- Email: [email protected]
Effect of slow-wave sleep deprivation on topographical distribution of spindles
Behavioural Brain Research 116 (2000) 55 – 59 www.elsevier.com/locate/bbr
Research report
Effect of slow-wave sleep deprivation on topographical distribution of spindles Luigi De Gennaro *, Michele Ferrara, Mario Bertini Dipartimento di Psicologia, Sezione di Neuroscienze, Uni6ersita` degli Studi di Roma ‘La Sapienza’ Via dei Marsi, 78, 00185 Roma, Italy Received 27 January 2000; received in revised form 26 May 2000; accepted 26 May 2000
Abstract Spindle activity, visually scored in the 12–15 Hz range over antero-posterior midline derivations, has been assessed in ten normal subjects during a baseline and a recovery sleep after 2 nights of selective slow-wave sleep (SWS) deprivation. The recovery sleep, characterized by a significant increase of stage 4 and SWS as compared to the baseline, revealed that sleep spindles are reduced in the first NREM sleep cycle. The size of this reduction in spindle density progressively decreased in the course of the night, paralleling the depletion of SWS rebound. Topographical distribution of spindle activity showed a global linear increase over consecutive NREM-REM sleep cycles, regarding to the whole antero-posterior midline EEG derivations except the occipital one. It is concluded that the SWS rebound after 2 nights of selective SWS deprivation is associated with a reduction of spindles during stage 2. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Sleep spindles; Slow wave sleep (SWS) deprivation; SWS rebound; Topographical distribution; Intra-night variations; Sleep cycles
1. Introduction Recent neurophysiological findings have enhanced the interest in changes of sleep spindles: variations of the membrane potentials in the thalamocortical network oscillate in the frequency range of spindles at an intermediate level of hyperpolarization and in the frequency range of delta at an higher level of hyperpolarization [19]. It has been hypothesized that a close relationship between changes at neuronal level in the thalamocortical network and at the level of macroscopic EEG, with a reciprocal relationship between sleep spindles and slow waves [20,5]: the increasing hyperpolarization of thalamocortical neurons when sleep begins [11] should generate first the spindle activity during stage 2 and, later, the prevalence of delta rhythm during slow-wave sleep (SWS). * Corresponding author. Tel.: +39-06-49917647; fax: + 39-064451667. E-mail addresses: [email protected] (L. De Gennaro), [email protected] (M. Ferrara), [email protected] (M. Bertini).
Quantification of EEG in the frequency range of delta and sigma by means of fast Fourier transform (FFT) shows the existence of an inverse relationship between sigma and delta activity [1,6,21]. Furthermore, power density in the frequency range of sleep spindles has been demonstrated to be higher during stage 2 than SWS [3,5,14]. The procedure of sleep deprivation allows a direct assessment of this inverse relationship since it is well known that delta EEG activity increases after total sleep deprivation [3,5] and partial sleep deprivation [2,4,7]. Contrasting results have been provided by the only two studies that analyzed the relationship between delta and spindle activity after total sleep deprivation: in humans, power spectra analysis and transient patterns detection algorithms revealed that spindle activity was reduced in the recovery night after 40 h of wakefulness, while delta activity was enhanced [5]; in the cat, total sleep deprivation enhanced slow-wave activity (SWA) but did not reduce spindle activity in the cortical EEG, while EEGs derived from thalamic structures showed an inverse relationship between these two activ-
0166-4328/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 6 - 4 3 2 8 ( 0 0 ) 0 0 2 4 7 - 3
56
L. De Gennaro et al. / Beha6ioural Brain Research 116 (2000) 55–59
ities [15]. This dissociation between cortical and thalamic EEGs in the cat has been explained by an enhancement of background activity in the frequency range of sleep spindles which FFT can not discriminate from organized sleep spindles [15]. Due to the intrinsic limitations of FFT analysis in distinguishing phasic activity from background EEG activity, a more direct evaluation of phasic spindle activity after sleep deprivation is needed. The aim of the present study is to assess spindle activity, visually scored in multichannel EEG recordings, after a selective SWS deprivation that caused a strong rebound of stage 4 in the recovery sleep [8]. Spindle activity changes across consecutive NREM cycles will be also evaluated.
2. Methods
2.1. Participants Ten normal right-handed male students (mean age= 23.4 years (SEM= 0.87); age range = 20 – 30 years) were selected as paid volunteers for the study. They signed an informed consent before participating in the study. The requirements for inclusion were: normal sleep duration and schedule, no daytime nap habits, no excessive daytime sleepiness, no other sleep, medical or psychiatric disorder, as assessed by a 1-week sleep log and by a clinical interview.
