes were calculated for each third of the night for the 8-h recording period for each subject. 6-Sleep onset was considered the first epoch of slow wave sleep during the recording period. The blood samples were immediately placed on ice and centrifuged, and aliquots were made before storage at —20 C. Plasma GH was measured by a double antibody RIA (19), with materials obtained from Dr. Albert F. Parlow and the National Pituitary Agency. RP-1 was used as the reference preparation and for iodination. GH was iodinated by the glucose oxidaselactoperoxidase method (20), and the suitability of the iodinated hormone was determined by the talc-resin-trichloroacetic acid method (21). All samples from the 3 nights for each subject were analyzed in duplicate in the same assay. Analysis of low, medium, and high GH plasma pools in each assay showed the maximum intra- and interassay variabilities to be 9.3% and 11.9%, respectively. Statistical analysis of the sleep and hormone data was performed by repeated measures analysis of variance (ANOVA) and paired t tests. Because the hormone values were not Gaussian distributed across subjects, they were log transformed before analysis, which normalized the distributions. All reported significance levels are two tailed. The mean plasma GH concentration was calculated as the average of the plasma GH concentrations measured every 15 min from 2300 h until 0700 h. Means for the first, second, and last thirds of the night were calculated as the averages of the GH values every 15 min from 2300-0130,0145-0415, and 0430-0700 h, respectively. The time of the GH peak was determined as the time of the sample with
91
the maximum GH concentration for that night. The time of the GH rise was calculated according to the method of Jarrett et al. (22), The GH nadir was calculated as the mean of the three lowest sequential values during the night before a GH peak, and GH rise time was taken as that time after the nadir when the GH concentration first exceeded the nadir plus 2 SD of the nadir. Analyses of the effects of SCOP on pulse frequency and amplitude of GH secretion were also performed. Pulses of GH were identified as described previously (23). An increase in GH concentration was considered a significant pulse if both its increment and decline exceeded twice the intraassay coefficient of variation for that concentration range. Increases or decreases below 2.0 Mg/L were not considered significant. The absolute amplitude of a pulse was determined by calculating the difference between the maximum concentration of the pulse and the minimum concentration of the trough preceding that pulse. The average pulse amplitude was considered the mean pulse amplitude of all pulses for that night.
Results SCOP effects on mean plasma GH concentrations Table 1 shows the effects of the two doses of SCOP compared to that of saline on nocturnal GH for the eight subjects. Repeated measures ANOVA revealed a highly significant treatment effect on the log-transformed 23000700 h mean GH concentrations (F2(23 = 10.69; P < 0.002). Both the 3.0 and the 6.0 Atg/kg doses of SCOP produced a significant reduction in the mean GH concentration compared to saline (mean, 3.0 and 2.3 vs. 4.3 Mg/L, respectively; t = 3.99, P < 0.006 for saline us. 3 Mg/ kg; t = 4.963, P < 0.002 for saline vs. 6 ixg/kg). There was a slight trend for a greater reduction of mean GH after the 6 /ug/kg compared to the 3.0 /ig/kg dose, but the difference was not statistically significant (t = 1.147; P > 0.29). SCOP effects on GH pulse frequency and amplitude As seen in Table 2, the high dose of SCOP produced a significant reduction in the number of GH pulses during TABLE 1. Effects of SCOP on mean (±SEM) 2300-0700 h GH concentrations (micrograms per L) in eight normal volunteers Subject no.
Saline
SCOP
SCOP
1 2 3 4 5 6 7 8
4.5 3.8 3.4 1.8 5.3 7.0 2.2 6.7
1.9 2.3 2.4 1.7 3.1 7.3 1.4 4.0
4.1 2.3 1.8 1.6 2.1 3.2 1.1 2.5
Mean"
4.3
3.0*
2.3 C
°F 2 , 23 = 10.69; P < 0.002. 1 = 3.99; P < 0.006 us. saline. c t = 4.963; P < 0.002 vs. saline. 6
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 02:48 For personal use only. No other uses without permission. . All rights reserved.
MCCRACKEN ET AL.
