Psychiatry and Clinical Neurosciences (1 995), 49, 299-304

Regular Article
Effects of hCRH on sleep and body temperature rhythms
YUICHIRO TSUCHIYAMA, MD, NAOHISA UCHIMURA, MI), TETSURO SAKAMOTO, ML),
HISAO MAEDA AND TATAYU KOTORII, MI)
Department of Neuropsychiatry, Kurume University School ofMedicine, Kurume, lapun

Abstract W e investigated the effects of human corticotropin-releasing hormone (hCRH) on the sleep structure and
body temperature rhythms of seven healthy young men. Polysomnographic recordings were obtained and
body temperatures were monitored continuously for 48 h in each subject following the intravenous
administration of saline or of 100 pg hCRH. The administration of hCRH was associated with a significant
phase-advancement in body temperature rhythm v5 the saline control. The administration of hCRH affected
non-rapid eye movement (NREM) sleep by reducing significantly slow wave sleep (SWS) and sleep
efficiency, whereas the percentage of sleep stages 1 and 2 were increased significantly. These changes in body
temperature rhythms and sleep structure induced by hCRH resembled those observed in patients with
endogenous depression, except for the findings during rapid eye movement (REM) sleep. W e hypothesize
that a hypersecretion of hCRH in patients with endogenous depression may help to explain the changes in
body temperature rhythms and sleep structure often reported in such patients.

Key words body temperature rhythm, depression, human corticotropin-releasing hormone, polysomnography.

INTRODUCTION latency, an increase in REM sleep in early sleep and in

interrupted sleep, a decrease in slow wave sleep (SWS) and
The hypothalamo-pituitary adrenal axis consists of human
sleep efficiency, and a tendency toward early morning
corticotropin-releasing hormone (hCRH), a hypothalamic
a ~ a k e n i n g . ~A- 'decrease
~ in the amplitude and an advance-
hormone, the adrenocorticotrophic hormone (ACTH), and
ment, or delay, in the phase of body temperature rhythm
cortisol. Human corticotropin-releasing hormone acts on
have been reported in such p a t i e n t ~ . l ~ -Although
'~ these
the anterior pituitary to induce the secretion of ACTH,
changes are generally regarded as markers for depression,
which stimulates the secretion of corticosteroids by the
few studies have evaluated the effects of an elevated serum
adrenal. The corticotropin-releasing hormone (CRH) nerve
level of CRH on sleep."-18 Accordingly, we studied the
cell bodies are distributed widely in the hypothalamus and
effects of the intravenous (i.v.) administration of hCRH in
surrounding limbic structure. These cells send nerve fibers
seven healthy young men in order to evaluate its effects on
to the autonomic nerve center in the brainstem and are
the sleep structure and body temperature rhythm. Such
believed to control the function of the autonomic nervous
effects were compared with data published previously on
system.'J The intraventricular administration of CRH in
endogenous C R H levels in depressed patients.
rats accelerates the cardiovascular function, elevates the
blood glucose concentration, and accelerates sympathetic
~ , ~addition to stimulating the
nervous system a ~ t i v i t y . In SUBJECTS AND METHODS
release of ACTH, CRH suppresses the secretion of growth
hormone (GH) and of luteinizing hormone (LH).5-7 W e evaluated seven healthy male medical students (aged
The secretion of CRH is increased by stress. The associa- 20-27 years; mean, 24.6) who remained at the sleep labora-
tion between the hypersecretion of CRH and depression has tory of Kurume University throughout the 6 day study.
received attention ever since Nemeroff et a / . reported that Each subject provided his informed consent for participa-
the concentration of CRH is increased in the cerebrospinal tion. Daily meals were provided at 08.30, 12.30, and
fluid of patients suffering from depression.8 Characteristic 18.30 h. During the course of the study, the subjects were
disturbances in sleep structure and body temperature have prohibited from napping, exercising, drinking alcohol, and
been reported in patients with depression. Studies of de- taking medication. The subjects slept at Kurume University
pressed patients have also shown a reduction in REM for two nights prior to treatment so as to facilitate their
acclimatization to our sleep laboratory. At 21.30 on the third
day, 5 mL of saline was administered intravenously via an
Correspondence address: Yuichiro Tsuchiyama, Department of Neuropsychia- indwelling catheter placed in the median vein. Recordings
try, Kurume University School of Medicine, 67 Asahi-machi, Kurume, Fuku-
aka, Japan. of polysomnography (PSG) were started at 22.00. Subjects
Recevived 25 May 1995; revised 3 July 1995; accepted 18 August 1995. were awakened the next morning at 06.00 if they had not
300
awoken spontaneously. Recordings were performed in ac-
cordance with the standard guidelines of Rechtschaffen and
Kales.” Records were evaluated visually every 20 s accord-
ing to those criteria. At 21.30 on the fifth day, each subject
received a single 100 pg dose of hCRH i.v., and the PSG
was again recorded. Prior to administration, the hCRH was
dissolved in 5 mL of sterile saline. Blood samples were
collected through an indwelling catheter 15 min prior to
administration and again at 15 and 30 min after the admin-
istration of saline and/or hCRH. Samples were immediately
centrifuged at 4 ” C , then stored at - 8 0 ° C until assayed.
Plasma concentrations of ACTH and cortisol were deter-
mined by radioimmunoassay.
A thermistor (Memory-mac Model VTM2-005, Vein
Co. Ltd, Toyo Physical Co. Ltd, Tokyo, Japan) was placed in
the rectum for the 48 h period following the administration
of saline and/or hCRH so as to record the body temperature.
Readings of body temperature were obtained using Mac-
reader Model VMS3-232. A personal computer was used for
data processing. Successive measurements of body tempera-
ture were obtained every 1 min. Body temperature rhythms
were analyzed by the visual assessment of the raw data and
by analysis of data collected every 30 min using the least-
squares spectrum method. The ambient temperature in the
ACTH
40 i
Y. Tsuchiyama et al.
sleep laboratory was 25 “ C , and the humidity was 70%. Each
subject completed the self-rating scale developed by Beck in
order to evaluate the incidence of depressive symptoms
before and after the administration of hCRH.” Each subject
also provided his own assessment of the sleep pattern (sleep
latency, sleep periods, number of awakenings).
The Student’s t-test and XY correlation were used in
statistical analysis. A level of P < 0.05 was accepted as
statistically significant.
RESULTS
Changes in serum concentrations of ACTH and
cortisol following administration of saline and
hCRH
Figure 1 shows the changes in serum concentrations of
ACTH and cortisol following the administration of saline
and hCRH. No significant differences were found between
the concentrations of ACTH or cortisol at baseline prior to
the administration of saline or hCRH. Following hCRH
administration, the serum ACTH concentration increased
significantly after 15 min and again at 30 min relative to
levels following saline administration. Thirty min after
Cortisol
Effects of hCRH on sleep and body temperature rhythms 30 1

