The pattern of rapid eye movement (REM) sleep was examined using a new scoring system for electroencephalogram (EEG) stages. An all-night polysomnogram (PSG) was recorded from 12 young healthy volunteers. First, all the data were scored according to
Traditional methods for nonlinear dynamic analysis, such as correlation dimension, Lyapunov exponent, approximate entropy, detrended fluctuation analysis, using a single parameter, cannot fully describe the extremely sophisticated behavior of electro
Electroencephalographic (EEG) activity recorded during the entire sleep cycle reflects various complex processes associated with brain and exhibits a high degree of irregularity through various stages of sleep. The identification of transition from w
This paper presents the contactless measuring of breathing using the MS Kinect depth sensor and compares the results obtained with records of breathing taken by polysomnography (PSG). We explore the methods of signal denoising, resampling, and spectr
Sleep stages are known to differ in the heart rate variability (HRV). REM sleep and wakefulness are characterized by long-range correlations in the heart beat rate. In SWS, a statistical correlation extends only to very few (3–6) of the heart beats t
We developed a simple computer-based, sleep scoring algorithm that categorizes three vigilance states of rats and mice as wakefulness, rapid eye movement (REM) sleep and non-rapid eye movement (NREM) sleep, based on fast Fourier transform analyses of
Variation in outcomes from surgery is a major challenge and defining surgical findings may help set benchmarks, which currently do not exist in laparoscopic cholecystectomy. This study outlines a new surgical scoring system incorporating key operativ
We performed this study to develop a new scoring system to stratify different levels of risk in patients admitted to hospital with a diagnosis of unstable angina (UA), which is a complex syndrome that encompasses different outcomes. Many prognostic v
Classically, professional assessment of sleep is done in the sleep laboratory using whole-night polysomnography (PSG). However, given a misbalance between accredited sleep laboratories and the large amount of patients suffering from sleep disorders,
We developed a machine learning methodology for automatic sleep stage scoring. Our time-frequency analysis-based feature extraction is fine-tuned to capture sleep stage-specific signal features as described in the American Academy of Sleep Medicine m
Sleep and Biological Rhythms 2013; 11: 245–253
Detection of electroencephalogram features of sleep stage 2 by a new scoring system Mitsuo HAYASHI,1 Aoi FUSHIMI1 and Hisashi IIZUKA2 1 Behavioral Sciences, Hiroshima University, Higashi-hiroshima and 2Higashifuji Technical Center, Toyota Motor Corporation, Susono, Japan
Abstract The aim of this study was to clarify the electroencephalogram (EEG) characteristics of sleep stage 2 (S2) by using a new scoring system of 5s-EEG stages. Twelve healthy university students (nine females and three males, 20 to 24 years) with no nocturnal sleep complaints took a nap for 1 h from 14:00. EEG during S2 from the onset of the ﬁrst spindle to sleep stage 3 was manually scored into 5-s epochs of ﬂattening, theta, spindle, K-complex, fast delta, and slow delta stages, according to criteria for EEG stages used for hypnagogic period and REM sleep. EEG stages of ﬂattening, theta, and fast delta occupied 97% of S2. Cumulative epochs of EEG stages of theta and fast delta signiﬁcantly increased as time elapsed from S2 onset. These results suggest that EEG components of ﬂattening, theta, and fast delta constitute the background of S2, and that the quality of S2 may be accounted for by cumulative epochs of theta or fast delta. Further studies are required to examine whether the EEG features observed during a nap could be applicable to nocturnal sleep. Key words: delta wave, electroencephalogram stage, K-complex, sleep stage 2, spindle, theta wave.
INTRODUCTION The importance of slow wave sleep (SWS) and rapid eye movement (REM) sleep has been emphasized in many sleep studies.1,2 The roles of salient electroencephalogram (EEG) components such as the delta wave,3 spindle,4 and K-complex5 have also been explored. However, sleep stage 2 (S2) has not attracted comparable attention, although it constitutes approximately half of nocturnal sleep.6 Laboratory and epidemiological studies have shown that sleep loss is associated with health problems and brain functioning.7 Sleep restriction shortens S2 and REM sleep but not always SWS.8 However, REM sleep depends on circadian body temperature rhythm rather than sleep length,9 and S2 and total sleep time decrease Correspondence: Professor Mitsuo Hayashi, Behavioral Sciences, Hiroshima University, Higashi-hiroshima 739-8521, Japan. Email: [email protected] Accepted 13 August 2013.
