- Email: [email protected]
Hippocampal EEG theta power density is similar during slow-wave sleep and paradoxical sleep. A long-term study in rats
ELSEVIER
Neuroscience Letters 172 (1994) 31 34
NEUROSClENCE LETTERS
Hippocampal EEG theta power density is similar during slow-wave sleep and paradoxical sleep. A long-term study in rats J.M. Gaztelu a'*, M. Romero-Vives ", V. Abraira a, E. Garcia-Austt b "Hospital Ram6n y Cajal, S" Neurologia Experimental, Dpto. Investigaci6n, Crta. Cohnenar, kin. 9, 28034 Madrid, Spain hFacultad de Ciencias, Montevideo, Uruguay Received 13 October 1993; Revised version received 7 February 1994; Accepted 1 March 1994
Abstract
Rat hippocampal EEG and motor activity were studied during 15 days in relation to the vigilance state and to the light-dark cycle with a computerized system. During slow-wave sleep (SS) the hippocampal EEG has an outstanding mean power density in the theta band, similar to the large values present during paradoxical sleep. A circadian modulation was found for motor activity in SS and wakefulness (i.e. day-sleep is more restless, day-wakefulness is more quiet than night), and for EEG mean theta power in SS (i.e. less during day-sleep than night-sleep). These data underline the importance of analyzing the dark period when studying nocturnal animals. Key words. Theta band; Motor activity; Slow-wave sleep; Circadian modulation: Hippocampus; Rat: Spectral analysis
The most striking feature of the rat hippocampal E E G is the occurrence of a clear rhythmicity in the range of 4-10 Hz, the so-called theta rhythm [6], which is prominent in wakefulness (W) and during paradoxical sleep (PS) and which has been extensively studied. Despite the fact that the rat is a nocturnal animal, most of the studies of hippocampal physiology have been performed during the day, without considering possible circadian influences. The purpose of this study was to analyze the changes of hippocampal E E G frequency content in relation both to the vigilance state (W, SS and PS) and to the light-dark cycle (L/D), during 15 consecutive days of recording. Although little theta rhythm has been observed in the hippocampal E E G during SS, we have found that this E E G has a substantial amount of power density in the theta band during this state, in particular in the nocturnal period heretofore unexplored. We have made a preliminary presentation on the subject [4]. Five albino male rats were kept in clear plastic cages (30 cm diameter x 40 cm height) and individually studied
* Corresponding author. Fax: (34) (1) 336-9016. 0304-3940/94/$7.00 © 1994 Elsevier Science Ireland Ltd. All rights reserved SSDI 0304-3940(94)00213-T
within a video-monitored, sound-attenuating, chamber with continuous low intensity red light, under a controlled 12:12 L/D. They were implanted for chronic recordings of the hippocampal E E G and of the temporalis muscle electromyogram (EMG) [7] and connected to a counterbalanced swivel with flexible leads. Head movements were recorded with a modified small electret microphone, generating a movement signal. An activity signal (ACT), resulting from adding the movement signal and the rectified EMG, was obtained. Signals were amplified with a polygraph. The filtered E E G and ACT were fed to a computer. A program sampled signals at 32 Hz, calculated the E E G power spectral density at delta ( 1 4 Hz) and theta (5-9 Hz) bands and ACT, in 1 s epochs, and determined the vigilance state of those periods [5] by an algorithm that compared E E G delta power and ACT with two threshold values (T1 and T2, respectively). These were adjusted for each rat during the adaptation to the experimental conditions, so that the staged states corresponded to the observed behavior and recordings. When delta power was larger than T1, the state for that period was SS. When it was smaller than T1, the activity was considered, determining W when ACT exceeded T2, and PS when it did not. The perform-
32
.I M, (;a:zcht ez ~d. !,\;~'uro~'c]('tt~( ~ Lctter~ 172
1W)4, 31 34
A J~in
i i i
i i
i
i
i f
i
i i I i i i
i
i
i i
i
i i
i
i
i i
i i i i
i i i i
i
i i i i i i i i i i i l l
i
f l l l l l l l l t
PS
ss l W
1-4
Hz
5-9
Hz ACT
.