hCRH administration, the serum concentration of cortisol Table 2. Mean values of sleep parameters assessed in seven subjects on
increased significantly as compared with saline. nights following the administration of saline or hCRH

Saline ( m e a n t s.d.) hCKH ( m e a n t s.d.1

Body temperature following hCRH
No significant differences were found in the actual measure- Total sleep (min) 403.5 k 4 1 . 6 375.4 ? 53.7*
ments of the mean temperature (TPmean, "C), the mean Sleep efficiencv (oh) 84.1 t 8.7 78.2 ? 11.2**
maximal temperature (TPmax, "C), the mean minimal Stage 1 t 2 (min) 286.0 k35.5 290.0 k 36.7
temperature (TPmin, " C), the mean difference between Stage 1 + 2 (0%) 70.8 k 3.8 77.8 t 6.8**
TPmax and TPmin (TPdif, "C), and the times coinciding SWS (miti) 44.4 t 19.3 25.8t15.1**
with TPmax and TPmin during the 48 h recording period sws ( O h ) 11.4 t 5 . 8 6.9 f3.7*
following saline and hCRH injection (Table 1). However, REM (min) 71.7k22.6 57.5 +22.9*
analysis of temperature rhythms by the least squares spec- KEM ("/o) 17.5 t 4.9 14.8 k 5.0
trum method demonstrated a significant advancement in the Time awake after 26.7 t 29.6 42.6t34.1*
acrophase with the administration of hCRH vs saline. No sleep onset (min)
significant differences were found between the best fitted Sleep latency (min) 54.8 t 28.9 67.1 t 3 0 . Y
period, amplitude, or mesor. REM latency (min) 77.0 k 46.9 74.4 k 22.6
SWS latency (min) 23.0 i6.5 22.1 t 5 . 6
Changes in sleep structure following hCRH REM number (min) 4.1? 1.0 4.1 k 1.3
REM cycle (min) 108.8t21.6 116.5 t 46.9
In Table 2, mean values of sleep parameters are compared for
KEM duration (min) 21.5t4.7 15.2 t 4 . 6
hCRH w saline. Sleep efficiency and SWS each decreased
significantly following hCRH. The percentage of sleep * P < 0.1, **P < 0.05
stages 1 and 2 increased significantly following the adminis-
tration of hCRH. The total sleep time (TST) ( P < 0.071), Table 3. Comparison of sleep parameters between the first and
percentage of SWS ( P < 0.77) and REM sleep ( P < 0.074) second halves of the night following the administration of saline or
all tended to decrease, but not to a significant extent, on the hCKH
night following the administration of hCRH. The amount
of time awake after sleep onset tended to increase with Saline ( m e a n t s.d.) hCRH (mean t s.d.)
hCRH but not to a significant extent. No change in E M
First halJ'of.n(qhr
latency was observed following hCRH.
Stage 1 + 2 (rnin) 119.0 k 31.1 124.0f29.1
The effects of hCRH on sleep structure were investigated
Stage 1 t 2 (Yo) 66.8k 8.3 75.2 k 9.8*
in the first us the second halves of the night (Table 3). In the
SWS (min) 37.1 k 14.9 20.0 t 11.5**
first half, only SWS decreased significantly with hCRH
SWS (90) 22.1 t 10.3 12.4f7.3*
administration, while the %SWS tended to decrease. The
%sleep stages 1 and 2 showed a tendency to increase in the KEM (min) 21.1 * 12.2 21.5 k 9 . 8
first half of the hCRH night. In the second half, no REM (Oh) 11.1k6.0 12.4k5.5

Second halfof'night

Table 1. Mean values of parameters of body temperature rhythm Stage 1 + 2 (min) 1 6 7 . 0 f 12.4 167.0t15.7
during a 48 h period following the administration of saline or hCKH Stage 1 t 2 (Yo) 74.4 k 4.6 80.5 k 6.5
SWS (min) 7.0 k 8.7 6.3 t 7.2
Saline (mean k s.d.) hCKH (mean f s.d.) sws (010) 3.2 k 4.2 3.0 k 3.4
REM (min) 50.5 k 14.2 35.8 ? 15.9*
Visual analysis $actual measurements
E M (90) 22.4k6.1 16.6 t 6.2"
TPmean (' C) 37.0 k 0.1 37.0 k 0.1
TPmax ("C) 37.7 * 0.2 37.8 k 0.2 * P < 0.1, **P < 0.05.
TPmin ("C) 36.2k0.3 36.2 k 0.3
TPdif (' C) 1.67k0.47 1.7320.42
significant changes were observed in any parameter, al-
Tmax 19.51 22.26 19.16 k 2.44
though a tendency towards a decreased REM and Y O E M
Tmin 05.57 f 1.34 05.14 k 2.05
was observed following hCRH administration.
Anaiysis by the least squares spectrum method
Best fitted period (h) 23.3 t 1.3 23.9k 1.5 Correlation between changes in ACTH and cortisol
Amplitude ("C) 0.45 k0.13 0.53 t 0.16 concentrations and sleep structure following hCRH
Acrophase 1 8 . 3 0 t 1.02 17.59?0.55*
Since we observed large variations in serum concentrations
Mesor ( C) 37.0k0.2 37.0 k 0.1
of ACTH and cortisol in response to hCRH administration,
*P < 0.05. we investigated the correlations between these concentra-
302 Y. Tsuchiyama et al.

tions and the change in sleep structure for each individual. tween the body temperatures obtained at the time of SWS
Increases in concentrations of ACTH and cortisol were onset and the total duration of SWS following the adminis-
obtained by comparing the values obtained 15 min before vs tration of saline (Fig. 3a), but not of hCRH (Fig. 3b).
those obtained 30 min after the administration of hCRH. A
significant positive correlation was observed between the Subjective estimates of sleep and mood
rate of increase in ACTH concentration and that in the
Subjective estimates of sleep latency, TST, and the frequency
%sleep stage 1 (Fig. 2a). A similar correlation was also
of sleep interruptions did not differ significantly following
observed between the rate of increase in cortisol concentra-
the administration of saline or hCRH. No significant differ-
tion and in the Yosleep stage 2 sleep (Fig. 2b). No significant
ences in depression were observed between the scores of
correlation was observed between the YoSWS or YoSWS and
Beck self-rating scale obtained after the administration of
the rate of increase in serum concentrations of ACTH and
saline (2.422.2) or hCRH (1.1 t 1.3).
cortisol.