proportionately during chronic sleep restriction,8 suggesting that sleepiness, performance deterioration, and health problems caused by shortened nocturnal sleep are substantially caused by the reduction of S2. It has been found that a short nap of 10 to 15 min improves performance and reduces sleepiness.10–13 These naps consist of sleep stage 1 (S1) and S2 and do not contain SWS and REM sleep. Hayashi et al.11 indicated that S2 plays an important role in the restorative function of such a nap, and these restorative effects are limited in S1. It has also been reported that S2 is involved in the consolidation of memory,14,15 and that subjective sleep quality in elderly people who have less SWS is positively related to the length of S2.16 These studies suggest that S2, as well as SWS and REM sleep, plays an important functional role in sleep. A possible reason why S2 has received so little attention may be its ambiguous characteristics. S2 is defined by the occurrence of sleep spindles or K-complexes.17 However, an epoch without such EEG markers is also scored as S2, not S1, if a spindle or K-complex occurs
within 3 min before or after the epoch.17 In addition, epochs that are less than 20% occupied by delta waves are also scored as S2.17 Thus S2 often contains EEG characteristics of S1, S2, and SWS that are present in several EEG components, such as low voltage and irregular waves, theta waves, spindles, K-complexes, and delta waves. Therefore, it is necessary to clarify the detailed EEG characteristics of S2 in order to explore its function. Recently, Kuriyama et al.18 found that EEG theta activity in S2 during a short daytime nap was significantly correlated with subjective fatigue, and the activity of the first spindle correlated with the performance of a vigilance task. Their results suggest that EEG activities may reflect the quality of S2. Hori and colleagues19 measured EEG characteristics of the hypnagogic period at 5-s intervals and categorized this period into nine stages. These EEG stages corresponded to the behavioral and subjective measures; different stages had different effects on the response speed, perception of having slept. Later, their group analyzed the EEG characteristics of REM sleep at 5 s intervals and classified it into six EEG stages.20 This 5-s-EEG scoring system may be useful in evaluating the quality of S2. The aim of the present study is to clarify the EEG characteristic of S2 by using a new scoring system of 5s-EEG stages. We focused on S2 during the first nonREM (NREM) sleep cycle from sleep onset to first SWS because EEG features, which characterize S1, S2, and SWS coexist in this period.
METHODS Participants Twelve undergraduate and graduate students participated in the study (nine females and three males, 20 to 24 years). They previously attended to other sleep experiments and were used to polysomnogram. They had regular sleep-wake habits and had no complaints of nocturnal sleep. None had a habit of napping. Before participation, they were told about the aim and the details of the study and signed informed consent forms. This study was approved by the local research ethics committee of the University. Because sleep stage 3 (S3) did not appear for one male participant, the data of the other 11 participants (21 to 24 years, mean 22.0 years) were analyzed.
Polysomnogram Electroencephalograms (C3-A2, C4-A1, O1-A2 and O2-A1), horizontal electrooculography (EOG),
submental electromyography (EMG), chest electrocardiography (ECG), and respiration curve by nasal thermometer were recorded by a digital electroencephalograph (Polymate AP1532, TEAC, Japan). Time constants were set to 0.3 s, 3.0 s, 0.03 s, 1.0 s, and 1.0 s for EEG, EOG, EMG, ECG, and respiration, respectively. The high cut filter was set to 120 Hz. Impedance between the electrodes was below 5 kΩ.
Procedure The participants reported to the sleep laboratory at 12:00 and were given lunch. Then electrodes for polysomnogram (PSG) were attached. At 14:00, they lay in a reclining seat or in a bed in a darkened, airconditioned, and sound-attenuated isolation box, and were instructed to try to sleep. They were awakened at 15:00. It has been observed that nocturnal sleep becomes lighter and SWS decreases when sleeping in a seat.21 To collect data on the longer S2 time, a nap was taken in a reclining seat by nine participants. The seat was positioned at backrest angles of 150° between the seat bottom and the seat back, which has been shown to be an effective angle for the recuperative effects of napping.12 The other three participants took a nap in a bed to compare the time course of S2 with the nap taken in the reclining seat.
