Effects of sleep deprivation on human immune ... - The FASEB Journal

2 downloads 0 Views 1MB Size Report
tion during a normal 24 h sleep-wake cycle was delayed by sleep deprivation, ... tered patterns in immune functions occurred indepen- dently of the cortisol ...
Effects

of sleep deprivation

on human

immune

functions H. MOLDOFSKY,’

R. DAVIDSON,

of the Toronto Hospital,

and Immunology,

Surgery,

J.

F. A. LUE,

Toronto Western Division

University

AND

R. GORCZYNSKI

Toronto, Ontario,

of Toronto, Toronto, Ontario,

Canada

M5T

Canada M5G

2S8 and Departments of Psychiatry, 1X5

METHODS ABSTRACT

Ten

The effect of 40 h of wakefulness on a variety of immunological parameters in the peripheral blood from 10 normal male subjects was studied. Sleep deprivation led to enhanced nocturnal plasma interleukin 1-like and interleukin 2-like activities. The rise in nocturnal response of lymphocytes to pokeweed mitogen stimulation during a normal 24 h sleep-wake cycle was delayed by sleep deprivation, but the response to the phytohemagglutinin mitogen was unaffected. With resumed nocturnal sleep, there was a prolonged decline in natural killer cell activity (measured as spontaneous cytolytic activity for human tumor cells) and return of an increased response to pokeweed mitogen. The altered patterns in immune functions occurred independently of the cortisol circadian rhythm, which remained unchanged. -MOLDOFSKY, H.; LUE, F. A.; DAVIDSON,J.

vation

on

R.; GoRczYNslu, human immune

1972-1977;

1989.

Key Words:

sleep

immunity

R. Effects functions.

of sleep depriFASEB J. 3:

lymphokines

natural killer

cells

RECENT STUDIES HAVE SHOWN a close connection between sleep and the immune system. A variety of peptides that have been shown to promote sleep in animals also have immunomodulatory effects. These agents include: factor S, muramyl peptides, interleukin 1 (IL 1), a-interferon, tumor necrosis factor (1), vasoactive intestine peptide (2-4), prostaglandin D2 (5), and delta sleep-inducing peptide (6). We have shown that aspects of cellular and humoral immune functions are altered with sleep in humans. Peripheral blood lymphocyte (PBL) proliferative responses to pokeweed mitogen (PWM), plasma IL 1-like, and IL 2-like activities were all greater during nocturnal sleep than during wakefulness. Circulating natural killer (NK) cell activity was

greater

determine

during

the

whether

daytime

these

than

immune

during

to sleep, we investigated the possibility vation would alter these patterns. 1972

sleep

measures that

(7).

normal

male

subjects

aged

19-2 7 years

To

are related sleep

healthy,

(mean age, 23 years) who maintained a regular sleepwake schedule were investigated for 5 consecutive days (Fig. 1). Subjects were physically and psychologically healthy as determined by physical examination, Cornell Medical Index, and sleep-wake questionnaires. They were requested to maintain a regular sleep-wake cycle during the week preceding the study and to abstain from the use of alcohol, nicotine, and caffeine during the study. The study was conducted in accordance with the Declaration of Helsinki. All subjects were required to complete the University of Toronto Human Review approved consent form. On the initial adaptation night subjects were screened to exclude those with sleep pathologies, including sleep apnea and nocturnal myoclonus. Upon awakening (day 2), serial venous blood samples (3 ml/sample) were taken every 2 h from 0800 to 2400 h and half-hourly from 2400 to 0730 h until the termination of the study (day 5). Blood was drawn using a manifold system connected to a heparinized drip (8). Blood samples were immediately taken to the laboratory for assessment of immune function. During this baseline period (day 2), subjects were trained and allowed 10 scheduled practices of a 20-mm battery of computerized tasks to assess performance and mood (9). Subjects slept at their usual bedtimes while being monitored polysomnographically. Beginning on day 3 (0930 h), subjects remained awake for 40 h while engaging in the computerized performance tasks hourly, participating in a variety of table games, and reading. Monitoring of subjects included electroencephalogram, electrooculogram, submental electromyogram, and electrocardiogram via portable recorder (Oxford, Mediog 9000) with constant laboratory observation. On day 4 (ca. 2300 h), subjects resumed their usual sleep habits but were allowed to extend their sleep period times. Polysomnographic recordings were scored manually according to standard criteria (10).