2.2. Procedure The protocol of the study was reviewed and approved by the local Institutional Review Board. Participants slept for 6 consecutive nights in a sound-proof, temperature controlled room: (1) adaptation; (2) baseline (BSL); (3) baseline with awakenings (BSL-A); (4) SWS deprivation-1 (DEP-1); (5) SWS deprivation-2 (DEP-2); (6) recovery (REC). Data regarding only nights c 3 (BSL-A) and c6 (REC) have been considered in the present study. An Esaote Biomedica VEGA 24 polygraph set at a paper speed of 10 mm/s was used for polygraphic recordings. EEG signals were high pass filtered with a time constant of 0.3 s and low pass filtered at 30 Hz; seven unipolar EEG channels (C3-A2, C4-A1, Fpz-A1, Fz-A1, Cz-A1, Pz-A1, Oz-A1) were applied using the international 10–20 system. Submental EMG was recorded with a time constant of 0.03 s. Bipolar horizontal and vertical eye movements were recorded with a time constant of 1 s. Bipolar horizontal EOG was recorded from electrodes placed about 1 cm from the medial and lateral canthi of the dominant eye, and bipolar vertical EOG from electrodes located about 3 cm above and below the right eye pupil. Electrode impedance was kept below 5 KV.
Every night, sleep recording started at about 23:30 h and ended after 7.5 h of accumulated sleep. During nights c 3–6 participants were awakened twice, and a psychophysiological test battery was administered in bed [9]. Each awakening (comprising the final morning awakening) was carried out from stage 2; the first night-time awakening was scheduled after 2 h and the second after 5 h of accrued sleep, after at least 5 min of stage 2. At the end of testing (duration= 13 min) subjects were asked to go back to sleep. With regard to selective SWS deprivation, during the nights c4 and c 5 a tone was delivered (frequency: 1000 Hz; intensity: 40–110 dB-spl) whenever at least two delta waves (0.5–3.5 Hz; \ 75 mV) appeared in a 15-s recording interval. The acoustic stimulation technique allowed to set SWS amount near to zero during both the deprivation nights; as a consequence, during the following undisturbed recovery night a significant SWS rebound was found [8].
2.3. Data analysis Left central EEG (C3-A2), EMG, and horizontal and vertical EOG were used to visually score sleep stages, according to the standard criteria [17]; with regard to SWS scoring, the \ 75 mV amplitude criterion was strictly followed. Spindles during BSL and REC nights were scored by two blind scorers. Cases in which they were in disagreement were solved by a third blind scorer. According to Dijk et al. [5], the following criteria were applied to score spindles: minimum frequency 12 Hz, maximum frequency 15 Hz, minimum amplitude 10 mV, maximum amplitude 80 mV, minimum duration 0.5 s. Spindle density was computed during stage 2 of each sleep cycle as expressed by the ratio: number of spindles/duration of stage 2. The start of the first cycle was set at sleep onset, while the epoch that immediately follows the end of each REM episode gives the start for the following sleep cycle. A 4× 2×5 repeated measure ANOVA, Sleep cycle (1st, 2nd, 3rd, 4th) × night (BSL-R, REC)× derivation (Fpz, Fz, Cz, Pz, Oz) was carried out on the spindle density of NREM sleep (stage 2+stage 3+ stage 4), using the Greenhouse–Geisser correction. Ortogonalpolynomial contrast analysis or Duncan’s multiple range test (alpha level for critical ranges set at 0.01) were used to compare means of significant interactions.
3. Results An almost complete selective SWS suppression during both the deprivation nights was achieved without affecting sleep duration [8]; in fact, the mean percentage of SWS was 0.29 and 0.65% in DEP-1 and DEP-2,
L. De Gennaro et al. / Beha6ioural Brain Research 116 (2000) 55–59
respectively. Consequently, a significant stage 4 and SWS rebound in the recovery night was found. Stage 4 percentage increased from 5.69 (SEM= 1.62) to 11.01 (SEM =2.59), and SWS percentage increased from 11.42 (SEM= 1.93) to 16.97 (SEM=2.81).
57
Table 1 Results of orthogonal-polynomial contrast analysis across levels of the factor cycle in the cycle×derivation ANOVAa Linear
Quadratic
Cubic
Derivation
F(1,9)
P
F(1,9)
P
F(1,9)
P
Fpz Fz Cz Pz Oz
20.27 4.28 7.22 4.76 0.49
0.001 0.06 0.02 0.05 0.50
2.11 1.28 2.13 3.67 3.75
0.18 0.29 0.18 0.09 0.08
0.0006 0.004 0.04 0.04 0.02
0.99 0.95 0.84 0.84 0.89
a
Comparisons have been carried out for each level of the factor derivation.
Fig. 1. Means (and SEM) of spindle density (number of spindles/min) recorded during the first four sleep cycles of baseline and recovery nights.