EEG scoring Sleep stages of the naps were manually scored every 20 s using C3-EEG according to the standard criteria.17 Data from one person who took a nap in the reclining seat were not used because SWS did not appear. Differences in naps taken in the reclining seat and those taken in the bed were unclear because of large individual differences in the time courses of sleep stages. Therefore the data of the eleven participants were analyzed together. Electroencephalogram recordings of S2 from the first spindle until S3 were divided into 5-s epochs and were visually scored, using the definitions of Hori et al.19,22 and Takahara et al.20 Unlike EEG stages of the hypnagogic period19 and REM sleep,20 EEG features which characterized S1, S2, and S3 often coexisted in the same epochs, so that it was difficult to classify the main EEG components of such epochs. Therefore the EEG stages of S2 were defined as epochs in which the following EEG wave forms were present for more than 50% of the 5-s duration. 1 Flattening: suppressed waves with amplitudes of less than 20 μV.
2 Theta: theta waves with frequencies of 4 to 7 Hz and amplitudes of more than 20 μV. 3 Spindle: spindles with a duration of more than 0.5 s or composed of more than six consecutive waves. 4 K-complex: the appearance of delta waves with a frequency less than 4 Hz followed by a positive peak of an amplitude such that the difference from the negative peak was greater than 200 μV. 5 Fast delta: delta waves of more than 2 to less than 4 Hz with amplitudes of more than 20 μV. 6 Slow delta: delta waves of 0.5 to 2 Hz with amplitudes of more than 20 μV. 7 Mixed epochs contained no more than 50% of any of the EEG waves listed above.
Data analysis To clarify the time characteristics of each EEG stage, the following parameters were analyzed. (1) Length and frequency of the EEG stages: Percentage of that each EEG stage occupied and length of each EEG stage was calculated. (2) Time course of EEG stages: Number of epochs in which the different stages occurred at each one minute was calculated. (3) Cumulative epochs of EEG stages: Fushimi et al.13 examined the effects of the length of S2 on sleepiness and performance after napping. Their participants were awakened from a daytime nap immediately after 3, 6, or 9 min of S2 elapsed; the longer S2 was, the more effective the nap. These results suggest that cumulative epochs of EEG stages of S2 may be related to napping effects. Therefore the cumulative number of sleep stages after 3, 6 and 9 min from S2 onset was also calculated, and a one-way analysis of variance (ANOVA) with repeated measures was conducted. Degrees of freedom were adjusted to reduce type I error using Huynh and Feldt’s epsilon for small samples. Post-hoc comparisons were performed using Bonferroni’s correction. The significance level was set to 0.05. To examine the relationship between the time courses of each EEG stage, Pearson’s product moment correlations were calculated. Correlation coefficients were calculated for every participant and transformed into ‘z’ scores. These scores were averaged across participants and then retransformed into correlation coefficients.
RESULTS Sleep variables of the nap During the nap, the mean latency to each sleep stage was 3.3 min to S1 (SD = 3.6; range 2.3 to 13.3 for
each participant), 8.2 min to S2 (SD = 5.3; 3.0 to 20.3), and 34.1 min to SWS (SD = 12.2; 16.7 to 51.0). REM sleep did not occur. Mean length of S2 was 18.0 min (SD = 5.2; 10.5 to 27.9).
Example of EEG stages Figure 1 shows an example of 20-s polysomnogram during S2. Although spindle, theta, or delta waves appeared in the first and second 5-s epochs, these epochs were scored as “theta” stages because theta waves occupied more than 50% of each epoch. The third 5-s epoch was scored as a “fast delta” stage because delta waves of 2 to 4 Hz and more than 20 μV occupied more than 50% of the epoch. Although theta waves and spindles appeared in the last 5-s epoch, this epoch was scored as a “flattening” stage. In this epoch, theta waves and spindles occupied less than 50% of the epoch; in contrast, flattening EEG waves with amplitudes of less than 20 μV occupied more than 50% of the epoch.
