Professional Documents
Culture Documents
sleepdeprivationpaper
sleepdeprivationpaper
net/publication/236602262
CITATIONS READS
11 288
10 authors, including:
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Ilan Wald on 20 May 2014.
cantly elevated state anxiety and psychological dissociation health outcome is important.
132.66.156.41 - 9/10/2013 10:29:08 AM
E-Mail karger@karger.com
Tel Aviv 69978 (Israel)
www.karger.com/nim
E-Mail shamgar @ post.tau.ac.il
Sleep is a critical biological process for normal func- potentially through common proinflammatory responses.
tioning of all known mammals [2, 3] and has an impor- Indeed, some psychological and health disorders, includ-
tant role in learning and memory consolidation [4, 5], ing depression and cardiovascular diseases, seem to in-
regulation of hormonal levels [6] and immune activities volve irregularities in two or three of these processes [32,
[7]. Sleep and its associated circadian rhythms are hy- 33], suggesting the biological and clinical significance of
pothesized to fulfill a restorative role, preventing accu- these interactions. Importantly, under naturalistic cir-
mulated perceptual overload of the cerebral cortex [8], cumstances, sleep deprivation is often accompanied by
and ensuring stable and healthy cytokine rhythms [9]. stress or continuous activity. For example, students study-
Sleep and immunity have been suggested to mutually ing for examinations, professionals in demanding jobs,
affect each other. Specifically, sleep deprivation has been parents to newborns or soldiers during active service
associated with a tendency towards a proinflammatory may all experience sleep deprivation alongside maintain-
state, as reflected by a higher potential of monocytes to ing activity and/or being subjected to stressful circum-
produce interleukin (IL)-6 and tumor necrosis factor-α stances. Thus, in order to simulate such circumstances
upon in vitro activation [10, 11]. Sleep deprivation was and measure their impact on the immune system, we em-
also reported to modulate natural killer (NK) cell cyto- ployed a research protocol combining sleep deprivation
toxicity, with some studies showing reduction and some with potentially stressful activities.
enhancement in their activity [9, 12, 13]. Additionally, We compared young male and female students sub-
human systemic levels of prolactin, estradiol and cortisol jected to a 30-hour sleep deprivation alongside psycho-
were found to correlate with total sleep latency, REM logical stress, with a control group of participants. Spe-
sleep latency and the percentage of slow-wave sleep fol- cifically, we assessed multiple immune, endocrine, cog-
lowing either total sleep deprivation or selective REM de- nitive and psychological indices within participants
privation [14]. Inversely, the impact of immunity on before and after the sleep deprivation protocol. The bio-
sleep has long been examined, and many cytokines were logical indices studied were those previously reported to
found to regulate sleep patterns. For example, proinflam- be affected by either stress or sleep deprivation [14, 34–
matory cytokines, including IL-1, are known to be elevat- 36], as we herein assess the combined effect of these con-
ed in response to infections [15], and both intracerebro- ditions on immunity. The psychological and cognitive
ventricular and peripheral injections of proinflammatory outcomes of this study will be described in detail else-
cytokines were shown to impact sleeping patterns [16, where, while this paper focuses on the immunological
17]. Similarly, humans subjected to a viral infection re- and endocrine outcomes of the sleep deprivation and
ported a growing desire to sleep [18]. Indeed, some stud- stress paradigm.
ies suggest that proinflammatory cytokines increase
slow-wave sleep and sleep duration, while anti-inflam-
matory cytokines inhibit sleep [19–21]. However, other Materials and Methods
studies suggest a more complex association between Participants
sleep and immunity [22–24]. Forty-one young healthy undergraduate psychology students
Immunity was repeatedly shown to be affected by psy- (mean age = 24 years; SD = 1.86; 25 females) chose to participate
chological and physiological stressors through various in the study after being briefly informed about the paradigm. Par-
mechanisms. Some investigators suggested a dichotomy ticipants were randomly assigned into a stress group (23 partici-
pants, mean age 24 years; SD 1.86; 14 females) and a control group
between the impact of acute and chronic stress, namely (18 participants, mean age 23.67; SD 1.46; 11 females). Study par-
that chronic stress suppresses immunity and reduces re- ticipation fulfilled students’ requirements for participating in ex-
sistance to diseases, while acute stress enhances immu- periments. The study was conducted in 2 replicates (4 weeks
nity and disease resistance [25, 26]. However, several apart), each including about half of the participants in each group.