A total of 80 085 spindles were scored during stage 2 sleep on antero-posterior midline leads; averaging over EEG derivations, mean spindle density was equal to 3.379 per min (SEM= 0.242: N=10). The comparison between BSL-A (M= 3.60; SEM= 0.37) and REC night (M = 3.14; SEM = 0.22) was not significant (F(1,9) = 1.30; P= 0.28), even though the significant night×sleep cycle interaction (F(3,27) =4.36; P= 0.01) indicated that the decrease of spindle density during REC night as compared to BSL-A was present in the first sleep cycle and progressively vanished in the course of the night (Fig. 1). In fact, the comparison between the 2 nights was significant in the first sleep cycle (F(1,9) = 6.28; P= 0.03), approached significance in second one (F(1,9) = 3.33; P= 0.10), and was not significant in the other sleep cycles. Spindle density showed a close to significance global increase across sleep cycle (F(3,27) = 2.69; P=0.07) and a significant main effect for the factor derivation (F(4,36) = 33.78; PB 0.00000001), indicating a centroparietal prevalence (Duncan’s multiple range test yielded the following significant differences: Cz=4.22, Pz= 4.17\ Fpz= 3.38\Fz=2.54, Oz=2.53). The sleep cycle × derivation interaction was also significant (F(12,108) = 6.34; PB 0.0000001) and is detailed in Fig. 2. Orthogonal-polynomial contrast analysis on spindle density revealed a significant linear increase across sleep cycles regarding the whole EEG derivations except Oz; quadratic and cubic components were not significant (Table 1).
4. Discussion
Fig. 2. Means (and SEM) of spindle density (number of spindles/min) recorded during stage 2 from the Fpz, Fz, Cz, Pz and Oz electrodes, as a function of the first four sleep cycles.
The present analysis of sleep spindles, visually scored in the 12–15 Hz range during stage 2, shows that spindle density is reduced in the first NREM sleep episode of the recovery sleep that follows 2 nights of selective SWS deprivation. Spindle activity also presents a global increasing linear trend over consecutive
58
L. De Gennaro et al. / Beha6ioural Brain Research 116 (2000) 55–59
NREM-REM sleep cycles on the whole antero-posterior midline EEG derivations with the exception of the occipital one. The size of the reduction in spindle density after the deprivation decreases in the course of the recovery sleep and parallels the depletion of SWS rebound. In fact, the recovery night is characterized by about a 50% of increase of SWS as compared to the baseline, and this sleep is recovered mostly in the first part of night [8]. This finding partially confirms data provided by the analysis of power spectra in the 12.75– 15.00 Hz range and by using a transient patterns detection algorithm: after a total sleep deprivation power spectra in the sigma frequency range were significantly reduced in the 2nd and 3rd NREM sleep cycles, while automatically detected spindle density was reduced in the first three NREM sleep cycles [5]. With regard to scoring spindle activity only during stage 2, we took this decision on the basis of the following considerations: (a) the visual scoring of spindles is less reliable during SWS (stages 3 and 4) due to the difficulty in detecting them superimposed on the high-amplitude delta activity; (b) the lack of SWS in the last sleep cycles and its increase in the recovery night as a consequence of SWS deprivation could be a confound respectively for the comparisons between sleep cycles and between nights, since SWS is characterized by a lower spindle density (e.g. [10,12]). Nevertheless, the same ANOVA design carried out on spindle density during stage 2 was also performed on spindle density of total NREM sleep (stage 2 +stage 3+ stage 4). The results of this ANOVA parallel those of the ANOVA regarding stage 2 sleep. Significant main effects were found for the cycle factor (F(3,27) = 6.38; P= 0.002) and for derivation (F(4,36) =36.44; PB 0.00000001). The interaction sleep cycle × derivation was again significant (F(12,108) =7.72; P B0.00000001) as well as the crucial sleep cycle× night interaction (F(3,27) = 2.86; P= 0.05). In this analysis too, the interaction indicated that the decrease of spindle density during the REC night as compared to BSL-A was present in the first sleep cycle and progressively vanished in the course of the night (BSL-A: 1st cycle = 2.94, 2nd cycle= 3.34, 3rd cycle= 3.50, 4th cycle = 3.32; REC: 1st cycle= 2.06, 2nd cycle= 2.70, 3rd cycle= 3.12, 4th cycle= 3.50). Intra-night variations of spindle density point out the existence of an increasing linear trend across NREM sleep cycles, reaching an asymptote during the 3rd or the 4th cycle, with the only exception of occipital lead that did not show any significant trend. This increase, previously found with a visual scoring [10,12] and with spectral analysis of sigma EEG frequencies [3,5,16,21], goes in an inverse direction as compared to the wellknown decrease of SWS and SWA across sleep cycles [3,5]. Furthermore, sigma and delta EEG power densities have been directly demonstrated to oscillate recip-