depri-

‘To whom correspondence should be addressed: Department Psychiatry, Toronto Western Division of the Toronto Hospital, Bathurst St., Toronto, Ontario, Canada M5T 2S8. 0892-6638/89/0003-1972/$01

of 399

.50. © FASEB

Adaptation/xcresfliflg Day

Blood

sa.pling

2

Day

3

Day

4

begins

extended

sampling

sleep

ends

5 Blood

clock time

q 2 hr

(B sa.p.)

#{149}.-

q 1/2 hr

(16)

2400

0800

#{149}

0600

1. Protocol summary. Subjects were observed for 5 consecutive days with blood drawn every 2 h from 0800 to 2400 h and halfhourly from 2400 to 0730 h from morning of day 2 until day 5. Subjects from

slept their usual time on days 1 and 2, but remained awake arising on day 3 until normal time of sleep on day 4, then resumed sleep until desired awakening. Subjects remained awake, with hourly mood and performance testing and continuous monitoring.

KILLER

and

IL

2-like

activity,

as

described

previously

(7).

Briefly, PBL were obtained by ficoll/Hypaque separation of whole blood. Cells were resuspended in aminimal essential medium (a-MEM) supplemented with 1 x 10-6 M 2-mercaptoethanol, 10% heat-inactivated fetal calf serum, and 10% heat-inactivated commercial human AB serum before use. This serum mixture has proved to be optimal in our hands for mitogen proliferation and human NK assays. All assays were performed in triplicate on PBL within 60 mm of sampling. Plasma samples were stored at 4#{176}C throughout the sampling period and were assayed together, again in triplicate, some 24-48 h after termination of the test. For all assays shown, SEM was less than 10-15%. LYMPHOCYTE

PROLIFERATION

supernatant

These assays are described perimental groups containing

elsewhere

(11).

Briefly,

ex-

x iO ficoll/Hypaque separated PBL were incubated in triplicate for 72 h in 250 el a-MEM with and without addition of phytohemagglutinin (PHA) (Weilcome Labs, Research Triangle Park, N.C.), final concentration, 1% or PWM (Gibco Labs, Chagrin Falls, Ohio), final concentration, 1%. Thereafter, individual wells were pulsed with 1 tCi/well [3H]thymidine and cells were harvested for analysis in a well-type /3-scintillation counter 16 h later. Background counts per minute (cpm) varied from 500 to 1000 in stimulated cultures over the course of the experiment. Data were expressed as arithmetic mean cpm over background. Within a 24-h period, cpm for PHA and PWM stimulation have been found to vary for any one individual by 30-60%; from individual to individual, the z cpm varied from 15,000-40,000 (PHA) or 6000-15,000 (PWM) in different experiments. However, when repeated samples (within 5 mm)

SLEEP DEPRIVATION

AND

IMMUNE

1

FUNCTION

analyzed,

the assays

15%.

CELL

of

INTERLEUKIN

ACTIVITY

individual

1-LIKE

IL 1-like activity

wells.

From

the

curve

ACTIVITY

in the plasma

(7).

Briefly,

was assayed

as described

1 x 106 mouse thymocytes were plates for 72 h in the presence

incubated in microtiter of concanavalin A (ConA) (Sigma, St. Louis, Mo.), 5 eg/ml. Control cultures received thymocytes and ConA only or, in addition, a standard source of IL 1 (obtained as

the

supernatant

confluent line,

16-18

of

lipopolysaccharide-stimulated

cultures

of the mouse macrophage tumor Thereafter, samples were pulsed for 1 tCi 3H-tritiated thymidine (3HTdR)

P388D1).

h with

(New England harvested, and tillation counter.