Body temperature and sleep DISCUSSION

There were no significant differences between the body
temperature obtained at sleep onset following either saline
or hCRH. (Sleep onset was defined as stage 2 sleep appear-
ing first on PSG.) The mean body temperature a t the time of
SWS onset was 37.1 kO.2"C for the saline night and
37.3 -+ 0.2"C for the hCRH, with no significant differ-
rences. A significant positive correlation was observed be-
Depression involves the endocrine system and circadian
rhythms as reflected by changes in sleep structure. Sleep
structures of depressed patients commonly include a reduc-
tion in REM latency, a proportionate increase in REM sleep
in the first half of the night, a decrease in SWS, and early
morning a ~ a k e n i n g . ' O - ' ~ . *The
~ - ~ ~first two characteristics
suggest that the phase of the circadian rhythm of REM sleep

(a)

r = 0.7577 P<0.05 0 v
c
0,
In r = 0.8654 Pc0.05

."
OC 01 I
-40 -20 0 20 40 60 80 100

Rate of increase in %stage 1 sleep

-0 -
,I
1

r = 0.7930 P<0.05
-
c

.- 300 Y =8.2148X+122.44 0,
rn 0. r=-0.3171
c
o 37.6 - Y = -0.0053X+37.4517
(I)
3
(I)