acute stressors were actually shown to suppress various The study was approved by the Institutional Review Board, and all
participants provided a written informed consent.
immune indices, including peripheral levels of NK acti-
vity and Th1 cytokine levels [27–30]. Acute stress was also Experimental Design and Specific Timelines
shown to increase proinflammatory cytokines, including The experimental design is illustrated in figure 1. All parti-
IL-6 [31]. cipants first reported to the laboratory on Monday morning
(09:00 AM), 3 days before the planned sleep deprivation night, to
Overall, it appears that immunity, stress and sleep are conduct a battery of psychological tests and to provide baseline
mutually interacting processes, and that sleep deprivation saliva and blood samples. Participants were then randomly as-
Tel-Aviv University Lib. of Life Science and Med.
and stress can often lead to similar immune consequences, signed to either the stress or the control group. Three days later,
132.66.156.41 - 9/10/2013 10:29:08 AM
Fri
09:15 PM Psychological tests,
to cognitive and social
06:30 AM stress tasks
Fig. 1. Participants from both groups first reported to the labora- laboratory at 09:00 PM. Batteries of psychological tests, and cogni-
tory on Monday morning (09:00 AM) to conduct a battery of psy- tive and social stress tasks were employed throughout the night
chological tests followed by the collection of saliva and blood sam- period. Participants of the control group reported to the labora-
ples. Three days later (Thursday), participants from the sleep-de- tory on Friday 09:00 AM, and both groups underwent a final bat-
prived group were awakened at 04: 00 AM and reported to the tery of psychological tests, and provided saliva and blood samples.
on the day of the experiment (Thursday), participants of the stress scale, continuous performance test and text presentation task; (ii)
group were awakened at 04:00 AM and allowed to maintain their analog mood scale, dot probe while putting the nondominant
daily routine, while continuously being monitored for staying hand in ice water and continuous performance test; (iii) arithme-
awake (by wearing an Actigraph, Mini Motionlogger, Ambula- tic task, analog mood scale and social stress task, and (iv) group
tory Monitoring Inc. and scheduled phone calls). At 09: 00 PM, text presentation.
these participants reported to the laboratory and were fully active
throughout the night until 10:00 AM Friday morning. At 4 time Saliva Collection
points during the night [(i) 09:15–11:25 PM, and (ii) 00:00–02:25, Saliva was collected using Salivatte swabs (Sarstedt, Germany)
(iii) 02:40–06:00 and (iv) 06:30–08:25 AM], different batteries of that were chewed for 60 s and placed back in the saliva tube. To
psychological tests and social stress tasks were employed (see be- overcome potential circadian-related cortisol alterations, the ex-
low). Participants of the control group reported to the laboratory act timing of saliva collection from each participant was matched
on Friday 09:00 AM, when participants of both the stress and the along the 2 collection days and counterbalanced between groups.
control groups underwent a final battery of psychological tests, Salivattes were then centrifuged at 1,000 g for 2 min, and the ex-
and provided the second set of saliva and blood samples. The time tract was collected and stored at –32 ° C for later analyses.
points of saliva and blood collection were counterbalanced be-
tween groups and maintained constant per participant for the Blood Collection
2 collection days in each replicate. Ten milliliter of blood were drawn by a certified nurse from
The following composition of tests and tasks were given the cubital vein into a heparinized syringe (50 units/ml) and
Tel-Aviv University Lib. of Life Science and Med.