rocally within NREM sleep [1,6,21]. Therefore, there is converging evidence of the existence of an inverse relationship between the time course of delta and spindle activities, independent of the method used to quantify EEG phasic changes (visual scoring, period amplitude analysis or spectral analysis). The centroparietal topographical prevalence confirms findings obtained analyzing EEG power spectra in the 12.75–13.50 Hz range [22], in the 12–14 Hz range [13] or using automatic detectors [23,24]. The relatively lower values of spindle activity on frontal leads, as compared to the other studies, is explained by our lower limit for sleep spindle scoring (12 Hz) that does not take into account the contribution of low-frequency (11.5–12.25 Hz) spindle activity, peaking on frontal derivations [22,23]. However, since intracellular recordings have recently demonstrated that low and high frequency spindle activity is attributable to a single mechanism, depending on the different duration of the hyperpolarization-rebound sequence in thalamocortical neurons (e.g. [18]), the discrimination between these EEG rhythms lost most of its relevance for the present study. In conclusion, the analysis of visually scored sleep spindles allowed to overcome intrinsic limitations of spectral power analysis and of automatic spindle detectors in distinguishing between background and phasic EEG activity, since both methods also find spindles in stage 1 and in REM sleep [5,23], and to show that the rebound of SWS after 2 nights of selective sleep deprivation is associated with a reduction of spindles during stage 2. References [1] Aeschbach D, Borbely AA. All-night dynamics of the human sleep EEG. J Sleep Res 1993;2:70 – 81. [2] Akerstedt S, Gillberg M. Sleep duration and the power spectral density of the EEG. Electroencephalogr Clin Neurophysiol 1986;64:119 – 22. [3] Borbely AA, Baumann F, Brandeis D, Strauch I, Lehmann D. Sleep deprivation: effect on sleep stages and EEG power density in man. Electroencephalogr Clin Neurophysiol 1981;51:483–93. [4] Brunner DP, Dijk DJ, Borbely AA. Repeated partial sleep deprivation progressively changes the EEG during sleep and wakefulness. Sleep 1993;16:100 – 13. [5] Dijk DJ, Hayes B, Czeisler CA. Dynamics of electroencephalographic sleep spindles and slow wave activity in men: effect of sleep deprivation. Brain Res 1993;626:190 – 9. [6] Dijk DJ. EEG slow waves and sleep spindles: windows on the sleeping brain. Behav Brain Res 1995;69:109 – 16. [7] Feinberg I, Baker T, Leder R, March JD. Response of delta (0 – 3 Hz) EEG and eye movements density to a night with 100 min of sleep. Sleep 1988;11:473 – 87. [8] Ferrara M, De Gennaro L, Bertini M. Selective slow-wave sleep (SWS) deprivation and SWS rebound: do we need a fixed SWS amount per night? Sleep Res Online 1999;2:15 – 9. [9] Ferrara M, De Gennaro L, Bertini M, Casagrande M. Selective slow-wave sleep deprivation and time-of-night effects on cognitive performance upon awakening. Psychophysiology, in press.
L. De Gennaro et al. / Beha6ioural Brain Research 116 (2000) 55–59 [10] Guazzelli M, Feinberg I, Aminoff M, Fein G, Floyd TC, Maggini C. Sleep spindles in normal elderly: comparison with young adult patterns and relation to nocturnal awakening, cognitive function and brain atrophy. Electroencephalogr Clin Neurophysiol 1986;63:526 – 39. [11] Hirsch JC, Fourment A, Marc ME. Sleep-related variations of membrane potential in the lateral geniculate body relay neurons of the cat. Brain Res 1983;259:308–12. [12] Jankel WR, Niedermeyer E. Sleep spindles. J Clin Neurophysiol 1985;2:1 – 35. [13] Jobert M, Poiseau E, Jahnig P, Schulz H, Kubichi S. Topographical analysis of sleep spindle activity. Neuropsychobiol 1992;26:210 – 7. [14] Johnson LC, Lubin A, Nute C, Austin M. Spectral analysis of the EEG of dominant and non-dominant alpha-subjects during waking and sleeping. Electroencephalogr Clin Neurophysiol 1969;26:361 – 70. [15] Lancel M, Van Riezen H, Glatt A. The time course of sigma activity and slow-wave activity during NREMS in cortical and thalamic EEG of the cat during baseline and after 12 h of wakefulness. Brain Res 1992;596:285–95. [16] Landolt HP, Dijk DJ, Achermann P, Borbely AA. Effect of age on the sleep EEG: slow-wave activity and spindle frequency activity in young and middle-aged men. Brain Res 1996;738:205 – 12.
.
59
[17] Rechtschaffen A, Kales A. A manual of standardized terminology: techniques and scoring system for sleep stages of human subjects. Los Angeles: Brain Information Service, 1968. [18] Steriade M, Amzica F. Coalescence of sleep rhythms and their chronology in corticothalamic networks. Sleep Res Online 1998;1:1 – 10. [19] Steriade M, McCormick DA, Sejnowski TJ. Thalamocortical oscillations in the sleeping and aroused brain. Science 1993;262:679 – 85. [20] Steriade M, Nunez A, Amzica F. Intracellular analysis of relations between the slow ( B 1 Hz) neocortical oscillation and other sleep rhythms of the electroencephalogram. J Neurosc 1993;13:3266 – 83. [21] Uchida S, Maloney T, March JD, Azari R, Feiberg I. Sigma (12 – 15 Hz) and delta (0.3 – 3 Hz) EEG oscillate reciprocally within NREM sleep. Brain Res Bull 1991;27:93 – 6. [22] Werth E, Achermann P, Dijk DJ, Borbely AA. Spindle frequency activity in the sleep EEG: individual differences and topographic distribution. Electroencephalogr Clin Neurophysiol 1997;103:535 – 42. [23] Zeitlhofer J, Gruber G, Anderer P, et al. Topographic distribution of sleep spindles in young healthy subjects. J Sleep Res 1997;6:149 – 55. [24] Zygierewicz J, Blinowska KJ, Durka PJ, Szelenberger W, Niemcewicz S, Androsiuk W. High resolution study of sleep spindles. Electroencephalogr Clin Neurophysiol 1999;110:2136 – 47.