Length and frequency of EEG stages During S2, the following lengths of EEG stages were found; flattening: 7.3 min (SD = 2.6), theta: 7.8 (SD = 3.5), spindle: 0.5 (SD = 1.2), K-complex: 0.02 (SD = 0.03), small delta: 2.3 (SD = 1.5), and slow delta: 0.05 (SD = 0.08). Table 1 shows the percentage of EEG stages during total S2 time. Both flattening and theta stages occupied approximately 40% of the total S2 time and were significantly longer than other EEG stages (F [6, 60] = 62.88, ∈ = 0.28, P < 0.001). The fast delta stage appeared for 12.6% of S2. These three stages occupied almost all of the 5-s epochs during S2 (96.7%, SD = 4.9; 82.8–99.7% for each participant). Therefore, these three EEG stages were subsequently analyzed. Figure 2 shows the distribution of lengths for flattening, theta, and fast delta EEG stages. Flattening was relatively stable, while fast delta was unstable. Although approximately half of flattening (47.1%) stages lasted for only 5 s (one epoch), a quarter (24.5%) lasted for more than 20 s. In contrast, 73.9% of fast delta stages lasted for 5 s and 91.1% lasted for less than 10 s. Half (50.0%) of the theta stages lasted for 5 s but only 14.2% for more than 20 s.
Time course of EEG stages To clarify the time course of each EEG stage, the numbers of epochs of flattening, theta, and fast delta stages in each minute were calculated, as shown in
Figure 1 Example of polysomnogram. Single, double, and bold lines indicate theta waves, spindles, and fast delta waves, respectively. Table 1 Percentage of electroencephalogram (EEG) stages during total time of sleep stage 2 EEG stage
Flattening Theta Spindle K-complex Fast delta Slow delta Mixed
Figure 3. Because S3 appeared for one participant 11 min after S2 onset, data from the first 10 min were analyzed. The number of epochs in the flattening stage monotonically decreased (F [9, 90] = 6.34, ∈ = 0.82, P < 0.001). Post-hoc comparisons showed that the 9th and 10th minutes of this stage had significantly fewer epochs than the first 5 min (ps < 0.05). The number of epochs in the theta stage significantly increased (F [9, 90] = 2.34, ∈ = 1.00, P < 0.05) and the fast delta stage tended to increase (F [9, 90] = 2.36, ∈ = 0.30, P = 0.09) as a function of time. However, no specific pairwise comparison of minutes was significant. Number of epochs at each one minute in the flattening stage was negatively correlated with that in the theta (r = −0.86, P < 0.0001) and fast delta (r = −0.65, P < 0.05) stages. Theta and fast delta were not significantly correlated.
Figure 2 Distribution of length for each electroencephalogram (EEG) stage. Vertical bars represent standard errors (SEs).
Figure 3 Number of epochs in each minute for electroencephalogram (EEG) stages of flattening, theta, and fast delta. Vertical bars represent standard errors (SEs).
Figure 4 Number of epochs of flattening, theta, and fast delta stages accumulated from the onset of sleep stage 2. Vertical bars represent standard errors (SEs).
Cumulative epochs of EEG stages Figure 4 shows the number of cumulative epochs of flattening, theta, and fast delta stages. These monotonically increased as a function of elapsed time from the onset of S2 (flattening: F [9, 90] = 85.61, ∈ = 0.15, P < 0.001; theta: F [9, 90] = 75.46, ∈ = 0.15, P < 0.001; fast delta: F [9, 90] = 9.37, ∈ = 0.15, P < 0.01). Flattening and theta significantly increased from the 3rd and 4th minute after S2 onset, respectively (ps < 0.05), while fast delta increased from the 8th minute (ps < 0.05). Figures 5 and 6 show the cumulative epochs of EEG stages of theta and fast delta for each participant. The theta stage monotonically increased as a function of elapsed time from the onset of S2 for all participants except for one participant, whose theta stage did not appear during the first 5 min. In contrast, the cumulative epochs of fast delta increased during the latter half of S2.
Figure 5 Number of epochs of theta stage for each participant accumulated from the onset of sleep stage 2. Open circles are individual data, and closed circles are mean data.
Figure 6 Number of epochs of fast delta stage accumulated from the onset of sleep stage 2. Open circles are individual data, and closed circles are mean data.
Table 2 shows cumulative epochs after 3, 6, and 9 min from S2 onset for flattening, theta, and fast delta EEG stages. The numbers of epochs for flattening and theta stages were significantly different after 3, 6 and 9 min of S2 (ps < 0.05). The number of epochs was significantly higher after 9 min than after 3 and 6 min (ps < 0.05), and there was no significant difference between 3 and 6 min for fast delta stage.