92
TABLE 2. Effects of SCOP on mean (±SEM) GH pulse frequency and the amplitude (micrograms per L) of the first pulse between 2300-0700 h in eight normal volunteers Saline Subject Pulse no. no. 1 2 3 4 5 6 7 8
3 2 2 3 3 2 2 2
3.0 Mg/kg SCOP
6.0 Mg/kg SCOP
Pulse Pulse Pulse Pulse Pulse amplitude no. amplitude no. amplitude 15.6 16.3 14.0 1.5
23.3 25.2 6.2
33.0
1 3 2 2 2 1 2 3
1 2 2 2 2
18.0
25.4
1
11.4
2.9 8.2
0
0.0
2
16.3
3.2 4.5 4.7 1.7 6.1
4.8 1.6 1.8 2.4
Mean
2.4
16.9
2.0
7.1°
1.5*
7.0°
SEM
0.2
3.6
0.3
2.7
0.3
2.6
°P23 = 19.014; P < 0.0001). Post-hoc comparisons showed that the mean GH concentration for the first third of the saline night (8.2 tig/L) was significantly greater than those for the 3.0 ^g/kg night (2.0 Mg/L) and the 6.0 Mg/kg night (1.4 /*g/L, t = 4.278, P < 0.004 and t = 4.503, P < 0.003, respectively). There was a trend for GH in the first third of the 6.0 ixg/kg night to be less than that for the 3.0 fig/kg night (t = 1.997; P < 0.09). No significant differences among mean GH concentrations for the three conditions were found for the second and last thirds of the night. SCOP effects on sleep variables After correcting significance levels for the performance of multiple t tests (24), all significant effects of both doses of SCOP on all-night sleep architecture were confined to measures of rapid eye movement (REM) sleep. SCOP produced a significant dose-related delay in REM latency: the first REM period latencies were 90,197, and 355 min for the saline, 3.0 Mg/kg, and 6.0 ng/kg nights, respectively (17). The total percentage of 5-sleep (slow wave; stages 3 and 4) showed a slight nonsignificant increase after both doses of SCOP (17). In addition, we calculated the time of 5-sleep onset and the distribution of 8-sleep during each third of the night. Table 3 shows the distribution of minutes of 5-sleep for each third of the night for the saline, 3.0 /xg/kg, and 6.0 fig/kg nights;
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 02:48 For personal use only. No other uses without permission. . All rights reserved.
EFFECTS OF SCOP ON GH SECRETION
93
20 Saline Scopolamine (3 ug/kg) Scopolamine (6 ug/kg) 15 -
LLJ
Z
o FIG. 1. Mean nocturnal GH concentration measured every 15 min after the im administration of saline and 3 and 6 fig/ kg SCOP at 2300 h on 3 separate nights in eight normal male volunteers.
cr
10 -
o I I I-
o
cr o
5-
TIME TABLE 3. Effects of SCOP on GH concentration and 6-sleep for each third of the night 2300-0700 h in eight normal volunteers Saline 1st GH (Mg/mL) Mean SEM
5-Sleep Min SEM
2nd
3.0 Mg/kg SCOP 3rd
8.2
3.0°
1.7°
1.6
0.6
0.1
34.5 6.3
22.1
7.0 2.8
6.4
2nd
1st
6.0 /xg/kg SCOP 3rd
1st
2.0 0.3
4.4 1.9
2.5 0.5
1.4 0.1
35.5
23.4 6.1
5.2 2.8
28.5
8.9
6.7
2nd
3rd
3.16
3.0"
0.7
0.6
36.2 10.5
17.1 7.9
' P < 0.005 vs. first third. 1 P < 0.05 us. first third.
after correcting for multiple t tests, 5-sleep time was not affected by either SCOP dose. Relation between sleep variables and GH release
The mean time of first appearance of 5-sleep was not affected by SCOP (saline, 3.0 Mg/kg, and 6.0 Mg/kg SCOP nights were 43, 50, and 56 min, respectively). For all 3 nights of sampling for the 8 subjects, 8-sleep preceded the GH rise time for that night in 21 of 24 cases. In 1 of the 3 discrepant cases, the subject had no scoreable bsleep for that night (saline condition); in another, the GH rise time occurred immediately after sleep onset (also saline condition). Correlational analyses were performed
to examine the relationship between GH secretion and sleep parameters: neither sleep onset nor 5-sleep onset was significantly correlated with the GH rise time for any treatment condition. Minutes of 5-sleep for each third of the night and the mean GH concentration for the corresponding third of the night for each condition also were not significantly correlated. There was no significant correlation between the percent decrease in mean GH concentration and the percent delay in the first REM period latency for any of the 3 study nights.