Research report
Effect of slow-wave sleep deprivation on topographical distribution of spindles Luigi De Gennaro *, Michele Ferrara, Mario Bertini Dipartimento di Psicologia, Sezione di Neuroscienze, Uni6ersita` degli Studi di Roma ‘La Sapienza’ Via dei Marsi, 78, 00185 Roma, Italy Received 27 January 2000; received in revised form 26 May 2000; accepted 26 May 2000
Abstract Spindle activity, visually scored in the 12–15 Hz range over antero-posterior midline derivations, has been assessed in ten normal subjects during a baseline and a recovery sleep after 2 nights of selective slow-wave sleep (SWS) deprivation. The recovery sleep, characterized by a significant increase of stage 4 and SWS as compared to the baseline, revealed that sleep spindles are reduced in the first NREM sleep cycle. The size of this reduction in spindle density progressively decreased in the course of the night, paralleling the depletion of SWS rebound. Topographical distribution of spindle activity showed a global linear increase over consecutive NREM-REM sleep cycles, regarding to the whole antero-posterior midline EEG derivations except the occipital one. It is concluded that the SWS rebound after 2 nights of selective SWS deprivation is associated with a reduction of spindles during stage 2. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Sleep spindles; Slow wave sleep (SWS) deprivation; SWS rebound; Topographical distribution; Intra-night variations; Sleep cycles
1. Introduction Recent neurophysiological findings have enhanced the interest in changes of sleep spindles: variations of the membrane potentials in the thalamocortical network oscillate in the frequency range of spindles at an intermediate level of hyperpolarization and in the frequency range of delta at an higher level of hyperpolarization [19]. It has been hypothesized that a close relationship between changes at neuronal level in the thalamocortical network and at the level of macroscopic EEG, with a reciprocal relationship between sleep spindles and slow waves [20,5]: the increasing hyperpolarization of thalamocortical neurons when sleep begins [11] should generate first the spindle activity during stage 2 and, later, the prevalence of delta rhythm during slow-wave sleep (SWS). * Corresponding author. Tel.: +39-06-49917647; fax: + 39-064451667. E-mail addresses: [email protected] (L. De Gennaro), [email protected] (M. Ferrara), [email protected] (M. Bertini).
Quantification of EEG in the frequency range of delta and sigma by means of fast Fourier transform (FFT) shows the existence of an inverse relationship between sigma and delta activity [1,6,21]. Furthermore, power density in the frequency range of sleep spindles has been demonstrated to be higher during stage 2 than SWS [3,5,14]. The procedure of sleep deprivation allows a direct assessment of this inverse relationship since it is well known that delta EEG activity increases after total sleep deprivation [3,5] and partial sleep deprivation [2,4,7]. Contrasting results have been provided by the only two studies that analyzed the relationship between delta and spindle activity after total sleep deprivation: in humans, power spectra analysis and transient patterns detection algorithms revealed that spindle activity was reduced in the recovery night after 40 h of wakefulness, while delta activity was enhanced [5]; in the cat, total sleep deprivation enhanced slow-wave activity (SWA) but did not reduce spindle activity in the cortical EEG, while EEGs derived from thalamic structures showed an inverse relationship between these two activ-
0166-4328/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 6 - 4 3 2 8 ( 0 0 ) 0 0 2 4 7 - 3
56
L. De Gennaro et al. / Beha6ioural Brain Research 116 (2000) 55–59
ities [15]. This dissociation between cortical and thalamic EEGs in the cat has been explained by an enhancement of background activity in the frequency range of sleep spindles which FFT can not discriminate from organized sleep spindles [15]. Due to the intrinsic limitations of FFT analysis in distinguishing phasic activity from background EEG activity, a more direct evaluation of phasic spindle activity after sleep deprivation is needed. The aim of the present study is to assess spindle activity, visually scored in multichannel EEG recordings, after a selective SWS deprivation that caused a strong rebound of stage 4 in the recovery sleep [8]. Spindle activity changes across consecutive NREM cycles will be also evaluated.
2. Methods
2.1. Participants Ten normal right-handed male students (mean age= 23.4 years (SEM= 0.87); age range = 20 – 30 years) were selected as paid volunteers for the study. They signed an informed consent before participating in the study. The requirements for inclusion were: normal sleep duration and schedule, no daytime nap habits, no excessive daytime sleepiness, no other sleep, medical or psychiatric disorder, as assessed by a 1-week sleep log and by a clinical interview.