.
.
.
•
B I l l l l l l l l l l l l i i l l l l l l l l l l l l f l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l
]
- _ _
Fig. 1. Vigilance states, hippocampal EEG power spectral density in the delta and theta bands, and activity, in the dark (A) and light (B) periods. Values are plotted for 24 s epochs for the vigilance states, [wakefulness (W), slow-wave sleep (SS), paradoxical sleep (PS)] and for 4 s epochs for the other variables, power density in the delta ( 1 4 Hz) and theta (5-9 Hz) bands, and activity (ACT). Calibrations: delta and theta, t75/tV/~/~; ACT, arbitrary units; time, min.
ance of the staging system was visually checked almost daily. Data were printed on-line every 24 s as follows: the predominant vigilance state during a 24 s period was plotted as a dash, and, below it, 6 values of the other variables averaged every 4s (Fig. 1). Data were also stored on the computer hard disk for further revision and processing. An off-line correction of the observed staging inconsistencies was performed in all cases (i.e. in an evaluation of 4750 epochs, less than 3.35% of epochs needed correction). Under our conditions for recording, W was quite homogeneous with respect to EEG power in the theta band; walking episodes were brief and sporadic. Thirty minutes of data corresponding to well identified epochs of W, SS and PS, respectively, were selected at random for three representative days (i.e. 3, 9 and 15) of each rat. They were analyzed off-line by multiple lineal regression (MLR) in two ways, depending on whether differences between light and dark conditions were considered or not. In each case, three MLR analyses were performed, with delta band power, theta band power and ACT as dependent variables, and rat, day, L/D, vigilance state, and interaction L/D-vigilance state as independent variables. Rat number, day and vigilance state were codified as dummy variables, taking as reference values, rat no. 1, day no. 1 and W. In all regression models the analysis of normalized residues showed a non-normal
distribution. Therefore, all comparisons between variables were made with Tchebyshev inequality [8], a very conservative test. Differences were considered significant when t -> 4.47, corresponding to P -< 0.05. The three dependent variables studied yielded mean values for the EEG spectral power in bands delta and theta, and for the motor activity. These values were not significantly different in the five animals nor in the three analyzed days. Therefore, variables 'day' and 'rat' were excluded from the final models. When data were analyzed in 24 h periods, i.e. without considering the photoperiod (Fig. 2; Fig. 3. left histograms), spectral power in the theta band was larger in SS than in W (t = 35.44). It was also larger in PS than in W (t = 35.36), but differences between PS and SS were not significant (t = 0.39). It is noteworthy that, although the hippocampal EEG does not show a sustained theta rhythm during SS, mean spectral power in the theta band in this state showed a value similar to the one found in PS, this condition exhibiting a theta rhythm with the largest rhythmicity (Fig. 2). The large theta power in SS does not make a theta peak because, instead of lower power just above and below it, as in W and PS, during SS there is a greater increase in those neighboring bands (alpha and delta) than in the theta band. Spectral power in the delta band displayed larger values in SS than in W
33
J.M. Gaztelu et al./Neuroscience Letters 172 (1994) 31 34
(t = 38.79) a n d PS (t = 21.90), c o r r e s p o n d i n g to the characteristic E E G slow-wave activity p r e s e n t in SS. D e l t a b a n d p o w e r was n o t statistically different b e t w e e n W a n d PS (t = 1.27). W h e n d a t a were a n a l y z e d in 12 h periods, i.e. c o n s i d ering L / D (Fig. 3, right h i s t o g r a m s ) , an i n t e r a c t i o n between v a r i a b l e s ' L / D ' a n d 'vigilance state' was found. S l o w - w a v e sleep t h e t a b a n d p o w e r was larger in the d a r k t h a n in the light p e r i o d (t = 7.12). Conversely, activity in SS was smaller in the d a r k t h a n in the light p e r i o d (t = 4.70). Therefore, d a y slow-wave sleep is m o r e restless a n d has lower m e a n E E G t h e t a p o w e r t h a n night sleep. In W, activity was larger in the