Discussion This study was undertaken to determine the extent to which the degree of reduction of nocturnal GH secretion
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 02:48 For personal use only. No other uses without permission. . All rights reserved.
94
MCCRACKEN ET AL.
by SCOP could reflect quantitative differences in cholinergic neurotransmission. To our knowledge, it is the first study to compare the effects of saline and two different doses of a cholinergic antagonist on nocturnal GH release in the same subjects. These results extend the findings of earlier reports that demonstrated abolition of GH secretion during the night following treatment with the muscarinic antagonists methscopolamine (2), atropine (10), and pirenzepine (3). These data further underscore the importance of cholinergic regulation of GH release during both sleep and pharmacological stimulation (1). This stands in contrast to the absence of any effect of either dose of SCOP on nocturnal cortisol secretion in these subjects (17), although there was a trend for SCOP to advance the cortisol rise time. By employing smaller doses of cholinergic antagonists
than those used in earlier studies, we observed reliable graded reductions in sleep-related GH secretion in six of eight subjects (75%). This suggests that the magnitude of reduction in nocturnal GH secretion in response to a cholinergic antagonist is a quantitative reflection of change in cholinergic neurotransmission. However, the effect of SCOP on nocturnal GH secretion was complex; both suppression and delay of release of GH occurred. Mean 2300-0700 h GH concentrations were reduced by 29.3% after the 3.0 /ig/kg SCOP dose and by 40.0% after the 6.0 ng/kg SCOP dose, and both doses delayed the appearance of the GH peak by over 3 h. Additionally, the profile of GH secretion after SCOP was altered, as evidenced by a reduction in both pulse frequency and the amplitude of the first GH pulse. To our knowledge, this is the first report to suggest that cholinergic blockade can affect GH pulse characteristics. It is conceivable that the partial resumption of GH release after SCOP was due to the clearance of SCOP; without corresponding plasma SCOP levels we cannot dismiss this possibility. The mechanism for the reduction by SCOP of nocturnal GH observed in these data is probably multifactorial. Because pirenzepine, an Mi muscarinic subtype-selective antagonist, is also a potent blocker of nocturnal GH secretion (11), the most likely mechanism for the suppression of GH release by SCOP involves the Mi muscarinic receptor subtype, although the role of other muscarinic receptor subtypes cannot be ruled out. Other data (25-27) indicate that the pathway by which cholinergic blockade exerts its effect on GH secretion in rodents is likely to be via an increase in somatostatin release. In spite of earlier notions of a linkage between GH secretion and
Vol. 72, No. 1 Printed in U.S.A.