2.2. Procedure The protocol of the study was reviewed and approved by the local Institutional Review Board. Participants slept for 6 consecutive nights in a sound-proof, temperature controlled room: (1) adaptation; (2) baseline (BSL); (3) baseline with awakenings (BSL-A); (4) SWS deprivation-1 (DEP-1); (5) SWS deprivation-2 (DEP-2); (6) recovery (REC). Data regarding only nights c 3 (BSL-A) and c6 (REC) have been considered in the present study. An Esaote Biomedica VEGA 24 polygraph set at a paper speed of 10 mm/s was used for polygraphic recordings. EEG signals were high pass filtered with a time constant of 0.3 s and low pass filtered at 30 Hz; seven unipolar EEG channels (C3-A2, C4-A1, Fpz-A1, Fz-A1, Cz-A1, Pz-A1, Oz-A1) were applied using the international 10–20 system. Submental EMG was recorded with a time constant of 0.03 s. Bipolar horizontal and vertical eye movements were recorded with a time constant of 1 s. Bipolar horizontal EOG was recorded from electrodes placed about 1 cm from the medial and lateral canthi of the dominant eye, and bipolar vertical EOG from electrodes located about 3 cm above and below the right eye pupil. Electrode impedance was kept below 5 KV.
Every night, sleep recording started at about 23:30 h and ended after 7.5 h of accumulated sleep. During nights c 3–6 participants were awakened twice, and a psychophysiological test battery was administered in bed [9]. Each awakening (comprising the final morning awakening) was carried out from stage 2; the first night-time awakening was scheduled after 2 h and the second after 5 h of accrued sleep, after at least 5 min of stage 2. At the end of testing (duration= 13 min) subjects were asked to go back to sleep. With regard to selective SWS deprivation, during the nights c4 and c 5 a tone was delivered (frequency: 1000 Hz; intensity: 40–110 dB-spl) whenever at least two delta waves (0.5–3.5 Hz; \ 75 mV) appeared in a 15-s recording interval. The acoustic stimulation technique allowed to set SWS amount near to zero during both the deprivation nights; as a consequence, during the following undisturbed recovery night a significant SWS rebound was found [8].
2.3. Data analysis Left central EEG (C3-A2), EMG, and horizontal and vertical EOG were used to visually score sleep stages, according to the standard criteria [17]; with regard to SWS scoring, the \ 75 mV amplitude criterion was strictly followed. Spindles during BSL and REC nights were scored by two blind scorers. Cases in which they were in disagreement were solved by a third blind scorer. According to Dijk et al. [5], the following criteria were applied to score spindles: minimum frequency 12 Hz, maximum frequency 15 Hz, minimum amplitude 10 mV, maximum amplitude 80 mV, minimum duration 0.5 s. Spindle density was computed during stage 2 of each sleep cycle as expressed by the ratio: number of spindles/duration of stage 2. The start of the first cycle was set at sleep onset, while the epoch that immediately follows the end of each REM episode gives the start for the following sleep cycle. A 4× 2×5 repeated measure ANOVA, Sleep cycle (1st, 2nd, 3rd, 4th) × night (BSL-R, REC)× derivation (Fpz, Fz, Cz, Pz, Oz) was carried out on the spindle density of NREM sleep (stage 2+stage 3+ stage 4), using the Greenhouse–Geisser correction. Ortogonalpolynomial contrast analysis or Duncan’s multiple range test (alpha level for critical ranges set at 0.01) were used to compare means of significant interactions.
3. Results An almost complete selective SWS suppression during both the deprivation nights was achieved without affecting sleep duration [8]; in fact, the mean percentage of SWS was 0.29 and 0.65% in DEP-1 and DEP-2,
L. De Gennaro et al. / Beha6ioural Brain Research 116 (2000) 55–59
respectively. Consequently, a significant stage 4 and SWS rebound in the recovery night was found. Stage 4 percentage increased from 5.69 (SEM= 1.62) to 11.01 (SEM =2.59), and SWS percentage increased from 11.42 (SEM= 1.93) to 16.97 (SEM=2.81).
57
Table 1 Results of orthogonal-polynomial contrast analysis across levels of the factor cycle in the cycle×derivation ANOVAa Linear
Quadratic
Cubic
Derivation
F(1,9)
P
F(1,9)
P
F(1,9)
P
Fpz Fz Cz Pz Oz
20.27 4.28 7.22 4.76 0.49
0.001 0.06 0.02 0.05 0.50
2.11 1.28 2.13 3.67 3.75
0.18 0.29 0.18 0.09 0.08
0.0006 0.004 0.04 0.04 0.02
0.99 0.95 0.84 0.84 0.89
a
Comparisons have been carried out for each level of the factor derivation.
Fig. 1. Means (and SEM) of spindle density (number of spindles/min) recorded during the first four sleep cycles of baseline and recovery nights.
A total of 80 085 spindles were scored during stage 2 sleep on antero-posterior midline leads; averaging over EEG derivations, mean spindle density was equal to 3.379 per min (SEM= 0.242: N=10). The comparison between BSL-A (M= 3.60; SEM= 0.37) and REC night (M = 3.14; SEM = 0.22) was not significant (F(1,9) = 1.30; P= 0.28), even though the significant night×sleep cycle interaction (F(3,27) =4.36; P= 0.01) indicated that the decrease of spindle density during REC night as compared to BSL-A was present in the first sleep cycle and progressively vanished in the course of the night (Fig. 1). In fact, the comparison between the 2 nights was significant in the first sleep cycle (F(1,9) = 6.28; P= 0.03), approached significance in second one (F(1,9) = 3.33; P= 0.10), and was not significant in the other sleep cycles. Spindle density showed a close to significance global increase across sleep cycle (F(3,27) = 2.69; P=0.07) and a significant main effect for the factor derivation (F(4,36) = 33.78; PB 0.00000001), indicating a centroparietal prevalence (Duncan’s multiple range test yielded the following significant differences: Cz=4.22, Pz= 4.17\ Fpz= 3.38\Fz=2.54, Oz=2.53). The sleep cycle × derivation interaction was also significant (F(12,108) = 6.34; PB 0.0000001) and is detailed in Fig. 2. Orthogonal-polynomial contrast analysis on spindle density revealed a significant linear increase across sleep cycles regarding the whole EEG derivations except Oz; quadratic and cubic components were not significant (Table 1).