through the nightly time points indicated above: (i) analog mood were kept at room temperature for a maximum of 2 h before be-
132.66.156.41 - 9/10/2013 10:29:08 AM
Depression (PHQ-9)
40 8
State anxiety
35 6
30 4
25 2
20 0
a Stress Control b Stress Control
Fig. 2. a Comparison of state anxiety levels between the control and levels, as indicated by PHQ-9 results, between the control and
sleep-deprived groups before (Monday) and after (Friday) the par- sleep-deprived groups before (Monday) and after (Friday) the par-
adigm. Significant differences were evident within the sleep-de- adigm. Significant differences were evident within the sleep-de-
prived group participants, but not within the control group par- prived group participants, but not within the control group par-
ticipants. Means ± SEM, p < 0.05. b Comparison of depression ticipants. Means ± SEM, p < 0.01.
A battery of both written and computerized tests was used to analysis of variance with a predetermined significance level of
assess trait and state anxiety, depression symptoms, dissociation, 0.05 was conducted. The independent between-participant vari-
behavior in stress situations, and coping styles and resilience. Ad- ables were group (stress/control) and sex (male/female). Time
ditionally, public presentations and social stress tasks were con- (before and after stress exposure) served as a repeated within-sub-
ducted during the night hours. All of the tests and tasks were given ject factor. Specific data observations that deviated from their
in their validated Hebrew versions and included the State-Trait group mean by more than 3 SDs were excluded from statistical
Anxiety Inventory-Trait Scale (STAI-T), State-Trait Anxiety In- analysis. These are identified in the Results section. Provided sig-
ventory-State Scale (STAI-S), patient health questionnaire (PHQ- nificant group differences were found, Fisher’s protected least sig-
9), Clinician Administered Dissociative State Scale, analog mood nificant difference analyses were performed to compare specific
scales, dot probe, continuous performance test, text presentation pairs of groups based on a priori hypotheses. All statistical analy-
and social stress tasks. ses were conducted using StatView software (SAS Institute, San
Data from the STAI-T, STAI-S and PHQ-9 were also used for Francisco, Calif., USA).
validating the psychological effect of the stressor, and thus these
tests are described in detail below:
The STAI [43] consists of two 20-item scales measuring situ- Results
ational variations in levels of perceived anxiety (state anxiety).
Likert scales ranged from 1 (not at all/almost never) to 4 (extreme-
ly/almost always) and scores ranged from 20 to 80, with higher Stress Manipulation Check
scores reflecting higher anxiety. The translation of these scales to Significant group-by-time interaction effects were
Hebrew was done according to the instructions of Spielberger and found for state anxiety (F1, 39 = 11.99, p < 0.01; fig. 2a) and
Sharma [44] and was found to be reliable and valid for use with depression (as indicated by PHQ-9 results; F1, 39 = 17.96,
Israelis [45]. p < 0.01; fig. 2b). Additionally, no differences were found
The PHQ-9 [46] is a self-reported depression rating scale con-
sisting of 9 items on which the diagnosis of DSM depressive dis- between the two experimental groups for trait anxiety.
order is based (designed to assess the severity of depressive These findings indicate a robust psychological impact of
symptoms). Reliability and diagnostic validity of the PHQ-9 the experimental procedure, as they indicate an immedi-
were established in several studies [46, 47]. ate change in state anxiety and depression within the
stress group only, which was not evident at all in the trait
Statistical Analysis
Data from both replicates were combined for the analysis, as anxiety data. We further examined these findings using
no noticeable differences between the two replicates were ob- paired Student’s t tests between the two experimental
Tel-Aviv University Lib. of Life Science and Med.
served. For each index examined, a 2 × 2 × 2 repeated-measure groups, which indicated no differences before the para-
132.66.156.41 - 9/10/2013 10:29:08 AM
Means ± SEM of the different psychological measures before (Monday) and after (Friday) sleep deprivation, and statistical outcomes
indicating significant differences. * These measures showed significant interaction between group and time point, resulting in elevated
anxiety and PHQ-9 in the stress group following the paradigm. NS = No significant outcome.