EEG components of spindle, K-complex, and slow delta wave Although the lengths of spindle, K-complex, and slow delta EEG stages were short (Table 1), these EEG components often appeared in S2. More than half of the EEG stages (54.6%) contained spindles (44.8%, SD = 14.9) and slow delta waves (9.8%, SD = 2.7). K-complexes appeared in only 1.7% (SD = 1.1) of the stages. The details about these EEG components will be reported elsewhere.
Table 2 Cumulative number of epochs after 3, 6, and 9 min from sleep stage 2 (S2) onset Elapsed time from sleep stage 2 onset (min)
EEG stage Flattening Theta Fast delta
(d.f. = 2.20)
23.0 (5.3) 12.0 (5.1) 0.6 (1.1)
44.2 (12.4) 24.3 (12.1) 2.1 (2.0)
57.5 (17.3) 41.1 (14.6) 6.2 (5.5)
64.05 70.43 10.17
0.68 0.68 0.54
<0.001 <0.001 <0.01
Parentheses indicate SDs. Underlined mean values are not significantly different. ANOVAs, analyses of variance; d.f., degrees of freedom; EEG, electroencephalogram.
DISCUSSION The present study classified EEG of S2 during the first NREM cycle from the onset of S2 until the first S3 into 5-s epochs in a procedure similar to that previously used with the hypnagogic period19 and with REM sleep.20 The EEG stages of flattening, theta, and fast delta occupied 97% of S2, and stages containing spindles and slow delta waves occupied 55%. These results suggest that S2 is characterized by flattening, theta, and fast delta waves as the background EEG, with the sporadic appearance of spindles and slow delta waves.
EEG stages of flattening and theta The EEG stages of flattening and theta occupied 84% of S2, suggesting that these EEG components mainly constitute the background EEG of S2. However, the percentages of these stages were negatively correlated; flattening decreased while theta increased as a function of time elapsed from S2 onset. This result suggests that these EEG stages reflect the depth of sleep. The EEG stage of flattening occupied 41.3% of S2 and decreased as a function of time. Although half of the occurrences of this EEG stage lasted for 5 s, a quarter lasted for more than 20 s. For the hypnagogic period, contrary to the present results, Tanaka et al.23 reported that the flattening stage was the most unstable and shortest and only occupied 5.1% (0.9 min) of 17.6 min from lights off until 5 min elapsed after the first spindle appeared. While this percentage corresponds to 11.0% of 8.2 min of S1, it is still lower than the comparable rate in the present study. In addition, the percentage of the EEG stage of theta was larger in S2 (42.9%) than in the hypnagogic period (10.8%).23 This discrepancy between S2 and the hypnagogic period may arise from the definition of the EEG stages. The hypnagogic EEG stages were defined by EEG components and its amplitudes,19,23 but not by the durations. The present study,
however, adopted a 50% criterion for the duration of the EEG components. As can be seen in Figure 1, various EEG components often coexisted in the same 5-s epochs during S2, so that a clear-cut distinction could not be made without the duration criterion. For the same reason, the present study did not distinguish vertex sharp waves from theta waves, as was done in the hypnagogic EEG stages. According to our 50% duration criterion, the epochs in which theta waves or vertex sharp waves occupied less than 50% were classified as flattening stage, while such epochs were classified as “ripples” (theta waves) or “vertex sharp wave(s)” stages in the hypnagogic period.23 In the present study, epochs in which theta waves or vertex sharp waves occupied more than 50% were scored as EEG theta stage. As a result, the proportion of EEG stages of flattening and theta may have been larger here than in the hypnagogic period. For REM sleep, Takahara et al.20 reported that flattening and theta stages occupied 34.9% and 57.8%, respectively. These percentages were comparable to those for S2 in the present study (41.3% and 42.9%). However, Takahara et al. also applied the same criteria used in the hypnagogic period study,19,23 and did not refer to the duration of the EEG components. Therefore, unlike the present study, the epochs with theta waves in REM sleep were scored as EEG theta stage even if these waves occupied less than 50% of the epochs. Thus, the flattening stage may have been more modestly scored in REM sleep.20 Nevertheless, one third of the 5-s EEG epochs during REM sleep were scored as flattening, suggesting that the low-voltage EEG pattern is more prominent in REM sleep than in S2.