Nuclear, Boston, Mass.), radioactivity was counted All groups were set up

Data

were

equals minus of IL

cpm (culture cpm (culture 1-like activity

first

cultures with preparation.

expressed

as

cpm,

6.7 Ci/mM, in a /3-scinin triplicate.

where

with ConA and putative with ConA only). Thereafter, were evaluated from the t

various

INTERLEUKIN

ASSAYS

and

immunologic

(effector:target vs. specific 5tCr release), the activity of each sample was expressed as lytic units/106 PBL (1 lytic unit defined as that required to give 20% lysis of I x iO target in 4 h).

previously

Blood samples were assayed for lymphocyte responses to mitogen stimulation, NK activity, IL 1-like activity,

of the

The NK activity assay has been described (12). In brief, 1 x iO 51Cr-K562 target cells were incubated in triplicate in 300 tl medium with individual PBL samples at ratios of 100:1, 60:1, and 30:1. After 4 h, specific 51Cr release was measured by sampling an aliquot (50 tl) of the

Figure

processed

in any

did not exceed

NATURAL

sleep

were

variation

described

wak#{149}tulness

Resu.ed

Blood

one individual

maximum Baseline

Day

Day

from

1

amounts

of a standard

cpm IL 1) units

cpm in IL I

2 ASSAY

IL 2 was assayed by using an IL 2-dependent human NK cell line (RMG-D4). Serial dilutions of plasma were added to 5 x 10 cells in 200 tl medium in flatbottomed microculture plates. At 48 h, cells were pulsed with 1 tCi [3H]thymidine for 16 h. Data were again expressed as cpm over cultures receiving no IL 2 (background range in these experiments: 300-700 cpm). Experiment-to-experiment variation with commercial IL 2 has given peak stimulation ranging from 4000 to 11,000 cpm. However, the dose of commercial material needed to give peak response remained constant (15 tl). PLASMA

CORTISOL

Cortisol levels were assayed by using a radioimmunoassay Products tectable

bilities

Corp., level was

were

in triplicate (RIA)

Los Angeles, Calif.). 0.2 tg/dl. Intraand

4.1 and

8.8%,

for all samples kit (Diagnostic Minimum interassay

devaria-

respectively.

1973

Technical difficulties Each subject provided and all assays were response assays were

TABLE

resulted in some loss of data. more than 76 blood samples, carried out in triplicate. Mitogen obtained

from

all 10 subjects;

Da y2

four

subjects were missing three or more points (i.e., 13, 5, 3, 3). NK assays were obtained for eight subjects, with four subjects missing three or more samples (i.e., 20, 21, 4, 3). IL 1 assay was completed for six subjects, with

three

subjects

5, 3).

IL

missing

2 assay

one subject

three

was

missing

or more

completed

samples

for

21 samples.

five

(i.e.,

subjects,

Samples

14, with

for the various

assays were lost for a variety of technical reasons (e.g., sporadic contamination of cultures, inactivation of IL

standard,

and failure

of proliferation

of cell lines).

values for the various assays were random. The of freedom (df) were appropriately adjusted statistical analyses.

1

Lost

degrees for all

To permit comparison of assays among subjects, the values obtained over the 76-h study for each parameter were standardized by using z statistical transformations. To determine the effects of day and time, data for each parameter were averaged for 12 time blocks (Mhours) (four 4-h blocks between 0700 and 2259 h: i.e., 0700-1059 h, 1100-1459 h, 1500-1859 h, 1900-2259 h, and

eight

1-h blocks

day. Two-way

between

2300

and

(3 days and 12 Mhours)

0659

h) for each

repeated-measures

analysis of variance (ANOVA) was carried parameter. This analysis was also repeated ing the time base for latencies to the various

out for each after adjustsleep stages

(1, 2, slow wave sleep [SWS] and rapid eye [REM]). To accommodate the more limited

1. Comparison of sleep physiology features on days 2 and 4 in means (and standard deviations)4

expressed

movement IL 1 and

Sleep Sleep

period latency,

time, mm

REM

latency,

Stage 1, mm Stage 2, mm Stages 3 and REM, mm

mm

462.7 25.0 109.0 28.2 252.6 60.0 92.6

mm

4, mm

Da y4

(22.7) (16.0) (65.6) (8.9) (33.5) (24.0) (24.2)