Dose-Dependent Effects of Scopolamine on Nocturnal Growth Hormone Secretion in Normal Adult Men: Relation to 5-Sleep Changes* JAMES T. McCRACKEN, RUSSELL E. POLAND, ROBERT T. RUBIN, AND LEONARDO TONDO Department of Psychiatry, University of California-Los Angeles School of Medicine, Harbor-University of California-Los Angeles Medical Center (J.T.M., R.E.P., R.T.R.), Torrance, California 90509; and the Department of Psychiatry, University of Cagliari (L. T.), Cagliari, Italy
a significantly greater delay observed after the larger dose. Similarly, a significant delay in the time of the GH rise was produced by SCOP. In contrast, the effects of both doses of SCOP on 5-sleep or sleep onset were small. These data confirm earlier reports demonstrating that cholinergic muscarinic input represents a potentially important source of regulation of nocturnal GH release and suggest that the magnitude of the reduction in GH and the extent of delay in the GH rise time may reflect quantitative differences in the degree of cholinergic blockade. These data are in agreement with recent studies suggesting that the timing of GH release need not be associated with 6sleep per se. (J Clin Endocrinol Metab 72: 90-95, 1990)
ABSTRACT. To explore the sensitivity of nocturnal GH secretion to different degrees of cholinergic blockade, we investigated the effects of two doses of the muscarinic receptor antagonist scopolamine (SCOP; 3.0 and 6.0 Mg/kg, im) and placebo, administered in a randomized fashion at 2300 h on three nights to eight normal male volunteers. Both doses of SCOP produced significant reductions in mean nocturnal GH concentration compared to the effects of the placebo; the higher dose of SCOP reduced GH to a greater degree than the lower dose, but this difference was not statistically significant (mean, 2.3 Mg/L after 6 Mg/kg vs. 3.0 jug/L after 3 Mg/kg). Both SCOP doses significantly shifted GH secretion into later portions of the night, with
T
ies have shown that nocturnal GH secretion can be abolished by the muscarinic antagonists methscopolamine (2), atropine (10), and pirenzepine (11), but whether the reduction is a consequence of suppression or delay of release, or both, is unknown. Since only one dose of a cholinergic antagonist was administered in these three studies, and GH secretion was completely abolished, it is unclear whether the degree of reduction of nocturnal GH secretion could reflect the degree of cholinergic blockade. Because of our interest in clinical studies of possible cholinergic abnormalities in severe depression, we elected to investigate whether nocturnal GH release would be sensitive to varying levels of muscarinic blockade by the muscarinic cholinergic antagonist scopolamine (SCOP). Our study also enabled us to examine the relationship between slow wave (5) sleep and GH secretion. This was of interest in view of reports suggesting that the apparent linkage of 5-sleep and GH release (12, 13) might, in fact, be fortuitous (14-16). We have previously described the effects of SCOP on allnight sleep architecture and nocturnal cortisol secretion in these subjects (17).
HE INFLUENCE of the cholinergic system on GH secretion has been appreciated for some time, beginning with early reports demonstrating daytime GH release after cholinergic agonist administration and reduction of sleep-related GH secretion after administration of medication with potent anticholinergic properties (1). Other studies have shown with more specificity that cholinergic muscarinic receptor antagonists, such as methylscopolamine, atropine, and pirenzepine, block the GH rise due to insulin-induced hypoglycemia (2, 3), clonidine, exercise, arginine (4), the enkephalin analog FK 33-824 (5), glucagon, L-dopa, apomorphine (6), and GH-releasing hormone (7-9). However, it has been suggested that the neurochemical regulation of nocturnal GH secretion may differ from that of GH release during wakefulness induced by pharmacological or physiological stimulation (1). Three stud-
Received December 5, 1989. Address requests for reprints to: Dr. James T. McCracken, Department of Psychiatry, Harbor-University of California-Los Angeles Medical Center, 1000 West Carson Street, Building F-5, Torrance, California 90509. * This work was supported by NIMH Grants MH-34471 (to R.E.P.) and MH-28380 (to R.T.R.), Physician Scientist Award MH-00722 (to J.T.M.), Research Scientist Development Award MH-00534 (to R.E.P.), Research Scientist Award MH-47363 (to R.T.R.), and NIH Clinical Research Center Grant RR-00425.
Subjects and Methods Eight normal male volunteers, aged 19-45 yr, gave their written informed consent to participate in the study. All sub90
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 02:48 For personal use only. No other uses without permission. . All rights reserved.