-
0
W
m

Figure 2. Correlation between rate of increase of (a) adrenocorti- Figure 3. Correlation between body temperatures obtained at the
cotropic hormone (ACTH) concentration and that of Yostage 1 sleep. time of slow wave sleep (SWS) onset and the total duration of SWS
Correlation between rate of increase of (b) cortisol concentation and following the administration of (a) saline and (b) human corticotropin-
that of %stage 2 sleep. releasing hormone (hCRH).
Effects of hCRH on sleep and body temperature rhythms
is advanced in depressed patients. In such patients, phase
advancement has also been observed in the circadian
rhythms of body temperature and in melatonin
secretion. 14*24 Moreover, the lability of the circadian rhythm
in depressed patients has been r e p ~ r t e d . ’As
~ hypersecretion
of cortisol and ACTH is commonly found in depressed
patients, it has been hypothesized that hyperfunction of the
hypothalamus-pituitary gland-adrenal cortex (HPA) secre-
tory axis is involved in d e p r e ~ s i o n . ~ ~ . ~ ~ - ~ ~
The effects of hCRH on human body temperature
rhythms apparently have not been investigated previously.
W e found that the acrophase of the body temperature
rhythm, identified after fitting the period analyzed by the
least squares spectrum method to 24 h, was advanced signi-
ficantly following the administration of hCRH vs saline.
This finding is noteworthy because a phase advancement in
body temperature rhythm has also been reported in patients
with
The effects of hCRH on the sleep of normal subjects have
been investigated.“.” Results in these studies did not
correlate. Although Britten et al. reported an increase in
‘awakening in rats’ following the intraventricular adminis-
tration of CRH,’’ Born et al. observed no change in human
sleep following the intravenous administration of hCRH
30pg every hour from 22.00 to 07.00.16 Holsboer et al.
reported a decrease in SWS and REM sleep and increases in
sleep stages 1 and 2 during the latter half of the night in
subjects administered 50 Fg of hCRH at 22.00 alone and in
subjects administered repeated doses of hCRH at 22.00,
23.00, 24.00, and Ol.00.”J8 In the present study, signifi-
cant decreases in sleep efficiency and SWS, and significant
increases in the %sleep stages 1 and 2, were observed on the
night following the intravenous administration of 100 pg of
hCRH. There was a trend toward an increase in the awake
time after sleep onset and a decrease in the TST and REM.
Some of the sleep changes induced by hCRH in this study
may have resulted from increases in ACTH and cortisol
secretion. W e correlated the rates of increase of ACTH and
cortisol secretion with changes in sleep parameters to test
this possibility. No significant correlations were observed
between the rate of decrease of SWS and the rate of increase
in ACTH or cortisol secretion, although the rates of increase
of sleep stages 1 and 2 were significantly correlated with the
rates of increase of ACTH and cortisol, respectively. These
results do not support the view that a decrease in SWS
following administration of hCRH resulted from a change
in the secretion of ACTH or cortisol. Born et al. described a
decrease in REM sleep during the first half of sleep follow-
ing the hourly administration of intravenous ACTH or
cortisol.” Those authors observed an increase in SWS
during the first half of sleep following the administration of
cortisol to healthy subjects. Their results also suggested that
the decrease in SWS observed in our study was not second-