(95.2) (3.2) (23.8) (10.9) (78.5) (30.3) (40.6)

628.6k

6.3’ 76.7 22.8 341.3’ 103.1’ 140.9’

Subjects’ baseline sleep (day 2) values are typical for age group. There were no significant differences between adaptation/screening (day 1) and baseline (day 2). Resumed sleep (day 4) after a night of sleep loss resulted in significantly longer sleep period time, stages 2, 3, 4, REM, and shortened sleep latency. bDiffers from baseline value, P < 0.0001, paired I test, d.f = 9. ‘Differs from baseline value, P < 0.01, paired test, d.f = 9.

that of baseline day 2 was less than on days 3 and 4, and sleep deprivation day 3 was less than resumed sleep day 4 (SNK, P < 0.05). Response to PWM stimulation was generally higher during nighttime than daytime (F = 47.3, d.f = 11,99, P < 0.0001). During sleep

deprivation, the nocturnal rise in activity did not occur until 0400-0559 h (F 6.4, d.f 22,188, P < 0.0001) as compared with increased activity from 2400 to 0559 h =

during during

baseline resumed

=

sleep on day sleep on

2 and from 2400 to 0559 h day 4 (SNK, P < 0.05)

(Fig. 2). The mitogen response was greater during all sleep stages than during daytime wakefulness (F 9.8, d.f 6,47, P < 0.0001), with maximum response occurring during SWS and minimum response during stage 1, REM, and wakefulness after sleep onset. =

=

IL 2 data sets, the repeated-measures ANOVA was performed on only nine nighttime blocks (i.e., 1900-0659 h). To further examine the effects of sleep stages, another two-way (2 days and 7 stages) repeated-measures ANOVA was carried out using average values from each stage (daytime wake, nighttime wake, 1, 2, 3, 4,

and REM). This latter analysis grouped 0700 and 2300 h as daytime wake. When significant, comparison

the Student-Newman-Keuls was employed.

data between F values were (SNK)

multiple

higher

stage (F

physiology

Table

1 shows

2.1,

d.f

and

day

of the sleep

was reduced

sleep).

of day 2

with there was a prolongation of sleep period time, stage 2, SWS, and REM. There were no significant differences in sleep physiology parameters between the adaptation and baseline nights.

day 2, sleep latency

4 (resumed

physiology Compared

on day 4 and

blood

lymphocyte

mitogen

stimulation

Pokeweed

mitogen

Vol. 3

June 1989

=

with 4.3,

mean

d.f

=

=

cell

h) after

onset

on days

to

response increas2,18, P < 0.05);

of

2 and

4

activity

progressively

decreased

over the 3 days of

observed; however at nighttime on days 2 and 4 there was greater NK cell activity during the 4 h before onset of sleep than during the remaining time asleep (F 4.2, df 8,56, P < 0.0001). There was a persis=

tent reduction in NK cell activity from 2300 to on day 4 during resumed sleep, compared with mum rise in NK cell activity from 2400 to 0059 sleep onset on baseline day 2 (F = 1.7, df = P < 0.05), followed by the nocturnal decline in

response

A day effect was observed, ing over the 3 days (F

1974

responsiveness

0600-0800

at any other time 11,99, P < 0.05).

the study (F = 7.4, df = 2,14, P < 0.01), i.e., mean NK activity on baseline day 2 was higher than on days 3 and 4, and higher on day 3 than on resumed sleep day 4 (SNK, P < 0.05). No particular daytime pattern was

=

Peripheral

no differences over the 3 days However, PHA response was

than

killer

NK activity aspects

showed stages.

6 h (approximately

1 sleep

Natural

Sleep

mitogen response

PHA stimulation or during sleep

=

RESULTS

(baseline)

Phytoliemagglutinin

(Fig.

3). However,

sleep

stages.

The FASEB Journal

there

were no differences

0559

22,136, activity

among

MOLDOFSKY

h

a maxih with

the

ET AL.