EFFECTS OF SCOP ON GH SECRETION jects were admitted three times, for 2 consecutive nights each time, to the Clinical Research Center at Harbor-University of California-Los Angeles Medical Center at approximately weekly intervals. All subjects were previously judged to be medically and psychologically healthy, as determined by a physical examination, electrocardiogram, full blood chemistry panel, thyroid function tests, urine drug screen, and clinical psychiatric interview. No subject had a past or present psychiatric illness or a history of major psychiatric disorder in any first degree relative. For each admission, the subjects spent 2 consecutive nights in the hospital, with the first night considered an adaptation night. On all nights, physiological sleep data (electroencephalogram, electrooculogram, and electromyogram) were recorded using Nihon Kohden polygraphs. Conventional electroencephalogram electrodes were attached at 1900 h, and recordings were made from 2300-0700 h. Lights out was from 2300-0700 h. On the second night of each 2-night pair, an iv catheter was inserted into the medial cubital vein at 1800 h and kept patent by periodic flushing with heparinized saline. A 10-ft extension allowed blood sampling to be performed from an adjacent room. Blood samples (5.0 mL) were taken every 15 min beginning at 2230 h and continuing until 0700 h. Also, on the second night of each admission at 2300 h just before lights out, the subjects were given an im injection of saline or SCOP (3.0 or 6.0 ng/ kg). The three doses were administered in a double blind, randomized fashion. Coded sleep records were blindly scored according to standard criteria (18). Sleep onset was defined as the first minute of stage 2 sleep, followed by at least 9 min of stage 2 sleep not interrupted by more than 1 min of waking or stage 1. 5-Sleep (stages 3 and 4) minutes were calculated for each third of the night for the 8-h recording period for each subject. 6-Sleep onset was considered the first epoch of slow wave sleep during the recording period. The blood samples were immediately placed on ice and centrifuged, and aliquots were made before storage at —20 C. Plasma GH was measured by a double antibody RIA (19), with materials obtained from Dr. Albert F. Parlow and the National Pituitary Agency. RP-1 was used as the reference preparation and for iodination. GH was iodinated by the glucose oxidaselactoperoxidase method (20), and the suitability of the iodinated hormone was determined by the talc-resin-trichloroacetic acid method (21). All samples from the 3 nights for each subject were analyzed in duplicate in the same assay. Analysis of low, medium, and high GH plasma pools in each assay showed the maximum intra- and interassay variabilities to be 9.3% and 11.9%, respectively. Statistical analysis of the sleep and hormone data was performed by repeated measures analysis of variance (ANOVA) and paired t tests. Because the hormone values were not Gaussian distributed across subjects, they were log transformed before analysis, which normalized the distributions. All reported significance levels are two tailed. The mean plasma GH concentration was calculated as the average of the plasma GH concentrations measured every 15 min from 2300 h until 0700 h. Means for the first, second, and last thirds of the night were calculated as the averages of the GH values every 15 min from 2300-0130,0145-0415, and 0430-0700 h, respectively. The time of the GH peak was determined as the time of the sample with
91
the maximum GH concentration for that night. The time of the GH rise was calculated according to the method of Jarrett et al. (22), The GH nadir was calculated as the mean of the three lowest sequential values during the night before a GH peak, and GH rise time was taken as that time after the nadir when the GH concentration first exceeded the nadir plus 2 SD of the nadir. Analyses of the effects of SCOP on pulse frequency and amplitude of GH secretion were also performed. Pulses of GH were identified as described previously (23). An increase in GH concentration was considered a significant pulse if both its increment and decline exceeded twice the intraassay coefficient of variation for that concentration range. Increases or decreases below 2.0 Mg/L were not considered significant. The absolute amplitude of a pulse was determined by calculating the difference between the maximum concentration of the pulse and the minimum concentration of the trough preceding that pulse. The average pulse amplitude was considered the mean pulse amplitude of all pulses for that night.
Results SCOP effects on mean plasma GH concentrations Table 1 shows the effects of the two doses of SCOP compared to that of saline on nocturnal GH for the eight subjects. Repeated measures ANOVA revealed a highly significant treatment effect on the log-transformed 23000700 h mean GH concentrations (F2(23 = 10.69; P < 0.002). Both the 3.0 and the 6.0 Atg/kg doses of SCOP produced a significant reduction in the mean GH concentration compared to saline (mean, 3.0 and 2.3 vs. 4.3 Mg/L, respectively; t = 3.99, P < 0.006 for saline us. 3 Mg/ kg; t = 4.963, P < 0.002 for saline vs. 6 ixg/kg). There was a slight trend for a greater reduction of mean GH after the 6 /ug/kg compared to the 3.0 /ig/kg dose, but the difference was not statistically significant (t = 1.147; P > 0.29). SCOP effects on GH pulse frequency and amplitude As seen in Table 2, the high dose of SCOP produced a significant reduction in the number of GH pulses during TABLE 1. Effects of SCOP on mean (±SEM) 2300-0700 h GH concentrations (micrograms per L) in eight normal volunteers Subject no.
