Arq Neuropsiquiatr 2006;64(2-B):388-393

1Laboratório de Mapeamento Cerebral e Integração Sensório-Motora, Instituto de Psiquiatria (IPUB), Universidade Federal do Riode Janeiro, Brasil (UFRJ); 2COPPE, Universidade Federal do Rio de Janeiro, Brasil (UFRJ); 3Departamento de Biociências e AtividadeFísica, Escola de Educação Física e Desportos (EEFD), UFRJ; 4Laboratório de Neurociências (LIM-27) Faculdade de Medicina daUniversidade de São Paulo, SP Brasil; 5C o o rdenador Laboratório de Mapeamento Cerebral e Integração Sensório-Motora, IPUB,UFRJ; 6Laboratório de Mapeamento Cerebral e Integração Sensório-Motora, IPUB-UFRJ, Professor Pesquisador UFRJ/IPUB e UniversidadeCastelo Branco (PROCIMH-UCB).

Received 26 September 2005, received in final form 16 January 2006. Accepted 18 February 2006.

Dra. Camila Ferreira - Rua Delfina 47 / 104 - 20511-270 Rio de Janeiro RJ - Brasil. E-mail: camilaferreira@libero.it

ELECTROENCEPHALOGRAPHIC CHANGESAFTER ONE NIGHT OF SLEEP DEPRIVATION

Camila Ferreira1, Andréa Deslandes1, Helena Moraes1, Maurício Cagy2,Fernando Pompeu3, Luis Fernando Basile4, Roberto Piedade5, Pedro Ribeiro6

ABSTRACT - Total or partial sleep deprivation (SD) causes degrading effects on diff e rent cognitive and psy-chomotor functions that might be related to electrophysiological changes frequently observed. In the pre s-ent study, we investigated the effects of one night of sleep deprivation on waking EEG. Experimental pro-tocol consisted of re c o rding electroencephalographic data from eleven healthy young subjects before(baseline) and after (time 2) one night of sleep deprivation. A natural log transformation was carried outand showed a significant increase in theta T6 (p=0.041), O2 (p=0.018) and OZ (p=0.028); and delta T6(p=0.043) relative power; and a decrease in alpha Fp1 (p=0.040), F3 (p=0.013), Fp2 (p=0.033), T4 (p=0.050),T6 (p=0.018), O2 (p=0.011) and Oz (p=0.025) and beta (p=0.022) absolute power. These outcomes showthat the EEG power spectra, after sleep deprivation, exhibit site-specific diff e rences in particular fre q u e n-cy bands and corroborate for the premise of local aspects of brain adaptation after sleep deprivation,rather than global.

KEY WORDS: sleep deprivation, qEEG, power spectral analysis, cortical activity.

Respostas eletrencefalográficas após uma noite de privação de sono

RESUMO - Privação total ou parcial de sono causa efeitos deletérios em diferentes funções cognitivas e psi-comotoras, que podem estar relacionados às mudanças eletrofísiológicas frequentemente observadas. Nop resente estudo, investigou-se os efeitos de uma noite de privação de sono nas respostas eletre n c e f a l o-gráficas de repouso. O protocolo experimental consistiu na coleta e gravação dos dados do qEEG de onzesujeitos jovens e saudáveis antes (momento baseline) e após (momento 2) uma noite de privação de sono.Todos os dados sofreram transformação logarítmica, que evidenciou um aumento significativo nas potên-cias relativas de teta T6 (p=0,041), O2 (p=0,018) e OZ (p=0,028); e delta T6 (p=0,043). As bandas de fre-qüência mais rápidas alfa Fp1 (p=0,040), F3 (p=0,013), Fp2 (p=0,033), T4 (p=0,050), T6 (p=0,018), O2 (p=0,011),Oz (p=0,025) e beta (p=0,022) apresentaram reduções significativas na potência absoluta. Os re s u l t a d o sdemonstram que o espectro de potência exibe diferenças específicas em freqüências de bandas difere n t e s ,e corroboram para a premissa de que o cére b ro, após privação ou perda de sono, promove ajustes e adap-tações locais, ao invés de globais.

PALAVRAS-CHAVE: privação de sono, qEEG, análise do espectro de potência, atividade cortical.