4. Discussion
Fig. 2. Means (and SEM) of spindle density (number of spindles/min) recorded during stage 2 from the Fpz, Fz, Cz, Pz and Oz electrodes, as a function of the first four sleep cycles.
The present analysis of sleep spindles, visually scored in the 12–15 Hz range during stage 2, shows that spindle density is reduced in the first NREM sleep episode of the recovery sleep that follows 2 nights of selective SWS deprivation. Spindle activity also presents a global increasing linear trend over consecutive
58
L. De Gennaro et al. / Beha6ioural Brain Research 116 (2000) 55–59
NREM-REM sleep cycles on the whole antero-posterior midline EEG derivations with the exception of the occipital one. The size of the reduction in spindle density after the deprivation decreases in the course of the recovery sleep and parallels the depletion of SWS rebound. In fact, the recovery night is characterized by about a 50% of increase of SWS as compared to the baseline, and this sleep is recovered mostly in the first part of night [8]. This finding partially confirms data provided by the analysis of power spectra in the 12.75– 15.00 Hz range and by using a transient patterns detection algorithm: after a total sleep deprivation power spectra in the sigma frequency range were significantly reduced in the 2nd and 3rd NREM sleep cycles, while automatically detected spindle density was reduced in the first three NREM sleep cycles [5]. With regard to scoring spindle activity only during stage 2, we took this decision on the basis of the following considerations: (a) the visual scoring of spindles is less reliable during SWS (stages 3 and 4) due to the difficulty in detecting them superimposed on the high-amplitude delta activity; (b) the lack of SWS in the last sleep cycles and its increase in the recovery night as a consequence of SWS deprivation could be a confound respectively for the comparisons between sleep cycles and between nights, since SWS is characterized by a lower spindle density (e.g. [10,12]). Nevertheless, the same ANOVA design carried out on spindle density during stage 2 was also performed on spindle density of total NREM sleep (stage 2 +stage 3+ stage 4). The results of this ANOVA parallel those of the ANOVA regarding stage 2 sleep. Significant main effects were found for the cycle factor (F(3,27) = 6.38; P= 0.002) and for derivation (F(4,36) =36.44; PB 0.00000001). The interaction sleep cycle × derivation was again significant (F(12,108) =7.72; P B0.00000001) as well as the crucial sleep cycle× night interaction (F(3,27) = 2.86; P= 0.05). In this analysis too, the interaction indicated that the decrease of spindle density during the REC night as compared to BSL-A was present in the first sleep cycle and progressively vanished in the course of the night (BSL-A: 1st cycle = 2.94, 2nd cycle= 3.34, 3rd cycle= 3.50, 4th cycle = 3.32; REC: 1st cycle= 2.06, 2nd cycle= 2.70, 3rd cycle= 3.12, 4th cycle= 3.50). Intra-night variations of spindle density point out the existence of an increasing linear trend across NREM sleep cycles, reaching an asymptote during the 3rd or the 4th cycle, with the only exception of occipital lead that did not show any significant trend. This increase, previously found with a visual scoring [10,12] and with spectral analysis of sigma EEG frequencies [3,5,16,21], goes in an inverse direction as compared to the wellknown decrease of SWS and SWA across sleep cycles [3,5]. Furthermore, sigma and delta EEG power densities have been directly demonstrated to oscillate recip-
rocally within NREM sleep [1,6,21]. Therefore, there is converging evidence of the existence of an inverse relationship between the time course of delta and spindle activities, independent of the method used to quantify EEG phasic changes (visual scoring, period amplitude analysis or spectral analysis). The centroparietal topographical prevalence confirms findings obtained analyzing EEG power spectra in the 12.75–13.50 Hz range [22], in the 12–14 Hz range [13] or using automatic detectors [23,24]. The relatively lower values of spindle activity on frontal leads, as compared to the other studies, is explained by our lower limit for sleep spindle scoring (12 Hz) that does not take into account the contribution of low-frequency (11.5–12.25 Hz) spindle activity, peaking on frontal derivations [22,23]. However, since intracellular recordings have recently demonstrated that low and high frequency spindle activity is attributable to a single mechanism, depending on the different duration of the hyperpolarization-rebound sequence in thalamocortical neurons (e.g. [18]), the discrimination