d a r k t h a n in the light p e r i o d (t = 20.70). C o n s e q u e n t l y , a c i r c a d i a n m o d u l a t i o n was o b s e r v e d in the a b o v e cases. Spectral p o w e r in the theta b a n d was larger in SS t h a n in W b o t h in the d a r k (t = 34.38) a n d in the light p e r i o d (t = 15.87). A m a j o r c o n c l u s i o n o f this study, p r e v i o u s l y unrep o r t e d , is t h a t d u r i n g SS the h i p p o c a m p a l E E G has an o u t s t a n d i n g m e a n p o w e r d e n s i t y in the t h e t a b a n d , similar to the large values p r e s e n t d u r i n g p a r a d o x i c a l sleep, being larger d u r i n g b o t h sleep stages t h a n d u r i n g W (Fig. 3). This is in a c c o r d a n c e with the c o n c e p t expressed b y A r n o l d s et al. [1], t h a t the h i p p o c a m p a l E E G s h o u l d n o t be c o n s i d e r e d in terms o f ' p r e s e n c e ' o r ' a b s e n c e ' o f t h e t a r h y t h m , as is d o n e in m o s t studies, b u t r a t h e r as a signal
A
THETA BAND 60 40
30
20 10
i
It
=
1
)
ACTIVITY 0 -5 -10 -15 -20 -25 -30 -35
B
=
|
|
80
60
40
40
~
tt
50
25
~
1
60
,o
~
|
DELTA BAND 70
W
I
0
6
12
18
24
30 30 20
21
10 0 ~
~ I
0
6
12
18
24
30
PS
ls
Hz
Fig. 2. Hippocampal EEG during the vigilance states. A: short samples of raw data during wakefulness (W), slow-wave sleep (SS) and paradoxical sleep (PS). B: power spectra of the EEG; 40 s were averaged. Stippled bar indicates the theta band (5 9 Hz). Note the high power in SS in this band, similar to PS and larger than in W.
W
II
~
SS
-20 4
P
SII
~
w
_~ II
ss
1 PS
Fig. 3. Mean hippocampal EEG power in the theta and delta bands and activity, with respect to the vigilance states, obtained by multiple lineal regression analyses, t,eft column, averages of 24 h in wakefulness (W), slow-wave sleep (SS) and paradoxical sleep (PS). Right column, considering light/dark periods separately, averages of 12 h in the dark (black bars) and light (white bars) periods. Histograms represent mean values (with their standard errors) of the EEG power m the theta and delta bands (upper and lower rows) and of activity (middle row), taking as zero references W (left) and W in the dark period (right). Note that without considering the photoperiod (left) mean power in the theta band was larger in SS than in W. Also (right), SS theta power was larger in the dark than in the light period, showing a circadian modulation. Note also that, conversely, SS activity is smaller in the dark than in the light period. Calibration: EEG power in the theta alld delta bands, ,uV/H',~zz;activity, arbitrary units. that d u r i n g different b e h a v i o r a l states c o n t a i n s a signific a n t a m o u n t o f activity in the theta b a n d . It can also be
34
j. .~I. Ga_-tehl ~,t a/. / Neuroscwnce Letters 172 :/1994) 31 34
stressed that the analysis performed in this study of rat hippocampal data during the night period has disclosed a heretofore unknown circadian modulation of the hippocampal electrographic activity during sleep. The high SS theta band power was a consistent finding in our experiments, independent of individual animals or days of recording, as was shown by M L R analysis. This method was selected as a convenient way to analyze simultaneously both the dependence between variables, and the existence of interactions between some of them. As a matter of fact, an interaction between variables ' L / D ' and 'vigilance state' was detected, implying that L/D has a different effect on E E G power and activity in the different vigilance states. Sensory information reaches the hippocampus during SS, as it has been demonstrated by the recording of sensory evoked potentials [2]. The existence of an important hippocampal E E G power in the theta band during SS could indicate that in this state the incoming information is probably processed in a similar way as during PS and W. It would be interesting to test the assumption that theta requires sensory input in SS as it does in W. This study provides an original characterization of several aspects of the hippocampal E E G correlated both with the vigilance states and the light-dark cycle. The differences noted should be further investigated in the domain of non-linear dynamics [3] and may provide important insights for the modelling of brain activity.