Sleep is essential to proper brain functioning andthe lack of it results in perf o rmance impairment dur-ing everyday cognitive tasks1 - 7 Sleep deprivation (SD)seems to exert local and specific effects on the brain,rather than global. Recent electro e n c e p h a l o g r a p h i cand neuroimaging studies revealed that the pre -f rontal cortex (PFC) is more responsive to sleep dep-rivation then other brain areas2,8-12, which is expect-ed since the frontal region has a greater restorative

need than other areas of the brain. In fact, duringnon-REM sleep, in the rebound sleep after SD, thei n c rease in delta-wave frequency (1-4 Hz, slow waveactivity related to deep sleep) occurs first in the PFCand later on other areas of the cort e x1 3 , 1 4, suggesti n gthe importance of the initial re c o v e ry in this re g i o n .A few recent studies2 , 3 , 8 p e rceived that, contrary toall expectations, after sleep deprivation, the PFC wassignificantly more activated during a diff e rent tasks

Arq Neuropsiquiatr 2006;64(2-B) 389

s u p p o rting the hypothesis of local dynamic compen-s a t o ry changes in the brain electrical activity aftersleep deprivation.

Studies concerning quantitative electro e n c e p h a-lography (qEEG) are still contradictory and have diver-gent results. The majority of EEG investigations re-p o rts the effects of SD on the rebound sleep, andconsidering the extended literature on sleep, veryfew studies have tried to observe electro p h y s i o l o g i-cal changes during prolonged wakefulness1 5 - 1 9. EEGis a useful and sensitive tool related to cognitive pro-cesses of vigilance and, there f o re, it can be used tounderstand the consequences of sleep loss in the b r a i nelectrical activity and functional organization whileindividuals are still awake. It has been demonstrat-ed by EEG spectral analysis that SD alters cortical activ-ity during prolonged wakefulness1 5 - 1 9. Spectral pow-er analysis1 7 has revealed an increase in the absolutepower of theta, alpha and beta bands after sleepdeprivation; but the same authors1 8 o b s e rved later areduction in alpha, although theta remained incre a s-ed. Some studies also re p o rted a reduction in theabsolute power of all bands, except for delta, and asignificant increase in theta absolute power in fro n t a land temporal are a s1 1. Corsi-Cabre r a6 o b s e rved ani n c rease in alpha and beta absolute power in theoccipital area. As noted, studies involving EEG arestill very much at odds, probably due to diff e re n tmethodologies and individual variability.

T h e re f o re, the aim of this study was to observ eand attempt to elucidate the electro p h y s i o l o g i c a lchanges followed by one night of sleep deprivation.

METHODSubjects – The sample consisted of 11 individuals, 4

males and 7 females, with ages varying between 21 and 40years (30.8±6.0 years). All subjects were healthy, right-hand-ed, non-smokers, and were not making use of any psychoac-tive or psychotropi c substance at the time of the test. Toe n s u re that subjects did not present any impairment oftheir physical and mental health, a questionnaire wasapplied. The questionnaire also aimed at identifying foodintake, body temperature, fatigue, and drugs use, amongothers. Subjects were not allowed to consume any alco-holic beverages or caffeine. The Edinburgh inventory wasused to assess laterality and exclude left-handed individu-als from the experiment. All subjects signed a consent form ,where the experimental condition was thoroughly descri-bed. The Psychiatric Institute’s Ethics Committee appro v e dthe experiment.

Study and pro c e d u res– E l e c t rophysiological re c o rd i n go c c u rred in two diff e rent times: pre- and post-sleep depri-vation (which itself took all night long). Subjects arrived atthe laboratory by 8:00 p.m. and performed the first qEEG

re c o rding (time 1- baseline). During the whole period ofsleep deprivation (night), volunteers played games, watchedvideos and carried out re c reational activities, and were alsofed each three or four hours. Two experimenters who werere q u i red to ensure continuous wakefulness of the subjectsm o n i t o red sleep deprivation continuously. Subjects abstai-ned from smoking or drinking xanthine-containing bever-ages (coffee, tea, cola or soft drinks). No concurrent tre a t-ment was allowed during the study. At 7:00 a.m. (time 2-after sleep deprivation) of the subsequent morning, all sub-jects repeated the same identical q EEG recording.