Saline
SCOP
SCOP
1 2 3 4 5 6 7 8
4.5 3.8 3.4 1.8 5.3 7.0 2.2 6.7
1.9 2.3 2.4 1.7 3.1 7.3 1.4 4.0
4.1 2.3 1.8 1.6 2.1 3.2 1.1 2.5
Mean"
4.3
3.0*
2.3 C
°F 2 , 23 = 10.69; P < 0.002. 1 = 3.99; P < 0.006 us. saline. c t = 4.963; P < 0.002 vs. saline. 6
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 02:48 For personal use only. No other uses without permission. . All rights reserved.
MCCRACKEN ET AL.
92
TABLE 2. Effects of SCOP on mean (±SEM) GH pulse frequency and the amplitude (micrograms per L) of the first pulse between 2300-0700 h in eight normal volunteers Saline Subject Pulse no. no. 1 2 3 4 5 6 7 8
3 2 2 3 3 2 2 2
3.0 Mg/kg SCOP
6.0 Mg/kg SCOP
Pulse Pulse Pulse Pulse Pulse amplitude no. amplitude no. amplitude 15.6 16.3 14.0 1.5
23.3 25.2 6.2
33.0
1 3 2 2 2 1 2 3
1 2 2 2 2
18.0
25.4
1
11.4
2.9 8.2
0
0.0
2
16.3
3.2 4.5 4.7 1.7 6.1
4.8 1.6 1.8 2.4
Mean
2.4
16.9
2.0
7.1°
1.5*
7.0°
SEM
0.2
3.6
0.3
2.7
0.3
2.6
°P23 = 19.014; P < 0.0001). Post-hoc comparisons showed that the mean GH concentration for the first third of the saline night (8.2 tig/L) was significantly greater than those for the 3.0 ^g/kg night (2.0 Mg/L) and the 6.0 Mg/kg night (1.4 /*g/L, t = 4.278, P < 0.004 and t = 4.503, P < 0.003, respectively). There was a trend for GH in the first third of the 6.0 ixg/kg night to be less than that for the 3.0 fig/kg night (t = 1.997; P < 0.09). No significant differences among mean GH concentrations for the three conditions were found for the second and last thirds of the night. SCOP effects on sleep variables After correcting significance levels for the performance of multiple t tests (24), all significant effects of both doses of SCOP on all-night sleep architecture were confined to measures of rapid eye movement (REM) sleep. SCOP produced a significant dose-related delay in REM latency: the first REM period latencies were 90,197, and 355 min for the saline, 3.0 Mg/kg, and 6.0 ng/kg nights, respectively (17). The total percentage of 5-sleep (slow wave; stages 3 and 4) showed a slight nonsignificant increase after both doses of SCOP (17). In addition, we calculated the time of 5-sleep onset and the distribution of 8-sleep during each third of the night. Table 3 shows the distribution of minutes of 5-sleep for each third of the night for the saline, 3.0 /xg/kg, and 6.0 fig/kg nights;
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 02:48 For personal use only. No other uses without permission. . All rights reserved.
EFFECTS OF SCOP ON GH SECRETION
93
20 Saline Scopolamine (3 ug/kg) Scopolamine (6 ug/kg) 15 -
LLJ
Z
o FIG. 1. Mean nocturnal GH concentration measured every 15 min after the im administration of saline and 3 and 6 fig/ kg SCOP at 2300 h on 3 separate nights in eight normal male volunteers.