ary to an increase in the serum concentration of ACTH and
cortisol.
Significant positive correlations were found between the
duration of SWS and the reduction of body temperature
303
during sleep or the temperature at sleep and SWS onset, as
well as the reduction of body temperature during a 4 h
period following sleep onset.2y-32 W e observed such a
correlation with the administration of saline, but not with
that of hCRH. As the body temperature at sleep or SWS
onset showed no difference between the administration of
saline and hCRH the reduction in SWS following hCRH
administration did not appear to depend on changes in body
temperature.
No subjects experienced depression on the morning fol-
lowing hCRH administration according to the results of the
Beck self-rating scale. Thus, the decrease in SWS and the
increase in sleep stages 1 and 2 that were induced by the
administration of hCRH were not accompanied by a depres-
sion. While the hypersecretion of hCRH is found in
depressed patients, it is not specific for depression as it occurs
also during stress and in patients with psychotic d i s e a ~ e . ~ , ~ ~
The observed hCRH-induced changes in body temperature
rhythms and sleep structure were similar to those observed
in patients with endogenous depression except for those
involving REM sleep. Our results lead us to hypothesize that
the hypersecretion of hCRH in patients with endogenous
depression is associated with a change in body temperature
rhythm and sleep structure.
ACKNOWLEDGMENTS
The author sincerely wishes to thank Drs Yoshishige Ida
and Hirotaka Mukasa for their assistance in the experiment.
REFERENCES
1. Olschowka JA, O’Donohue TL, Mueller C P et a/. Hypothalamic
and extrahypothalamic distribution of CRH-like immunoreactive
neurons in the rat brain. Neuroetiducrinulugy 1982; 35: 305-308.
2. Pelletier G, D’esy L, Cote J et a/. Immunoelectron microscopic
localization of CRF in the bovine hypothalamus. Neuroenducrinol-
ugy 1982; 35: 402-404.
3. Brown MR, Fisher LA, Spiess J rt a/. Corticotropin-releasing
factor: Actions on the sympathetic nervous system and metabo-
lism.]. Etidocritiol. 1982; 111: 928-931.
4. Brown MR, Fisher LA. Corticotropin-releasing factor: Effects on
the autonomic nervous system and visceral systems. Fed. Proc.
1985; 44: 243-248
Hashimoto K, Brodish A. The effect of hypothalamic lesion on
5.
stress-induced change of pituitary hormones (2). The effect on
plasma G H and TSH. Chi. Endocrind 1979; 27: 505-508.
6. Hashimoto K, Brodish A. The effect of hypothalamic lesion on
stress-induced change of pituitary hormones (3). Plasma prolactin
and LH after etherlaparotomy stress. C/iti. E n d o c r i d . 1979; 27:
625-628.
7. Hashimoto K, Murakami K, Ohno N ef a / . Physiological action of
CKF. Clin. Eitdocritiu!. 1985; 33: 583-589.
8. Nemeroff CB. The role of corticotropin-releasing factor in the
pathogenesis of major depression. Pharrnacopsychiatry 1988; 21:
76-82.
9. Hoofdakker RH, Bezrsma GM. On the explanation of short E M
latencies in depression. Psychiat.Res. 1985; 16: 155-163.
10. Kupfer DJ, Foster FG, Coble P e t a / . T h e application of EEG sleep
for the differential diagnosis of affective disorders. A m . ] . Psychiat.
1987; 135: 69-74.
304
11 Reynolds C F 111, Kupfer DJ. EEG sleep in elderly depressed,
demented, and healthy subjects. Bid. Psychiat. 1985; 20: 431-
442.
12. Reynolds CF 111, Kupfer DJ. Sleep research in affective ilhess.
State of the art circa 1987. Sleep 1987; 10: 199-215.
13. Wehr TA, Sack DA, Duncan W C ef a / . Sleep and circadian