E l e c t roencephalogram re c o rding – The study designrespected the International Pharmaco-EEG group guide-lines. Subjects were seated comfortably in a sound-light-attenuated room, while EEG was collected from 20 mono-polar derivations for five minutes (eyes closed, alert butresting). Data were collected with eyes closed in order toobserve the cortex electrical activity without any externalstimuli, minimizing possible visual artifacts. Electrodes werepositioned according to the International 10/20 System (re-f e rred to linked earlobes with ground at FPZ). All electro d eimpedances were kept below 5 k . The signal was ampli-fied with a gain of 22,000, analogical ly filtered between0.01 Hz (high-pass) and 100 Hz (low-pass), and sampled at2 4 0 Hz using a Braintech-3000® (EMSA-Medical Instru m e n t s ,Rio de Janeiro, RJ, Brazil) EEG acquisition system. The EEGwas re c o rded by means of the software ERP Acquisition(Delphi 5.0®, Borland-Inprise), developed at the BrainMapping and Sensorimotor Integration Lab, employing thefollowing digital filters: notch ( 6 0 Hz), high-pass of 0.3 H zand low-pass of 25 Hz. Visual inspection was employed fordetection and elimination of artifacts. Eye-movement (EOG)a rtifact was monitored with a bipolar electrode montageusing two 9-mm-diameter electrodes attached superior toand on the external cantus of the right eye.

Data analysis – At least two minutes of art i f a c t - f ree dataw e re extracted from the EEG’s total re c o rd for quantita-tive analysis. A Matlab 5 . 3® (The Mathworks Inc., M a s s a c h u-setts, USA) routine was implemented to perf o rm a spectralanalysis. For each of the 20-monopolar derivations, absoluteand relative powers (µV2) were computed for the delta( 1 . 0 – 3 . 5 Hz), theta (4.0–7.5 Hz), alpha (8.0–12.0 Hz), andbeta (13–25 Hz) frequency bands.

Statistical analysis – Data were expressed as mean±s . d .The statistical software SPSS for Windows was used for alldata analysis. Pre l i m i n a rydescriptive examinations re v e a l e dthat none of the evaluated EEG indexes had a normal dis-tribution. To obtain a better approximation of a norm a ldistribution, data from the power spectra were submittedto a log transformation (LOG 10, log absolute power). Signi-ficance levels were set at p 0.05.

RESULTSOne night of sleep deprivation caused a signifi-

cant power decrease in the frontal, temporal and

390 Arq Neuropsiquiatr 2006;64(2-B)

Fig 1. Alpha absolute power cortical map showing both frontal and occipital views. One night of sleep deprivation

caused a significant power decrease in frontal, temporal and occipital areas in this frequency band. Electrodes Fp1

(p=0.040), F3 (p=0.013), Fp2 (p=0.033), T4 (p=0.050), T6 (p=0.018), O2 (p=0.011) and Oz (p=0.025) illustrate re m a r k -

able reductions. The beta absolute power cortical map also presents a significant power decrease in temporal re g i o n ,

specially, in the T6 electrode (p=0.022) after one night of sleep deprivation (*p<0.05).

occipital areas of the alpha frequency band, in the

following electrodes: Fp1 (p=0.040), F3 (p=0.013), Fp2

(p=0.033), T4 (p=0.050), T6 (p=0.018), O2 (p=0.011)

and Oz (p=0.025). The beta frequency band also expe-

rienced power reduction in the temporal area (p=

0.022). Temporal delta (p=0.043) and temporal-occip-

ital theta T6 (p=0.041), O2 (p=0.018) and OZ (p=

0.028), instead, exhibited power increases after SD.