Dsy

2,

Bas.lins

s1s.p

Interleukin

1-like activity

1.0

Despite potential difficulties in measuring plasma IL 1 activity because of suspected inhibitors associated with the biologic assay used in this study (13), peaks of plasma IL 1-like activity occurred in five of the six sUbjects at various times during baseline sleep. In contrast to the previous study (7) where IL I activity was observed with SWS after sleep onset, these grouped data did not show any specific diurnal or sleep stage-related pattern. Mean nighttime plasma IL 1-like activity was higher during sleep deprivation and resumed sleep compared with baseline (F 4.6, df 2,10, P < 0.05).

z s C 0

r e -0.

=

=

-1.

-1.

Interleukin

2-like

activity

-1.

Once

-1.

more,

1 h during -1.

3, !xtended

IL

part

2-like

activity

of sleep,

was

as in the

greater initial

for study

(7). Mean nighttime plasma IL 2-like activity was highest from 0100 to 0159 h on all three nights (F 2.4, df 8,32, P < 0.05). Furthermore, the plasma IL 2like activity was higher during 0100-0159 h of day 3 sleep deprivation in comparison with that during 2400-0059 h of both day 2 baseline sleep and day 4 resumed sleep (F 2.2, c/f 16,58, P < 0.01).

-2.

Day

plasma the early

=

wakefulness

=

z

=

=

$

Plasma

C 0

cortisol

The pattern of cortisol production showed a typical daytime peak (0700-1100 h) and nocturnal trough (2300-0300 h). This circadian rhythmicity for all subjects was not altered by either sleep deprivation or by

r

e

the Day

4, Rssuaed

recovery

sleep

for

any

individual.

sleep

1.#{149}

DISCUSSION

0.

z

Despite

0..

with and

$ C 0

r e -I.’ -1.

7-li.

11-15

15-19

19-23

1357

time (Mhour) Figure 2. Summary of PBL response

to PWM. Change in cpm for each sample was z transformed within subjects by setting mean to 0 and standard deviation (SD) to 1. Transformed values were then grouped into 12 time blocks (Mhours) and two-way (3 days and 12 Mhours) repeated-measures ANOVA was performed. Mean cpm over background for all 10 subjects was 10737 for day 2, 11911 for day 3, and 13210 for day 4. One SD was 5553, 4723, and 3624 for days 2, 3, and 4, respectively. Response was higher on days 3 and 4 than on day 2 (F = 4.3, d.f = 2,18, P < 0.05), higher during nighttime than daytime (F = 47.3, d.f - 11,99, P < 0.0001); nocturnal increase was delayed until 0400 h during the sleep deprivation night (F = 6.4, d.f = 22,188, P < 0.0001). Bars indicate maximum and minimum SEM. PWM response vs. time (n 10). SLEEP DEPRIVATION

AND

IMMUNE

FUNCTION

great

regard illness,

scientific

advances

in the

past

30 years

to sleep physiology and its role in health the function of sleep remains unresolved.

The theory that sleep is a restorative process for tissue anabolic activities to counter catabolic functions has been controversial, with evidence presented both in favor of (14) and in opposition to the theory (15). This study largely confirms our initial observations that aspects of the cellular and humoral immune functions are related to sleep (7). If sleep serves as a restorative process, then the observed increased proliferative response of peripheral blood lymphocytes during baseline sleep may be an indicator of this process. This study shows a potential for increased responses to PWM, a B cell stimulator, during sleep. However, it also confirms our previous findings that responses to PHA, a T cell stimulator in vitro, do not show increased activity with onset of sleep (7). In fact, no

special

diurnal

pattern

was

observed

except

that

PHA hours

mitogen response increased inexplicably for a few in the early part of the morning, approximately 6 h after the onset of sleep. As in the first report, NK cell activity shows a particular pattern of increased activity before sleep and a precipitous drop after onset of sleep (7). This decline in peripheral blood NK cell 1975

Day 2, Baseline

sleep

sleep might relate to reduced cellular which would not be consistent with the theory. On the other hand, in support of the

1.I

function

1.6

production, restorative

with

1.4

theory,

i.2

NK cells into tissues

during

sleep

of pathogenic 1.0

z 0.4

fl

S 0.2

r e

o.c

TI

-

-0.2 -0.4

Day 3, Sleep

deprivation

1.0

0.$

40 h

PWM

W

Whereas

during

a reduction

-0. 2

and lysis

the

24-h

mean

PWM

sleep

deprivation

in the number

re-

or whether

the

is accompanied

of B cells or their

resumption

of sleep

by

functional

would

sociated with restoration of B cells and/or tional capabilities. Similarly, the significance

be as-

their funcof the per-

sistent decline in NK cell activity with resumption of sleep after extended wakefulness remains to be established.