rhythms in affective patients isolated from external time cues.
Psyrhiat. Re.?. 1985; 15: 327-339.
14. Mendlewicz J , Linkowski P. Hypothalamic Futictioris, S/eep arid
Circadian Rhythms in Affective Disorders. Raven Press, New York,
1987; 22 1-236.
15. Tsujimoto T, Yamada N , Shimoda K, Hanada K, Takahashi S.
Circadian rhythms in depression, Part 11. Circadian rhythms in
inpatients with various mental disorders.]. AJective Dis. 1990: 18:
199-210.
16. Born J , Spath-Schwalbe E, Schwakenhoffer H e r a / . Influences of
corticotropin-releasing hormone, adrenocorticotropin, and corti-
sol on sleep in normal man.]. C/ift, Eridocrird. Metab. 1989; 68:
904-91 1.
17. Holsboer F, Bardelehen U. Human corticotropin-releasing
hormone (h-CRH): Clinical studies in patients with affective
disorders, alcoholism, panic disorders and in normal controls.
Prog. Neuro-Psyrhopharmacol Bio/. Psychiat. 1988; 12: 165- 187.
18. Holshoer F, Bardeleben U , Steiger A. Effects of intravenous
corticotropin-releasing hormone upon sleep-related growth hor-
mone surge and sleep EEG in man. Neuroeridocrifdogy 1988; 48:
32-38.
19. Rechtschaffen A, Kales A, eds. A Mariiral ojstaridardized Termitlo/-
ogy, Technique arid Sroriri'r System for Sleep S t a p ifHirmari Subjectz.
Public Health Service, Washington, DC: US Government Print-
ing Office, 1968.
20. Beck AT. Depressiorr, Causes arid Tretlrmrrrt. University of Pennsyl-
vania Press, Philadelphia, 1972.
21. Goetz KR, Puig-Antich J , Ryan N et a / . Electro-encephalographic
sleep of adolescents with major depression and normal controls.
Arch. Gec Psychiar. 1987; 44: 61-68.
Y . Tsuchiyama e t a ! .
22. Hawkins DK, Taub JM, Castle RL. Extended sleep (hypersomnia)
in young depressed patients. A m . ] . Psychiat. 1985; 142: 905-910.
23. Kupfer DJ, Ulrich RF, Borbely AA. et al. Electroencephalographic
sleep of younger depressives. Arch. Gem Psychiat. 1985; 42:
806-810.
24. Moore-Ede MC, Sulzman FM, Fuller CA. The C h k s that Time
Us. Physiolopy d t h e Circadian Timirrg System. Harvard University,
Cambridge, 1982.
25. Souetre E, Salvati E, Rix H et a / . Effect of recovery on the cortisol
circadian rhythm of depressed patients. Bid. Psychiat. 1988; 24:
336-340.
26. Bardeleben V, Holsboer F. Cortisol response to corticotropin-
releasing hormone in dexamethasone pretreated patients with
depression.]. Neuroendorrimd. 1989; 1: 485-488.
27. Linkowski P, Mendlewicz J , Leclercq R. The 24-hour profile of
adrenocorticotropin and cortisol in major depressive illness. ].
Clitr. Errdocrinol. Metab. 1985; 61: 429-438.
28. Schlesser MA, Winokur G, Sherman BM. Hypothalamic-
pituitary-adrenal axis activity in depressive illness. Arch. Gen.
Psychiat. 1980; 37: 737-743.
29. I3ritten D, Koob GF, Rivier J , Vale W . Intraventricular
corticotropin-releasing factor enhances behavioral effects of
novelty.$!I Sci. 1982; 31: 363-367.
30. Czeisler CA, Weitzman ED, Moore-Ede M C eta/. Human sleep:
Its duration and organization depend on its circadian phase. Science
1980; 210: 1264-1267.
31. Czeisler CA, Zimmerman J C , Ronda J M et a[. Timing of REM
sleep is coupled to the circadian rhythm of body temperature in
man. SIeep 1980; 2: 329-346.
32. Berger RJ, Palca J W , Walker J M rt a/. Correlations between body
temperatures, metabolic rate and slow wave sleep in humans.
Nerrroscierrce Letters 1988; 86: 230-234.
33. Ehlers CL, Kupfer DJ. Hypothalamic peptide modulation of EEG
sleep in depression: A further application of the S-process hypoth-
esis. Bid. Psychiat. 1987; 22: 513-517.