EEG stages of spindles, K-complexes, slow and fast delta waves Spindles, K-complexes, and delta waves often appeared in S2, unlike the hypnagogic period and REM sleep. The
fast delta stage significantly increased after S2 onset and occupied 12.6% of S2. In addition, fast delta was negatively correlated with flattening (r = −0.65). These results suggest that fast delta is one of the EEG components, which constitute the background EEG of S2. Although spindles appeared in approximately half of the 5-s S2 EEG epochs (44.8%), S2 was occupied by only small percentages of spindle (2.5%), K-complex (0.1%), and slow delta (0.2%) EEG stages. This discrepancy is attributable to the 50% criterion; that is, based on this criterion, the spindle stage was restricted to the epochs that contained more than two spindles or one spindle with a duration of more than 2.5 s. In future studies, it may be preferable to score the spindle stage as any epoch that contains spindles, regardless of the 50% criterion.
SWA is traditionally measured by EEG spectral power within the frequency range of approximately 0.5 to 4.5 Hz and has been considered to be a homeostatic marker of sleep.1 Brooks and Lack and Lamarche et al. also used these frequency bands (0.5–4.0 Hz,10 and 0.75–3.75 Hz25), which contain both slow (0.5–2.0 Hz) and fast (2–4 Hz) delta. However, fast delta stage (12.6%) was greater than slow delta stage (0.2%) during S2, suggesting that the effects of SWS during a short daytime nap may be dependent on fast delta. Taken together, the results of the present study suggest that the quality of S2 may be accounted for by the amount of fast delta or theta. Further study is required to evaluate these relationships.
Time course of EEG structure of S2 Cumulative epochs of EEG stages It has been reported that an ultra-short nap, which is composed of 0.5 or 1.5 min of S1 has no recuperative power.24 Hayashi et al.11 also found that a short daytime nap of 5 min, which was composed of only S1 had limited positive effects, while a 9-min nap composed of 6 min of S1 and 3 min of S2 improved performance of vigilance task and reduced subjective sleepiness. In the first 3 min of S2 in the present study, the number of cumulative epochs of 5s-EEG stages was less than 1 for fast delta but 23.0 (corresponding to 115 s) and 12.0 (60 s) for flattening and theta, respectively. Although flattening EEG often occurs in the early period of S1, S1 has no, or little recuperative power.11,24 These results suggest that theta waves, and not the flattening EEG, may contribute to the restorative power of an ultra-short nap of less than 10 min. The effectiveness of the nap is enhanced by increasing the time of S2. Fushimi et al.13 compared the effects of 3, 6, and 9 min of S2 on subsequent sleepiness and performance after napping, and found that longer S2 resulted in a more effective nap. In the present study, cumulative epochs of theta stage increased, and differed significantly after 3, 6, and 9 min of S2. This change was observed for all participants except one. These results support the possibility that theta activity plays an important role in the restorative function of a short nap. On the other hand, several studies have pointed out that delta activity contributes to the benefits of a short nap. Brooks and Lack10 noted that the onset of some delta activity contributes to the benefits of a short daytime nap. Lamarche et al.25 also found that the amount of slow wave activity (SWA) of the parietal area in women during the late-luteal phase correlated with cognitive performance.
During S1, people can conduct a certain degree of behavioral response to auditory stimuli and the perception being awake is high.19 These phenomena give rise to the question whether S1 is true sleep. Or rather, should it be considered a “sleep onset period”.26 Contrary to S1, behavioral response decreases and the perception of being awake is low during S219, supporting the consensus that S2 is true sleep. It could be argued that true sleep begins with the appearance of spindles.26 Using spectral analysis of EEG from waking to SWS during nocturnal sleep, it has been observed that sigma band activity, which corresponds to the frequency band of spindles, increases 0.4 min after S2 onset.27 In the present study, however, within 5 min after S2 onset, flattening EEG predominantly appeared. After that, the number of epochs in flattening stage monotonically decreased, and those in theta and fast delta alternatively increased. Cumulative number of epochs in EEG stage of fast delta also increased after 9 min from S2 onset, compared to that of 6 min. These results fundamentally correspond to the findings of spectral analysis of EEG during nocturnal sleep, in which delta (2.5–3.5 Hz) and theta (4.0– 7.5 Hz) band EEG activities begins to increase from several minutes before S2 onset, and reached a plateau level after several minutes of S2 onset.28 These results suggest that S2 may be classified into three major periods, by the predominant EEG components: flattening, theta, and delta. The present study examined the time course of EEG structure from S2 onset, and could clarify the behavior of flattening and theta activities as the background EEG of S2, which predominantly appear within 10 min after S2 onset. However, delta activities that are thought to predominantly appear around the termination of S2
could not be fully examined in this study. It is suggested that the time structure until S2 offset should be further explored in order to clarify the dynamics of delta activity during S2.