The authors wish to thank Dr. T.H. Bullock tbr helpful comments and for correcting the English style. This work was supported by FIS Grants 89/0162, 92/0207 and by CEC G r a n t ClI.0165.U(H). [1] Arnolds, D.E.A.T., Lopes da Silva, F.H., Aitink, J.W., Kamp, A. and Boeijinga, P., The spectral properties of hippocampal EEG related to behaviour in man. In G. Pfurtscheller, E Buser, F.H. Lopes da Silva and H. Petsche (Eds.), Rhythmic EEG Activitiesand Cortical Functioning, Developments in Neuroscience, Vol. 10, Elsevier, 1980, pp. 91 102. [2] Brankack, J. and Buzsaki, G., Hippocampal responses evoked by tooth pulp and acoustic stimulation: depth profiles and effect of behavior, Brain Res., 378 (1986) 303-314. [3] Bullock, T.H., Integrative system research on the brain: resurgence and new opportunities, Annu. Rev. Neurosci., 16 (1993) 1-15. [4] Gaztelu, J.M., Romero-Vives,M., Abraira, V. and Garcia-Austt, E., Hippocampal electrogram analysisduring sleep and wakefulness in long-term recordings in rats, Eur. J. Neurosci., Suppl. 6 (1993) 61. [5] Gaztelu, J.M., Abraira, V., Romero, M. and Garcia-Austt, E., Long-term, on-line, automated system for rat sleep staging and continuous hypnogram print-out, X Congress of the European Sleep Res. Soc., 1990. [6] Green, J.D. and Arduini, A.A., Hippocampal electrical activity in arousal, J. Comp. Neurol., 17 (1954) 533-557. [7] Rosenberg, R.S., Bergmann, B.M. and Rechtschaffen, A., Variations in slow wave activity during sleep in the rat, Physiol. Behav., 17 (1976) 931-938. [8] Walpole, R. and Myers, R.H., Prohabilidad y Estadistica, McGraw-Hill, 1986, 258 pp.
Neuroscience Letters 172 (1994) 31 34
NEUROSClENCE LETTERS
Hippocampal EEG theta power density is similar during slow-wave sleep and paradoxical sleep. A long-term study in rats J.M. Gaztelu a'*, M. Romero-Vives ", V. Abraira a, E. Garcia-Austt b "Hospital Ram6n y Cajal, S" Neurologia Experimental, Dpto. Investigaci6n, Crta. Cohnenar, kin. 9, 28034 Madrid, Spain hFacultad de Ciencias, Montevideo, Uruguay Received 13 October 1993; Revised version received 7 February 1994; Accepted 1 March 1994
Abstract
Rat hippocampal EEG and motor activity were studied during 15 days in relation to the vigilance state and to the light-dark cycle with a computerized system. During slow-wave sleep (SS) the hippocampal EEG has an outstanding mean power density in the theta band, similar to the large values present during paradoxical sleep. A circadian modulation was found for motor activity in SS and wakefulness (i.e. day-sleep is more restless, day-wakefulness is more quiet than night), and for EEG mean theta power in SS (i.e. less during day-sleep than night-sleep). These data underline the importance of analyzing the dark period when studying nocturnal animals. Key words. Theta band; Motor activity; Slow-wave sleep; Circadian modulation: Hippocampus; Rat: Spectral analysis
The most striking feature of the rat hippocampal E E G is the occurrence of a clear rhythmicity in the range of 4-10 Hz, the so-called theta rhythm [6], which is prominent in wakefulness (W) and during paradoxical sleep (PS) and which has been extensively studied. Despite the fact that the rat is a nocturnal animal, most of the studies of hippocampal physiology have been performed during the day, without considering possible circadian influences. The purpose of this study was to analyze the changes of hippocampal E E G frequency content in relation both to the vigilance state (W, SS and PS) and to the light-dark cycle (L/D), during 15 consecutive days of recording. Although little theta rhythm has been observed in the hippocampal E E G during SS, we have found that this E E G has a substantial amount of power density in the theta band during this state, in particular in the nocturnal period heretofore unexplored. We have made a preliminary presentation on the subject [4]. Five albino male rats were kept in clear plastic cages (30 cm diameter x 40 cm height) and individually studied