Cortical maps represented by Figures 1 and 2 repre-

sent the significant electrophysiological results found

in the present study

DISCUSSION

The real implication of sleep is still not entire l y

clear and there f o re the effects of sleep deprivationor sleep loss also cannot be fully understood yet. The

p resent study demonstrated that one night of sleep

deprivation caused specific electrophysiological chan-

ges, such as relative delta and theta power incre a s e ,and absolute alpha and beta power reduction. Our

findings are in agreement with previous re s u l t s5 , 6 , 1 1 , 1 8 , 2 0

which observed the same alterations in the brain elec-

trical activity. Electrophysiological alterations such as

Arq Neuropsiquiatr 2006;64(2-B) 391

an increase in theta relative power in occipital andtemporal areas, and in delta relative power in tem-poral regions were expected since the subjects werefeeling each time sleepier and less alert (no scalesw e re used, subjects simply were frequently askedabout their subjective feelings of sleepiness thro u g h-out the whole experiment). It is clear that the amountof theta becomes more pronounced as the period ofcontinuous wakefulness pro g resses; despite the factthat cognitive workload (e.g., task demands) re q u i re-ments did not change. Increased theta activity isf o u n d2 1 to correlate with increased workload, as wellas increased fatigue. However, the widespread aug-m e n t a t i o n2 2 of theta activity has more traditionallybeen associated with impaired alertness. In addition,these generalized low-frequency EEG patterns havebeen associated2 3 with decrements in vigilance tasks.Although changes in EEG theta activity can be inter-preted as something other than an indication of anindividual state of alertness (for example, a correla-tion has been observed between theta activity and

cognitive workload), experimental evidences have

shown that theta is in fact a tru s t w o rthyindicator of

great central nervous system inhibition associated24

with low arousal and diminished information pro-

cessing. Based on the present knowledge about sleep

loss and sleep deprivation, physiological arousal andp e rf o rmance; both EEG and behavioral findings ob-

s e rved in the present study are coherent with the lit-

e r a t u re. In sleep-deprivation paradigms, it has been

well established2 2 that prolonged wakefulness is ac-

companied by systematic power increases in theta

activity. Contextual stimulation is known to attenu-

a t e2 5 the effects of sleep deprivation, and thus, theimpact of sleepiness on alertness may have been re-

duced modestly preventing greater delta power in-

c rease (e.g., frank sleep episodes), while still not alto-

gether eliminating its appearance in slow-wave EEG.

This could help explain why delta power did not show

a large increase all over the scalp, but instead, was

limited to the temporal area (T6 electrode).

Fig 2. Theta relative power cortical map. The occipital view of this frequency band exhibits the power incre a s e

o b s e rved after one night of sleep deprivation in the following electrodes: T6 (p=0.041), O2 (p=0.018), and Oz (p=

0.028). An also significant power augmentation was observed in the delta band. The delta relative power cortical

map, in particular electrode T6 (p=0.043) displays the power changes observed after one night of sleep deprivation

(*p<0.05).

392 Arq Neuropsiquiatr 2006;64(2-B)

Alpha power during waking EEG is also of part i c-ular interest for the re s e a rch on sleep deprivation,sleep loss and sleepiness. Throughout active wake-fulness (eyes open), alpha power is usually low, butif the individual is severely fatigued2 6, it might incre a-se. However, in resting conditions (eyes closed), alphapower is high also when the subject is fully re s t e d .Several studies observed that during the transitionf rom resting conditions with eyes closed to sleeping,a gradual reduction in alpha power occurs, all togeth-er with an increase in theta power2 6 - 2 8. So, if the sub-jects experience pro g ressively higher sleepiness sen-sation as the hours of wakefulness proceed, the ob-s e rved reduction in alpha power is expected. Con-s e q u e n t l y, reduced alpha power during waking peri-ods may indicate a high motivation for sleeping;indeed it has been already found2 8 that subjectivesleepiness during sleeplessness periods correlates neg-atively with alpha power and positively with thetapower during the waking EEG after sleep depriva-tion. Also, the early stages of drowsiness, experiencedh e re by the subjects after prolonged wakefulness,a recharacterized by what it’s called an “alpha pow-er dropout”. The trains of alpha waves become lessand less continuous and the last alpha fragments giveplace to a low voltage pattern of mixed slow fre-quencies (e.g., delta and theta). Alpha activity coun-teracts with theta activity; i.e., a slowing of posteri-or alpha is substituted gradually by the posterior the-ta activity (and delta), as the motivation to sleep in-c reases. These adjustments in the spectrum toward sslower frequency bands were better observed hereby the temporal activity changes (T6 electrode), sincewe found an alpha and beta power reduction andan accordingly a theta and delta power increase inthis area. Since these changes are very common inthe elderly2 9 or in patients suffering from dementia,we can only speculate about the substantial corre l a-tion between these EEG abnormalities observed aftersleep deprivation (spectrum shift to slower fre q u e n-cies) and a cognitive deficit.