cr
10 -
o I I I-
o
cr o
5-
TIME TABLE 3. Effects of SCOP on GH concentration and 6-sleep for each third of the night 2300-0700 h in eight normal volunteers Saline 1st GH (Mg/mL) Mean SEM
5-Sleep Min SEM
2nd
3.0 Mg/kg SCOP 3rd
8.2
3.0°
1.7°
1.6
0.6
0.1
34.5 6.3
22.1
7.0 2.8
6.4
2nd
1st
6.0 /xg/kg SCOP 3rd
1st
2.0 0.3
4.4 1.9
2.5 0.5
1.4 0.1
35.5
23.4 6.1
5.2 2.8
28.5
8.9
6.7
2nd
3rd
3.16
3.0"
0.7
0.6
36.2 10.5
17.1 7.9
' P < 0.005 vs. first third. 1 P < 0.05 us. first third.
after correcting for multiple t tests, 5-sleep time was not affected by either SCOP dose. Relation between sleep variables and GH release
The mean time of first appearance of 5-sleep was not affected by SCOP (saline, 3.0 Mg/kg, and 6.0 Mg/kg SCOP nights were 43, 50, and 56 min, respectively). For all 3 nights of sampling for the 8 subjects, 8-sleep preceded the GH rise time for that night in 21 of 24 cases. In 1 of the 3 discrepant cases, the subject had no scoreable bsleep for that night (saline condition); in another, the GH rise time occurred immediately after sleep onset (also saline condition). Correlational analyses were performed
to examine the relationship between GH secretion and sleep parameters: neither sleep onset nor 5-sleep onset was significantly correlated with the GH rise time for any treatment condition. Minutes of 5-sleep for each third of the night and the mean GH concentration for the corresponding third of the night for each condition also were not significantly correlated. There was no significant correlation between the percent decrease in mean GH concentration and the percent delay in the first REM period latency for any of the 3 study nights.
Discussion This study was undertaken to determine the extent to which the degree of reduction of nocturnal GH secretion
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 02:48 For personal use only. No other uses without permission. . All rights reserved.
94
MCCRACKEN ET AL.
by SCOP could reflect quantitative differences in cholinergic neurotransmission. To our knowledge, it is the first study to compare the effects of saline and two different doses of a cholinergic antagonist on nocturnal GH release in the same subjects. These results extend the findings of earlier reports that demonstrated abolition of GH secretion during the night following treatment with the muscarinic antagonists methscopolamine (2), atropine (10), and pirenzepine (3). These data further underscore the importance of cholinergic regulation of GH release during both sleep and pharmacological stimulation (1). This stands in contrast to the absence of any effect of either dose of SCOP on nocturnal cortisol secretion in these subjects (17), although there was a trend for SCOP to advance the cortisol rise time. By employing smaller doses of cholinergic antagonists
than those used in earlier studies, we observed reliable graded reductions in sleep-related GH secretion in six of eight subjects (75%). This suggests that the magnitude of reduction in nocturnal GH secretion in response to a cholinergic antagonist is a quantitative reflection of change in cholinergic neurotransmission. However, the effect of SCOP on nocturnal GH secretion was complex; both suppression and delay of release of GH occurred. Mean 2300-0700 h GH concentrations were reduced by 29.3% after the 3.0 /ig/kg SCOP dose and by 40.0% after the 6.0 ng/kg SCOP dose, and both doses delayed the appearance of the GH peak by over 3 h. Additionally, the profile of GH secretion after SCOP was altered, as evidenced by a reduction in both pulse frequency and the amplitude of the first GH pulse. To our knowledge, this is the first report to suggest that cholinergic blockade can affect GH pulse characteristics. It is conceivable that the partial resumption of GH release after SCOP was due to the clearance of SCOP; without corresponding plasma SCOP levels we cannot dismiss this possibility. The mechanism for the reduction by SCOP of nocturnal GH observed in these data is probably multifactorial. Because pirenzepine, an Mi muscarinic subtype-selective antagonist, is also a potent blocker of nocturnal GH secretion (11), the most likely mechanism for the suppression of GH release by SCOP involves the Mi muscarinic receptor subtype, although the role of other muscarinic receptor subtypes cannot be ruled out. Other data (25-27) indicate that the pathway by which cholinergic blockade exerts its effect on GH secretion in rodents is likely to be via an increase in somatostatin release. In spite of earlier notions of a linkage between GH secretion and