* Corresponding author. Fax: (34) (1) 336-9016. 0304-3940/94/$7.00 © 1994 Elsevier Science Ireland Ltd. All rights reserved SSDI 0304-3940(94)00213-T
within a video-monitored, sound-attenuating, chamber with continuous low intensity red light, under a controlled 12:12 L/D. They were implanted for chronic recordings of the hippocampal E E G and of the temporalis muscle electromyogram (EMG) [7] and connected to a counterbalanced swivel with flexible leads. Head movements were recorded with a modified small electret microphone, generating a movement signal. An activity signal (ACT), resulting from adding the movement signal and the rectified EMG, was obtained. Signals were amplified with a polygraph. The filtered E E G and ACT were fed to a computer. A program sampled signals at 32 Hz, calculated the E E G power spectral density at delta ( 1 4 Hz) and theta (5-9 Hz) bands and ACT, in 1 s epochs, and determined the vigilance state of those periods [5] by an algorithm that compared E E G delta power and ACT with two threshold values (T1 and T2, respectively). These were adjusted for each rat during the adaptation to the experimental conditions, so that the staged states corresponded to the observed behavior and recordings. When delta power was larger than T1, the state for that period was SS. When it was smaller than T1, the activity was considered, determining W when ACT exceeded T2, and PS when it did not. The perform-
32
.I M, (;a:zcht ez ~d. !,\;~'uro~'c]('tt~( ~ Lctter~ 172
1W)4, 31 34
A J~in
i i i
i i
i
i
i f
i
i i I i i i
i
i
i i
i
i i
i
i
i i
i i i i
i i i i
i
i i i i i i i i i i i l l
i
f l l l l l l l l t
PS
ss l W
1-4
Hz
5-9
Hz ACT
.
.
.
.
•
B I l l l l l l l l l l l l i i l l l l l l l l l l l l f l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l
]
- _ _
Fig. 1. Vigilance states, hippocampal EEG power spectral density in the delta and theta bands, and activity, in the dark (A) and light (B) periods. Values are plotted for 24 s epochs for the vigilance states, [wakefulness (W), slow-wave sleep (SS), paradoxical sleep (PS)] and for 4 s epochs for the other variables, power density in the delta ( 1 4 Hz) and theta (5-9 Hz) bands, and activity (ACT). Calibrations: delta and theta, t75/tV/~/~; ACT, arbitrary units; time, min.
ance of the staging system was visually checked almost daily. Data were printed on-line every 24 s as follows: the predominant vigilance state during a 24 s period was plotted as a dash, and, below it, 6 values of the other variables averaged every 4s (Fig. 1). Data were also stored on the computer hard disk for further revision and processing. An off-line correction of the observed staging inconsistencies was performed in all cases (i.e. in an evaluation of 4750 epochs, less than 3.35% of epochs needed correction). Under our conditions for recording, W was quite homogeneous with respect to EEG power in the theta band; walking episodes were brief and sporadic. Thirty minutes of data corresponding to well identified epochs of W, SS and PS, respectively, were selected at random for three representative days (i.e. 3, 9 and 15) of each rat. They were analyzed off-line by multiple lineal regression (MLR) in two ways, depending on whether differences between light and dark conditions were considered or not. In each case, three MLR analyses were performed, with delta band power, theta band power and ACT as dependent variables, and rat, day, L/D, vigilance state, and interaction L/D-vigilance state as independent variables. Rat number, day and vigilance state were codified as dummy variables, taking as reference values, rat no. 1, day no. 1 and W. In all regression models the analysis of normalized residues showed a non-normal