between these EEG rhythms lost most of its relevance for the present study. In conclusion, the analysis of visually scored sleep spindles allowed to overcome intrinsic limitations of spectral power analysis and of automatic spindle detectors in distinguishing between background and phasic EEG activity, since both methods also find spindles in stage 1 and in REM sleep [5,23], and to show that the rebound of SWS after 2 nights of selective sleep deprivation is associated with a reduction of spindles during stage 2. References [1] Aeschbach D, Borbely AA. All-night dynamics of the human sleep EEG. J Sleep Res 1993;2:70 – 81. [2] Akerstedt S, Gillberg M. Sleep duration and the power spectral density of the EEG. Electroencephalogr Clin Neurophysiol 1986;64:119 – 22. [3] Borbely AA, Baumann F, Brandeis D, Strauch I, Lehmann D. Sleep deprivation: effect on sleep stages and EEG power density in man. Electroencephalogr Clin Neurophysiol 1981;51:483–93. [4] Brunner DP, Dijk DJ, Borbely AA. Repeated partial sleep deprivation progressively changes the EEG during sleep and wakefulness. Sleep 1993;16:100 – 13. [5] Dijk DJ, Hayes B, Czeisler CA. Dynamics of electroencephalographic sleep spindles and slow wave activity in men: effect of sleep deprivation. Brain Res 1993;626:190 – 9. [6] Dijk DJ. EEG slow waves and sleep spindles: windows on the sleeping brain. Behav Brain Res 1995;69:109 – 16. [7] Feinberg I, Baker T, Leder R, March JD. Response of delta (0 – 3 Hz) EEG and eye movements density to a night with 100 min of sleep. Sleep 1988;11:473 – 87. [8] Ferrara M, De Gennaro L, Bertini M. Selective slow-wave sleep (SWS) deprivation and SWS rebound: do we need a fixed SWS amount per night? Sleep Res Online 1999;2:15 – 9. [9] Ferrara M, De Gennaro L, Bertini M, Casagrande M. Selective slow-wave sleep deprivation and time-of-night effects on cognitive performance upon awakening. Psychophysiology, in press.
L. De Gennaro et al. / Beha6ioural Brain Research 116 (2000) 55–59 [10] Guazzelli M, Feinberg I, Aminoff M, Fein G, Floyd TC, Maggini C. Sleep spindles in normal elderly: comparison with young adult patterns and relation to nocturnal awakening, cognitive function and brain atrophy. Electroencephalogr Clin Neurophysiol 1986;63:526 – 39. [11] Hirsch JC, Fourment A, Marc ME. Sleep-related variations of membrane potential in the lateral geniculate body relay neurons of the cat. Brain Res 1983;259:308–12. [12] Jankel WR, Niedermeyer E. Sleep spindles. J Clin Neurophysiol 1985;2:1 – 35. [13] Jobert M, Poiseau E, Jahnig P, Schulz H, Kubichi S. Topographical analysis of sleep spindle activity. Neuropsychobiol 1992;26:210 – 7. [14] Johnson LC, Lubin A, Nute C, Austin M. Spectral analysis of the EEG of dominant and non-dominant alpha-subjects during waking and sleeping. Electroencephalogr Clin Neurophysiol 1969;26:361 – 70. [15] Lancel M, Van Riezen H, Glatt A. The time course of sigma activity and slow-wave activity during NREMS in cortical and thalamic EEG of the cat during baseline and after 12 h of wakefulness. Brain Res 1992;596:285–95. [16] Landolt HP, Dijk DJ, Achermann P, Borbely AA. Effect of age on the sleep EEG: slow-wave activity and spindle frequency activity in young and middle-aged men. Brain Res 1996;738:205 – 12.
.
59
[17] Rechtschaffen A, Kales A. A manual of standardized terminology: techniques and scoring system for sleep stages of human subjects. Los Angeles: Brain Information Service, 1968. [18] Steriade M, Amzica F. Coalescence of sleep rhythms and their chronology in corticothalamic networks. Sleep Res Online 1998;1:1 – 10. [19] Steriade M, McCormick DA, Sejnowski TJ. Thalamocortical oscillations in the sleeping and aroused brain. Science 1993;262:679 – 85. [20] Steriade M, Nunez A, Amzica F. Intracellular analysis of relations between the slow ( B 1 Hz) neocortical oscillation and other sleep rhythms of the electroencephalogram. J Neurosc 1993;13:3266 – 83. [21] Uchida S, Maloney T, March JD, Azari R, Feiberg I. Sigma (12 – 15 Hz) and delta (0.3 – 3 Hz) EEG oscillate reciprocally within NREM sleep. Brain Res Bull 1991;27:93 – 6. [22] Werth E, Achermann P, Dijk DJ, Borbely AA. Spindle frequency activity in the sleep EEG: individual differences and topographic distribution. Electroencephalogr Clin Neurophysiol 1997;103:535 – 42. [23] Zeitlhofer J, Gruber G, Anderer P, et al. Topographic distribution of sleep spindles in young healthy subjects. J Sleep Res 1997;6:149 – 55. [24] Zygierewicz J, Blinowska KJ, Durka PJ, Szelenberger W, Niemcewicz S, Androsiuk W. High resolution study of sleep spindles. Electroencephalogr Clin Neurophysiol 1999;110:2136 – 47.