Some investigators also observed that the alphapower reduction after sleep deprivation may be re l a t-ed to a reduction of activation in sub-cortical brains t ru c t u res with general cortical activation pro p e rt i e s ,e.g., brain stem, midbrain, hypothalamus and otherp a rts of the limbic system, since these stru c t u res showa positive correlation between regional cere b r a lblood flow and cortical alpha power3 0 in resting con-ditions. Moreover, the positive correlation betweenbeta absolute power increase (e.g., temporal re g i o n )and the hours of prolonged wakefulness might bei n t e r p reted as an index of eff o rtand temporary acti-

vation. A speculative interpretation may be that theo b s e rved beta power increase reflects pro c e s s e sunderlying an increasing eff o rt to maintain wakeful-ness, i.e., the motivation to stay awake, that coinci-des16 with an increased feeling of sleepiness.

In conclusion, the present study demonstratedthat one night of sleep deprivation was associatedwith: (1) increased resting EEG theta and delta rela-tive power with a specific occipital/temporal focus,(2) decreased resting EEG beta absolute power in thetemporal area and (3) decreased resting EEG alphaabsolute power centered in the frontal, temporaland occipital areas of the cortex. The electro p h y s i o l o-gical adjustments of the brain after sleep depriva-tion or sleep loss are still not entirely clear, but it ishighly plausible to be local, rather than global, pro-cesses. Future re s e a rch is still necessary to re p l i c a t ethese findings and maybe correlate qEEG changeswith behavioral responses. The processes underlyingthese electroencephalographic changes are connect-ed to each other and might reflect the perf o rm a n c ei m p a i rments that commonly accompany the loss ordeprivation of sleep, leading us to our choice to goto sleep.

REFERENCES1. Franken P, Dijk DJ, Tobler I, Borbely AA. Sleep deprivation in rats:

e ffects on EEG power spectra, vigilance states, and cortical tempera-ture. Am J Physiol 1991;261:198-208.

2. D rummond SP, Brown GG. The effects of total sleep deprivation onc e rebral responses to cognitive performance. Neuro p s y c h o p h a r m a c o-logy 2001;25:68-73.

3. D rummond SP, Gillin JC, Brown GG. Increased cerebral response dur-ing a divided attention task following sleep deprivation. J Sleep Res2001;10:85-92.

4. D rummond SP, Brown GG, Gillin JC, Stricker JL, Wong EC, Buxton RB.A l t e red brain response to verbal learning following sleep deprivation.Nature 2000;10:655-657.

5. Patat A, Rosenzweig P, Enslen M, et al. Effects of a new slow re l e a s eformulation of caffeine on EEG, psychomotor and cognitive functionsin sleep-deprived subjects. Hum Psychopharmacol 2000;15:153-170.

6. C o r s i - C a b rera M, A rce C, Ramos J, Lorenzo I, Guevara MA. Time courseof reaction time and EEG while performing a vigilance task duringtotal sleep deprivation. Sleep 1996;19:563-569.

7. Batejat D, Lagarde D. Circadian rhythm and sleep deprivation: eff e c t son psychomotor performance. Med Sci Res 1992; 20:167-168.

8. Chee MW, Choo WC. Functional imaging of working memory after 24hr of total sleep deprivation. J Neurosci 2004;24:4560-4567.

9. Cajochen C, Foy R, Dijk DJ. Frontal predominance of a relative incre a s ein sleep delta and theta EEG activity after sleep loss in humans. SleepRes Online 1999;2:65-69.

10. Schwierin B, Achermann P, Deboer T, Oleksenko A, Borbely AA, To b l e rI. Regional diff e rences in the dynamics of the cortical EEG in the ratafter sleep deprivation. Clin Neurophysiol 1999;110:869-875.