distribution. Therefore, all comparisons between variables were made with Tchebyshev inequality [8], a very conservative test. Differences were considered significant when t -> 4.47, corresponding to P -< 0.05. The three dependent variables studied yielded mean values for the EEG spectral power in bands delta and theta, and for the motor activity. These values were not significantly different in the five animals nor in the three analyzed days. Therefore, variables 'day' and 'rat' were excluded from the final models. When data were analyzed in 24 h periods, i.e. without considering the photoperiod (Fig. 2; Fig. 3. left histograms), spectral power in the theta band was larger in SS than in W (t = 35.44). It was also larger in PS than in W (t = 35.36), but differences between PS and SS were not significant (t = 0.39). It is noteworthy that, although the hippocampal EEG does not show a sustained theta rhythm during SS, mean spectral power in the theta band in this state showed a value similar to the one found in PS, this condition exhibiting a theta rhythm with the largest rhythmicity (Fig. 2). The large theta power in SS does not make a theta peak because, instead of lower power just above and below it, as in W and PS, during SS there is a greater increase in those neighboring bands (alpha and delta) than in the theta band. Spectral power in the delta band displayed larger values in SS than in W
33
J.M. Gaztelu et al./Neuroscience Letters 172 (1994) 31 34
(t = 38.79) a n d PS (t = 21.90), c o r r e s p o n d i n g to the characteristic E E G slow-wave activity p r e s e n t in SS. D e l t a b a n d p o w e r was n o t statistically different b e t w e e n W a n d PS (t = 1.27). W h e n d a t a were a n a l y z e d in 12 h periods, i.e. c o n s i d ering L / D (Fig. 3, right h i s t o g r a m s ) , an i n t e r a c t i o n between v a r i a b l e s ' L / D ' a n d 'vigilance state' was found. S l o w - w a v e sleep t h e t a b a n d p o w e r was larger in the d a r k t h a n in the light p e r i o d (t = 7.12). Conversely, activity in SS was smaller in the d a r k t h a n in the light p e r i o d (t = 4.70). Therefore, d a y slow-wave sleep is m o r e restless a n d has lower m e a n E E G t h e t a p o w e r t h a n night sleep. In W, activity was larger in the d a r k t h a n in the light p e r i o d (t = 20.70). C o n s e q u e n t l y , a c i r c a d i a n m o d u l a t i o n was o b s e r v e d in the a b o v e cases. Spectral p o w e r in the theta b a n d was larger in SS t h a n in W b o t h in the d a r k (t = 34.38) a n d in the light p e r i o d (t = 15.87). A m a j o r c o n c l u s i o n o f this study, p r e v i o u s l y unrep o r t e d , is t h a t d u r i n g SS the h i p p o c a m p a l E E G has an o u t s t a n d i n g m e a n p o w e r d e n s i t y in the t h e t a b a n d , similar to the large values p r e s e n t d u r i n g p a r a d o x i c a l sleep, being larger d u r i n g b o t h sleep stages t h a n d u r i n g W (Fig. 3). This is in a c c o r d a n c e with the c o n c e p t expressed b y A r n o l d s et al. [1], t h a t the h i p p o c a m p a l E E G s h o u l d n o t be c o n s i d e r e d in terms o f ' p r e s e n c e ' o r ' a b s e n c e ' o f t h e t a r h y t h m , as is d o n e in m o s t studies, b u t r a t h e r as a signal
A
THETA BAND 60 40
30
20 10
i
It
=
1
)
ACTIVITY 0 -5 -10 -15 -20 -25 -30 -35
B
=
|
|
80
60
40
40
~
tt
50
25
~
1
60
,o
~
|
DELTA BAND 70
W
I
0
6
12
18
24
30 30 20
21
10 0 ~
~ I
0
6
12
18
24
30
PS
ls
Hz
Fig. 2. Hippocampal EEG during the vigilance states. A: short samples of raw data during wakefulness (W), slow-wave sleep (SS) and paradoxical sleep (PS). B: power spectra of the EEG; 40 s were averaged. Stippled bar indicates the theta band (5 9 Hz). Note the high power in SS in this band, similar to PS and larger than in W.