On

the

other

hand,

the

lack

of a specific

sleep

4

deprivation suggests

-0.0

4,

Resuced

Sleep

effect that

T cell

on stimulation activity

with

or certain

PHA subsets

mitogen of T cell

function may not be as readily affected by 40 h of wakefulness. If, as suggested by animal studies (16-19), IL 1 has sleep-inducing properties, then sleep deprivation might be expected to alter IL 1 levels. OUr previous analysis

z S

r

of the

surveillance

cells.

of wakefulness.

activity

C 0

be redistribution

sponse showed an overall increase in activity from baseline day 2 to day 4, 24-h mean NK cell activity was decreased over the same time interval. Furthermore, a specific nocturnal change occurred during sleep deprivation. The expected rise in response to stimulation by PWM commencing with sleep on the baseline night was delayed until 0400-0559 h on day 3. With resumed sleep, the nocturnal rise in response to stimulation with PWM was reestablished. It is as yet unknown whether the kinetic delay in response to stimulation with

z

-0.

may

for immune

Sleep deprivation is a technique that has been commonly used to assess the function of sleep (15). This study shows that aspects of the immune cellular functions, i.e., PWM and NK cell activity, are altered with

0.0

C 0

there

0.

(7) led us to anticipate

an increase

in IL 2 activity

after

increased IL 1 activity. The observed increase in plasma IL 1-like and IL 2-like activities during the night of sleep deprivation might reflect changes in these interleukins within the central nervous system. IL 1 has been shown to be a product of astroglial cells (20). The normal brain produces IL 1/3 mRNA, and receptors

0.

e

are widely

distributed

throughout

the brain

(21).

A net-

work of IL 1 immunoreactive fibers has been observed in the human hypothalamus (22). Furthermore, IL 1-like biologic activity and IL 1/3 RIA measurements are augmented 7-11

11-15

15-19

19-23

1357

time (Mhour) Figure 3. Summary

of NK

activity.

Percent

lysis

for each

in

the

cerebrospinal

fluid

of

cats

while

asleep as compared with wakefulness (23). A frequent confounding influence to the interpretation of the physiological effects of sleep deprivation sample

the possibility

that

the emotional

stress

is

of the disruptive

experience may contribute to the observed changes. was z transformed within subjects by setting mean to 0 and stanThis idea was raised by Home (24) in his review of the dard deviation (SD) to 1. Transformed values were then grouped studies of assessments of aspects of the iminto 12 time blocks (Mhours) and two-way (3 days and 12 Mho#{252}rs) two earlier repeated-measures ANOVA was carried Out. Mean percent lysisfor all eight subjects was 27.9% for day 2, 26.3% for day 3, and 20.9% for day 4. One 5D was 11.2%, 12.9%, and 11.6% for days 2, 3, and 4, respectively. NK activity was higher on day 2 than on days 3 and 4 (F = 7.4, d.f 2, P < 0.01), and was lower during resumed sleep on day 4 from 2400 to 0059 h than during baseline sleep on day 2 (F 1.7, a’.f = 22,136, P < 0.05). Bars indicate maximum and minimum SEM. NK cell activity vs. time (n = 8).

1976

Vol. 3

June 1989

mune

system

study involved phagocytosis

after

human

sleep

deprivation.