11. F o rest G, Godbout R. Effects of sleep deprivation on performance andEEG spectral analysis in young adults. Brain Cogn 2000;43:195-200.

12. Jones K, Harrison Y. Frontal lobe function, sleep loss and fragmentedsleep. Sleep Med Rev 2001;5:463-475.

13. Maquet P. The role of sleep in learning and memory. Science 2001;2;294:1048-1052.

14. Werth E, Achermann P, Borbely AA. Brain topography of the humansleep EEG: antero-posterior shifts of spectral power. Neuro report 1996;20;8:123-127.

Arq Neuropsiquiatr 2006;64(2-B) 393

15. G a rc i a - G a rcia F, Beltran-Parrazal L, Jimenez-Anguiano A, Ve g a - G o n-zalez A, Dru c k e r-Colin R Manipulations during forced wakefulnesshave diff e rential impact on sleep arc h i t e c t u re,EEG power spectru m ,and Fos induction. Brain Res Bull 1998;47:317-324.

16. Strijkstra AM, Beersma DGM, Drayer B, Halbesma N, Dann S.Subjective sleepiness correlates negatively with global alpha (8-12 Hz)and positively with central frontla theta (4-8 Hz) frequencies in thehuman awake electroencephalogram. Neurosci Letters 2003;340:17-20.

17. C o r s i - C a b rera M, Ramos J, A rce C, Guevara MA, Ponce-de-Leon M,Lorenzo I. Changes in the waking EEG as a consequence of sleep andsleep deprivation. Sleep 1992;15:550-555.

18. Corsi-Cabrera M, Sanchéz AI, Del-Rio-Portilla Y, Villanueva Y, Pérez-Garcí. Effect of 38 h of sleep deprivation on the waking EEG in wom-em: sex differences. Internat J Psychophysiol 2003;50:213-224.

19. Ugalde E, Corsi-Cabrera M, Juarez J, Ramos J, A rce C. Waking elec-t roencephalogram activity as a consequence of sleep and total sleepdeprivation in the rat. Sleep 1994;17:226-230.

20. Kim H, Guilleminault C, Hong S, Kim D, Kim S, Go H, Lee S. Patternanalysis of sleep-deprived human EEG. J Sleep Res 2001;10:193-201.

21. Hankins TC, Wilson GF. A comparison of heart rate, eye activity, EEGand subjective measures of pilot mental workload during flight. Av i a tSpace Environ Med 1998;69:360-367.

22. Caldwell JA, Prazinko BF, Hall KK. The effects of body posture on re s t-ing electroencephalographic activity in sleep-deprived subjects. ClinNeurophysiol 2000;111:464-470.

23. Belyavin A, Wright NA. Changes in electrical activity of the brain withvigilance. Electroencephalogr Clin Neurophysiol 1987;66:137-144.

24. Ray WJ. The electrocortical system. In Cacioppo JT, Ta s s i n a r y, LG (eds).Principles of psychophysiology. New York: Cambridge University Pre s s ,1990:382-412.

25. Dijkman M, Sachs N, Levine E, et al. Effects of reduced stimulation onn e u robehavioral alertness depend on circadian phase during humansleep deprivation. Sleep Res 1997;26:265-272.

26. Klimesch W. EEG alpha and theta oscillations reflect cognitive and me-mory performance: a review and analysis. Brain Res Brain Res Rev1999; 29:169-195.

27. Tanaka H, Hayashi M, Hori T. Topographical characteristics and prin-cipal component structure of the hypnagogic EEG. Sleep 1997;20:523-534.

28. Dijk DJ, Czeisler CA. Contribution of the circadian pacemaker and thesleep homeostasis to sleep pro p e n s i t y, sleep stru c t u re, electro e n c e p h a l o-graphic slow waves, and sleep spindle activity in humans. J Neurosci1995;15:3526-3538.

29. N i e d e r m e y e r, Lopes da Silva F. Electroencephalography: basic princi-ples, clinical applications and related fields. Philadelphia: Wilkins &Wilkins, 1993.

30. Sadato N, Nakamura S, Oohashi T, et al. Neural networks for genera-tion and suppression of alpha rhythm: a PET study. Neuro report 1998;9:893-897.