W
II
~
SS
-20 4
P
SII
~
w
_~ II
ss
1 PS
Fig. 3. Mean hippocampal EEG power in the theta and delta bands and activity, with respect to the vigilance states, obtained by multiple lineal regression analyses, t,eft column, averages of 24 h in wakefulness (W), slow-wave sleep (SS) and paradoxical sleep (PS). Right column, considering light/dark periods separately, averages of 12 h in the dark (black bars) and light (white bars) periods. Histograms represent mean values (with their standard errors) of the EEG power m the theta and delta bands (upper and lower rows) and of activity (middle row), taking as zero references W (left) and W in the dark period (right). Note that without considering the photoperiod (left) mean power in the theta band was larger in SS than in W. Also (right), SS theta power was larger in the dark than in the light period, showing a circadian modulation. Note also that, conversely, SS activity is smaller in the dark than in the light period. Calibration: EEG power in the theta alld delta bands, ,uV/H',~zz;activity, arbitrary units. that d u r i n g different b e h a v i o r a l states c o n t a i n s a signific a n t a m o u n t o f activity in the theta b a n d . It can also be
34
j. .~I. Ga_-tehl ~,t a/. / Neuroscwnce Letters 172 :/1994) 31 34
stressed that the analysis performed in this study of rat hippocampal data during the night period has disclosed a heretofore unknown circadian modulation of the hippocampal electrographic activity during sleep. The high SS theta band power was a consistent finding in our experiments, independent of individual animals or days of recording, as was shown by M L R analysis. This method was selected as a convenient way to analyze simultaneously both the dependence between variables, and the existence of interactions between some of them. As a matter of fact, an interaction between variables ' L / D ' and 'vigilance state' was detected, implying that L/D has a different effect on E E G power and activity in the different vigilance states. Sensory information reaches the hippocampus during SS, as it has been demonstrated by the recording of sensory evoked potentials [2]. The existence of an important hippocampal E E G power in the theta band during SS could indicate that in this state the incoming information is probably processed in a similar way as during PS and W. It would be interesting to test the assumption that theta requires sensory input in SS as it does in W. This study provides an original characterization of several aspects of the hippocampal E E G correlated both with the vigilance states and the light-dark cycle. The differences noted should be further investigated in the domain of non-linear dynamics [3] and may provide important insights for the modelling of brain activity.
The authors wish to thank Dr. T.H. Bullock tbr helpful comments and for correcting the English style. This work was supported by FIS Grants 89/0162, 92/0207 and by CEC G r a n t ClI.0165.U(H). [1] Arnolds, D.E.A.T., Lopes da Silva, F.H., Aitink, J.W., Kamp, A. and Boeijinga, P., The spectral properties of hippocampal EEG related to behaviour in man. In G. Pfurtscheller, E Buser, F.H. Lopes da Silva and H. Petsche (Eds.), Rhythmic EEG Activitiesand Cortical Functioning, Developments in Neuroscience, Vol. 10, Elsevier, 1980, pp. 91 102. [2] Brankack, J. and Buzsaki, G., Hippocampal responses evoked by tooth pulp and acoustic stimulation: depth profiles and effect of behavior, Brain Res., 378 (1986) 303-314. [3] Bullock, T.H., Integrative system research on the brain: resurgence and new opportunities, Annu. Rev. Neurosci., 16 (1993) 1-15. [4] Gaztelu, J.M., Romero-Vives,M., Abraira, V. and Garcia-Austt, E., Hippocampal electrogram analysisduring sleep and wakefulness in long-term recordings in rats, Eur. J. Neurosci., Suppl. 6 (1993) 61. [5] Gaztelu, J.M., Abraira, V., Romero, M. and Garcia-Austt, E., Long-term, on-line, automated system for rat sleep staging and continuous hypnogram print-out, X Congress of the European Sleep Res. Soc., 1990. [6] Green, J.D. and Arduini, A.A., Hippocampal electrical activity in arousal, J. Comp. Neurol., 17 (1954) 533-557. [7] Rosenberg, R.S., Bergmann, B.M. and Rechtschaffen, A., Variations in slow wave activity during sleep in the rat, Physiol. Behav., 17 (1976) 931-938. [8] Walpole, R. and Myers, R.H., Prohabilidad y Estadistica, McGraw-Hill, 1986, 258 pp.