Table 2. Results for white blood cell counts assessed 3 days before the paradigm and on the morning following it
Means ± SEM of white blood cell subpopulation count before (Monday) and after (Friday) sleep deprivation, and statistical outcomes
indicating significant differences. * Monocyte counts were significantly higher in males than in females regardless of group and time
point. NS = No significant outcome. a One male deviant subject was removed from the control granulocyte group.
digm and significant differences following it (p < 0.01). lem with one set of tubes, IL-6 and monocytes were only
All findings are presented in table 1. assessed in approximately half of the samples.
Leukocyte Counts. The experimental paradigm had no
Immunological Indices effect on the total leukocyte numbers or on the number
Assessments of NK cytotoxicity and FACS analysis of any leukocyte subtype. Further examination of cellular
were conducted on all 23 stress and 18 control partici- marker expression levels (see Methods) on all related leu-
pants, excluding 1 male subject from the stress group in kocyte subtypes yielded also no effect for sleep depriva-
whom the second blood withdrawal was unsuccessful. tion, although baseline levels were within the detection
For all cytokine and hormone measurements, we ran- limits and indicated individual differences. The number
domly and a priori chose 19–22 participants from the of lymphocytes, monocytes and granulocytes did not sig-
stress group, and matched those to 13–17 participants nificantly differ between the groups, nor between baseline
from the control group, enforcing similar representation values (Monday) and values after sleep deprivation (Fri-
for each replicate and sex. We conducted this random day) in both groups. When examined for gender differ-
selection in order to keep within the available number of ences, males had a significantly higher number of mono-
Tel-Aviv University Lib. of Life Science and Med.
wells in the ELISA plates. Also, due to a technical prob- cytes than females, independently of group condition
132.66.156.41 - 9/10/2013 10:29:08 AM
70
60
40
30
20
Control Fri.
Stress Mon.
10 Control Mon.
Stress Fri.
0
E:T 1 E:T 2 E:T 3 E:T 4 E:T 5 E:T 6 E:T 7 E:T 8
a E:T ratio (whole blood:K562)
80
70
60
K562 cytotoxicity (%)
50
(F1, 11 = 10.89, p < 0.01). No additional gender differences blood NPY levels between the groups, nor between levels
were found among the other leukocyte subpopulations at baseline and after sleep deprivation. Additionally, no
examined. All data are presented in table 2. gender differences were found among these hormone lev-
NK Cell Counts and Cytotoxic Activity. The experi- els. All levels are presented in table 3, including minimal
mental paradigm neither affected NK numbers or cyto- detection levels.
toxicity between the groups nor between both time points Cytokines. The experimental paradigm had neither an
(at baseline and after sleep deprivation). When examined effect on IL-6 and IL-10 levels between the groups, nor
for gender differences, males had significantly higher kill- between levels at baseline and after sleep deprivation. No
ing activity levels than females, independent of group significant sex differences were found in these indices. All
condition (F1, 36 = 8.49, p < 0.01; fig. 3). levels are presented in table 4, including minimal detec-
Hormone Levels. The experimental paradigm had no tion levels.
Tel-Aviv University Lib. of Life Science and Med.
Means ± SEM of hormonal levels before (Monday) and after (Friday) sleep deprivation, and statistical outcomes indicating signifi-
cant differences. NS = No significant outcome.
Table 4. Results of cytokine levels (in pg/ml) assessed 3 days before the paradigm and on the morning following it
Means ± SEM of cytokine levels before (Monday) and after (Friday) sleep deprivation and statistical outcomes indicating significant
differences. NS = No significant outcome.a Most, but not all, readings of IL-10 levels were below the minimal detection level provided
by the manufacturer.