The

first

72 h of sleep deprivation. The reduced by polymorphonuclear granulocytes and

increased interferon production by lymphocytes were accompanied by raised adrenocortical and adrenomedullary activities (25). The second study involving 64 h of wakefulness with subjects in constant light and

The FASEBJournal

MOLDOFSKY ET AL.

in visual and world showed

isolation from the surrounding response to PHA but no changes in endocrine function (26). These studies are also flawed by the reliance on single morning blood samples taken before and after sleep deprivation and a

reduced

several days later. The diurnal fluctuations observed in our previous (7) and current study suggest that earlier studies using single sample techniques are likely to be difficult to interpret. Although the experimental procedures

may

be stressful

(e.g.,

repeated

blood

sampling),

of stress,

subjects.

showed

a normal

Furthermore,

circadian

no day-to-day

pattern

change

2

during

nocturnal

sleep deprivation

in immune

functions

that have been

ety of stressful circumstances (27, 28) may expected disordered sleep that accompanies tionally distressing situations or the diurnal

of the immune

13.

secretion at the changes

reported

12.

occurred

were unrelated

to any changes in the pattern of cortisol night. On the other hand, it is possible that

10.

for all

in this circadian pattern throughout the study. Finally, the changes in PWM, PHA, NK cell activity, IL 1, and IL

9.

11.

there was no clinical or physiological evidence that the subjects who participated in this study were distressed. The patterns of plasma cortisol, a physiological indicator

chronobiological

auditive

in a varirelate to the these emofluctuation

14. 15. 16.

functions. 17.

We

thank Dr. A. W. MacLean

Angus

and R. Heslegrave

testing

protocol,

research

for statistical

for providing assistants

advice,

computerized

M. Gregoris,

Drs. R. G.

performance

research

was

Environmental cal Research

supported

by

Medicine Council

the

contract of Canada

Defence

and

Civil

no. 02E.97711-5-8483 grant

no.

MA

for asThis

Instmtute and

19. Shoham,

Krueger,

of

Medi-

7733.

REFERENCES

20.

21.

1. Krueger, J. M., Toth, L. A., Cady, A. B., Johannsen, L., and Obal, F. (1988) Immunomodulation and sleep. Sleep Peptides: Basic and Clinical Approaches (Inoue, S., Schneider-Helmert, D., eds) pp. 95-129, Japan Sci. Soc. Press/Springer Verlag, Berlin 2. Drucker-Colin, R., Bernal-Pedraza, J., Fernandez-Cancino, F., and Oksenberg, A. (1984) Is vasoactive intestinal polypeptide (VIP) a sleep factor? Peptides 5, 837-840 3. Obal, E, Sary, G., Alfoldi, P., Rubicsek, G., and Obal, E (1986) Vasoactive intestinal polypeptide promotes sleep without effects on brain temperature in rats at night. Neurosci. Leti. 64, 236-240 4. Riou, F., Cespuglio, R., and Jouvet, M. (1982) Endogenous peptides and sleep in the rat. III. The hypnogenic properties of vasoactive intestinal polypeptide (VIP). Ne’uropeptides 2, 265-277 5. Ueno, R., Honda, K., Inoue, S., and Hayaishi, 0. (1983) Prostaglandin D2 a cerebral sleep-inducing substance in rats. Proc. Nail. Acad. Sci. USA 80, 1735-1737 6. Yehuda, S., Shredny, B., and Kalechman, Y. (1987) Effects of DSIP, 5-HTP and serotonin on the lymphokine system in a preliminary study. mt. j Neurosci. 33, 185-197 7. Moldofsky, H., Lue, F. A., Eisen, J., Keystone, E., and Gorczynski, R. M. (1986) The relationship of interleukin-1 and immune functions to sleep in humans. Psychosom. Med. 48, 309-318 8. Weitzman, E. D., Czeisler, C. A., Zimmerman, J. C., Ronda, J. M., and Knauer, R. S. (1982) Blood sample collection in

SLEEP DEPRIVATION

AND

IMMUNE

FUNCTION

analytical

18. Susic, V., and Totic, S. (1987) Short interleukin-l on sleep and temperature

K. Carayanniotis,

and L. Poplonski and research fellow Dr. J. Jephthah-Ochola sistance with the assays, and members of the sleep laboratory.

disorders:

22.