Limitations There are several limitations to the present study. First, the small sample size may emphasize individual differences. More participants are required to confirm the present findings. Second, the participants in the present study were young adults. In general, younger age is associated with larger EEG amplitudes, and EEG amplitude differs among different age groups.29 Further studies are required to examine age effects on the EEG structure of S2. Third, the present study analyzed S2 during the first NREM cycle of a daytime nap, not during the latter half of nocturnal sleep when S2 lasts longer. Although it has been reported that the occurrence of slow wave sleep during a nap is reproducible in nocturnal sleep,30 it is not clear whether all EEG features of S2 during the nap are applicable to nocturnal sleep. In addition, Brandenberger et al.31 pointed out that S2 is not uniform; it contains a “quiet” period toward SWS and “active” period toward REM sleep. The present study only focused on the former period. Nevertheless, the EEG stages of flattening and theta occurred stably, similar to REM sleep, suggesting that the present EEG scoring system may be applicable to S2 toward REM sleep. Fourth, the present study used only visual scoring. Quantitative EEG analysis should advance the understanding of the quality of S2. Finally, the present study did not clarify the features of spindle which appeared in approximately half (44.8%) of the EEG stages during S2. It is reported that the occurrence of slow and fast spindles are affected by the phase of slow oscillations32 or K-complexes.33 The relationships among the EEG components during S2 such as spindles, K-complexes, theta, slow and fast delta waves should be further explored.
REFERENCES 1 Achermann P, Borbély AA. Sleep homeostasis and models of sleep regulation. In: Kryger MH, Roth T, Dement W, eds. Principles and Practice of Sleep Medicine, 5th edn. Elsevier Saunders: St. Louis, MO, 2011; 431–44. 2 Siegel JM. REM sleep. In: Kryger MH, Roth T, Dement W, eds. Principles and Practice of Sleep Medicine, 5th edn. Elsevier Saunders: St. Louis, MO, 2011; 92–111.
3 Amzica F, Steriade M. Electrophysiological correlates of sleep delta waves. Electroencephalogr. Clin. Neurophysiol. 1998; 107: 69–83. 4 De Gennaro L, Ferrara M. Sleep spindles: an overview. Sleep Med. Rev. 2003; 7: 423–40. 5 Colrain IM. The K-complex: a 7-decade history. Sleep 2005; 28: 255–73. 6 Williams RL, Karacan I, Hursch CJ. Electroencephalography (EEG) of Human Sleep: Clinical Applications. Wiley: New York, 1974. 7 Grandner MA, Patel NP, Gehrman PR, Perlis ML, Pack AL. Problems associated with short sleep: bridging the gap between laboratory and epidemiological studies. Sleep Med. Rev. 2010; 14: 239–47. 8 Banks S, Van Dongen HPA, Maislin G, Dinges DF. Neurobehavioral dynamics following chronic sleep restriction: dose-response effects of one night recovery. Sleep 2010; 33: 1013–26. 9 Czeisler CA, Zimmerman JC, Ronda JM, Moore-Ede MC, Weitzman ED. Timing of REM sleep is coupled to the circadian rhythm of body temperature in man. Sleep 1980; 2: 329–46. 10 Brooks A, Lack L. A brief afternoon nap following nocturnal sleep restriction: which nap duration is most recuperative? Sleep 2006; 29: 831–40. 11 Hayashi M, Motoyoshi N, Hori T. Recuperative power of a short daytime nap with or without stage 2 sleep. Sleep 2005; 28: 829–36. 12 Hayashi M, Abe A. Short daytime nap in a car seat to counteract daytime sleepiness: the effect of backrest angle. Sleep Biol. Rhythms 2008; 6: 34–41. 