tion levels of the different bioassays and showed the ex- the ability of young people to cope with occasional sleep
132.66.156.41 - 9/10/2013 10:29:08 AM
vitro cytokine ‘induced production’, but not by direct collection. Unlike other studies, in which blood samples
132.66.156.41 - 9/10/2013 10:29:08 AM
References
1 Ohayon MM: Epidemiology of insomnia: 10 Irwin MR, Wang M, Campomayor CO, et al: 16 Shoham S, Davenne D, Cady AB, et al: Re-
what we know and what we still need to learn. Sleep deprivation and activation of morning combinant tumor necrosis factor and inter-
Sleep Med Rev 2002;6:97–111. levels of cellular and genomic markers of in- leukin 1 enhance slow-wave sleep. Am J
2 Staunton H: Mammalian sleep. Naturwissen- flammation. Arch Intern Med 2006;166:1756– Physiol 1987;253:R142–R149.
schaften 2005;92:203–220. 1762. 17 Krueger JM, Walter J, Dinarello CA, et al: Sleep-
3 Laposky AD, Bass J, Kohsaka A, et al: Sleep 11 Aldabal L, Bahammam AS: Metabolic, endo- promoting effects of endogenous pyrogen (in-
and circadian rhythms: key components in crine, and immune consequences of sleep de- terleukin-1). Am J Physiol 1984; 246:R994–
the regulation of energy metabolism. FEBS privation. Open Respir Med J 2011;5:31–43. R999.
Lett 2008;582:142–151. 12 Matsumoto Y, Mishima K, Satoh K, et al: To- 18 Roger J, Broughton RDO (eds): Sleep, Arous-
4 Frank MG, Benington JH: The role of sleep in tal sleep deprivation induces an acute and al, and Performance. Boston, Birkhäuser,
memory consolidation and brain plasticity: transient increase in NK cell activity in 1992.
dream or reality? Neuroscientist 2006;12:477– healthy young volunteers. Sleep 2001;24:804– 19 Majde JA, Krueger JM: Links between the in-
488. 809. nate immune system and sleep. J Allergy Clin
5 Maquet P: The role of sleep in learning and 13 Shakhar K, Valdimarsdottir HB, Guevarra JS, Immunol 2005;116:1188–1198.
memory. Science 2001;294:1048–1052. et al: Sleep, fatigue, and NK cell activity in 20 Patel SR, Zhu X, Storfer-Isser A, et al: Sleep
6 Morselli L, Leproult R, Balbo M, et al: Role of healthy volunteers: significant relationships duration and biomarkers of inflammation.
sleep duration in the regulation of glucose revealed by within subject analyses. Brain Be- Sleep 2009;32:200–204.
metabolism and appetite. Best Pract Res Clin hav Immun 2007;21:180–184. 21 Krueger JM, Obál FJ, Fang J, et al: The role of
Endocrinol Metab 2010;24:687–702. 14 Martins RC, Andersen ML, Garbuio SA, et al: cytokines in physiological sleep regulation.
7 Bollinger T, Bollinger A, Oster H, et al: Sleep, Dopamine transporter regulation during four Ann NY Acad Sci 2001;933:211–221.
immunity, and circadian clocks: a mechanis- nights of REM sleep deprivation followed by 22 Gamaldo CE, Shaikh AK, McArthur JC: The
tic model. Gerontology 2010;56:574–580. recovery – an in vivo molecular imaging sleep-immunity relationship. Neurol Clin
8 Muzur A: Toward an integrative theory of study in humans. Sleep 2010;33:243–251. 2012;30:1313–1343.
sleep and dreaming. J Theor Biol 2005; 233: 15 Julkunen I, Melén K, Nyqvist M, et al: Inflam- 23 Motivala SJ: Sleep and inflammation: psycho-
103–118. matory responses in influenza A virus infec- neuroimmunology in the context of cardio-
9 Irwin M: Effects of sleep and sleep loss on im- tion. Vaccine 2000;19(suppl 1):S32–S37. vascular disease. Ann Behav Med 2011; 42:
munity and cytokines. Brain Behav Immun 141–152.
Tel-Aviv University Lib. of Life Science and Med.
2002;16:503–512.
132.66.156.41 - 9/10/2013 10:29:08 AM