23.

and

therapeutic

tech-

nmques. Sleeping and Waking Dtsorders: Indications and Techniques (Guillemin#{227}ult, C., ed) pp. 311-313, Addison-Wesley, Menlo Park, California Angus, R. 0., and Heslegrave, R. J. (1985) The effects of sleep loss and sustained cognitive performance. Behav. Res. Methods Instrum. & Comput. 17, 55-67 Rechtschaffen, A., and Kales, A. (1968) A Manual of Standardized Terminology, Techniques and Scoring System for Sleep Stages of Human Subjects, UCLA Brain Information Service/Brain Research Institute, Los Angeles Gorczynski, R. M., and MacRae, S. (1981) Inhibition of cell proliferation rather than of cell lysis as a measure of immune reactivity in embryo-antigen-challenged mice. Br j Cancer 43, 19-31 Harris, J. J., Chin, J., Jewett, M. A. S., Kennedy, M., and Gorczynski, R. M. (1984) Monoclonal antibodies against SSEA-l antigen: binding properties and inhibition of human natural killer cell activity against target cells bearing SSEA-1 antigen. J. Immunol. 132, 2505-2509 Cannon, J. G., and Dinarello, C. A. (1985) Increased plasma interleukin-1 activity in women after ovulation. Science 227, 1247-1249 Adam, K., and Oswald, I. (1983) Protein synthesis, bodily renewal and sleep-wake cycle. Clin. Sci. 65, 561-567 Home, J. (1983) Human sleep and tissue restitution: some qualifications and doubts. Clin. Sci. 65, 569-579 Krueger, J. M., Walter, J., Dinarello, C. A., Wolff, S. M., and Chedid, L. (1984) Sleep-promoting effects of endogenous pyrogen (interleukin-l). Am. j Physiol. 246, R994-R999 Tobler, I., Borbely, A. A., Schwyzer, M., and Fontana, A. (1984) Interleukin-l derived from astrocytes enhances slow activity in sleep EEG of the rat. Eur. j Pharmacol. 104, 191-192

S., Davenne, M. (1987) 1 enhance

J.

D., Cady, Recombinant slow-wave

and long term effects of in cat. Sleep Res. 16, 150

A. B., Dinarello, C. A., and tumor necrosis factor and sleep. Am. J. Physiol. 253,

interleukin R142-R149 Fontana, A., Kristensen, F., Dubs, R., Gemsa, A., Weber, E. (1982) Production of prostaglandin E and an interleukin-l-like factor by cultured astrocytes and C6 glioma cells. j Immunol. 129, 2416-2419 Farrar, W. L., Hill, j. M., Harel-Bellan, A., and Vinocour, M. (1987) The immune logic brain. ImmunoL Rev. 100, 361-378 Breder, C., Dinarello, C. A., and Saper, C. B. (1988) Interleukin-l immunoreactive innervation of the human hypothalamus. Science 240, 321-324 Lue, E A., Bail, M., Jephthah-Ochola, J., Carayanniotis, K.,

Gorczynski,

R., and Moldofsky,

H. (1988) Sleep and cerebro-

spinal fluid interleukin-l-like activity in the cat. mt. j Neurosci. 42, 179-183 24. Home, J. (1988) Why We Sleep: The Functions of Sleep in Humans and Other Mammals, Oxford University Press, Oxford 25. Palmblad, J., Cantell, K., Strander, H., Froberg, J., Karlsson, C. C., Levi, L., Granstrom, M., and Unger, P. (1976) Stressor exposure and immunological response in man: interferonproducing capacity and phagocytosis. J. Psychosom. Res. 20, 193-199 26. Palmblad, J., Petrini, B., Wasserman, J., and Akerstedt, T. (1979) Lymphocyte and granulocyte reactions during sleep deprivation. Psychosom. Med. 41, 273-278 27. Baker, G. H. B. (1987) Psychological factors and immunity. ,j Psychosorn. Res. 31, 1-10

28. Tecoma, immune

E. S., and Huey, L. H. (1985) Psychic distress response. L!fe Sci. 36, 1799-1812

and the

Receivedfor publication January 20, 1989. Accepted for publication March 21, 1989.

1977