13 Fushimi A, Hayashi M. The examination of the best duration of nap time taken in a car seat. The 33rd Annual Meeting of Japanese Society of Sleep Research, Abstract book 2008; 116. 14 Walker MP, Brakefield T, Morgan A, Hobson JA, Stickgold R. Practice with sleep makes perfect: sleepdependent motor skill learning. Neuron 2002; 35: 205– 11. 15 Fogel S, Smith C. Learning-dependent changes in sleep spindles and stage 2 sleep. J. Sleep Res. 2006; 15: 250–5. 16 O’Donnell D, Silva EJ, Munch M, Ronda JM, Wang W, Duffy JF. Comparison of subjective and objective assessments of sleep in healthy older subjects without sleep complaints. J. Sleep Res. 2009; 18: 254–63. 17 Rechtschaffen A, Kales A. A Manual of Standardized Terminology, Techniques and Scoring System for Sleep Stages of Human Subjects. Public Health Service, U.S. Government Printing Office: Washington, DC, 1968. 18 Kuriyama A, Ogawa K, Hayashi M. EEG activity underlining the recuperative power of a short nap. The 38th Annual Meeting of Japanese Society of Sleep Research, Abstract Book 2013; 175. 19 Hori T, Hayashi M, Morikawa T. Topographical EEG changes and the hypnagogic experience. In: Ogilvie RD,
Harsh JR, eds. Sleep Onset: Normal and Abnormal Processes. American Psychological Association: Washington, DC, 1994; 237–53. Takahara M, Kanayama S, Nittono H, Hori T. REM sleep EEG pattern: examination by a new EEG scoring system for REM sleep period. Sleep Biol. Rhythms 2006; 4: 105–10. Aeschbach D, Cajochen C, Tobler I, Dijk DJ, Borbély AA. Sleep in a sitting position: effect of triazolam on sleep stages and EEG power spectra. Psychopharmacology (Berl) 1994; 114: 209–14. Sleep Computing Committee, Hori T. chair. Learning Manual of PSG Chart. Polysomnogram, Sleep Stage Scoring, Interpretation. Japanese Society of Sleep Research: Tokyo, 1999. Tanaka H, Hayashi M, Hori T. Statistical features of hypnagogic EEG measured by a new scoring system. Sleep 1996; 19: 731–8. Tietzel AJ, Lack LC. The recuperative value of brief and ultra-brief naps on alertness and cognitive performance. J. Sleep Res. 2002; 11: 213–18. Lamarche LJ, Driver HS, Forest G, Koninck JDE. Napping during the late-luteal phase improves sleepiness, alertness, mood and cognitive performance in women with and without premenstrual symptoms. Sleep Biol. Rhythms 2010; 8: 151–9. Ogilvie RD. The process of falling asleep. Sleep Med. Rev. 2001; 5: 247–70.
27 Morikawa T, Hayashi M, Hori T. Spatiotemporal variations of alpha and sigma band EEG in the wakingsleeping transition period. Percept. Motor. Skill 2002; 95: 131–54. 28 Morikawa T, Hayashi M, Hori T. Auto power and coherence analysis of delta-theta band EEG during the waking-sleeping transition period. Electroencephalogr. Clin. Neurophysiol. 1997; 103: 633–41. 29 Carrier J, Land S, Buysse DJ, Kupfer DJ, Monk TH. The effects of age and gender on sleep EEG power spectral density in the middle years of life (ages 20–60 years old). Psychophysiology 2001; 38: 342–42. 30 Gander P, Signal L, Van Dongen HPA, Muller D, Van den Berg M. Stable inter-individual differences in slow-wave sleep during nocturnal sleep and naps. Sleep Biol. Rhythms 2010; 8: 239–44. 31 Brandenberger G, Ehrhart J, Buchheit M. Sleep stage 2: an electroencephalographic, automatic, and hormonal duality. Sleep 2005; 28: 1535–40. 32 Mölle M, Bergmann TO, Marshall L, Born J. Fast and slow spindles during sleep slow oscillation: disparate coalescence and engagement in memory processing. Sleep 2011; 34: 1411–21. 33 Kokkinos V, Kostopoulos K. Human non-rapid eye movement stage II sleep spindles are blocked upon spontaneous K-complex coincidence and resume as higher frequency spindles afterwards. J. Sleep Res. 2011; 20: 57–72.