REM Sleep Episodes Research Articles

Melatonin secretion in relation to sleep in epileptic and control children and adolescents.

Serum melatonin (MT) levels were measured by radioimmunoassay at 4 hour intervals from 12.00 am to 8.00 am (darkness 8.00 pm - 7.00 am) with simultaneous EEG tracing in 22 epileptic patients aged 8 to 21 years. Mean levels for all patients at 12.00 pm and 4.00 am were respectively 43±7 and 55±10 pg/ml (mean ± SEM). There was no significant difference in mean peak levels of MT at 4.00 am among non-treated (mean = 62±19 pg/ml, n = 6), valproic acid-treated (mean = 48±13 pg/ml, n = 10) and patients treated with other medications (mean 62±29 pg/ml, n = 3). Three treated patients who had epileptic discharges through the night had a mean peak of only 32±5 pg/ml and were not included in the subsequent analysis. Analysis of MT levels according to sleep stages revealed that MT values after 2 or more episodes of REM sleep (86±20 pg/ml) were significantly higher than those after no preceeding episodes (31±6 pg/ml, P<.0001) and even than those after one preceeding episode (48±8 pg/ml, P<.05). For comparison 8 normal males were studied hourly in the same fashion. Mean MT levels at 12.00 am and 4.00 am were 36±10 pg/ml and 50±9 pg/ml, respectively. Likewise, MT values at 4.00 am varied directly with the number of preceeding episodes of REM sleep although the differences were not statistically significant. These results suggest that MT secretion may be related to REM sleep.

REM sleep episodes during the multiple sleep latency test in narcoleptic patients

Forty narcoleptic patients were given the Multiple Sleep Latency Test, consisting of 20 min opportunities to sleep offered at 10.00, 12.00, 14.00, 16.00 and 18.00 o'clock. Eleven patients had 2 episodes of REM sleep, 5 had 3, 11 had 4, and 13 had 5 before they were awakened. Fourteen control subjects given similar opportunities to sleep (reported in a companion article (Richardson et al. 1978)) had no REM sleep episodes. For the 10.00–18.00 o'clock opportunities respectively, there were 32, 29, 30, 28 and 27 REM sleep episodes. We conclude that this procedure can provide physicians with data useful in the diagnosis of narcolepsy.

Sleep-deprivation: Effects on sleep and EEG in the rat

1. The vigilance states (waking, rapid eye movement (REM) sleep, and non-REM (NREM) sleep), motor activity, food intake and water intake were continuously recorded by telemetry in unrestrained rats. In addition, an amplitude measure and a frequency measure (number of zero-crossings (ZCR) per 10 s) of the telemetered EEG-signal was obtained. The animals were recorded during a control day, then subjected to 12-h or 24-h sleep-deprivation (SD) by means of a slowly rotating cylinder, and subsequently recorded for further 1–2 days. The EEG-parameters were recorded also during SD. 2. On the control day, the EEG-amplitude of NREM-sleep exhibited a decreasing trend in the 12-h light-phase (Figs. 3, 4). The occurence of slow wave sleep (SWS; defined as the NREM-sleep fraction with less than 40 ZCR/10 s) was practically limited to the first part of the light-phase (Figs. 2, 4). Cumulative plots of the zero-crossing bands (Fig. 2) revealed a prominent daily rhythm in the EEG-frequency distributionwithin NREM-sleep. 3. The percentage of NREM-sleep and REM-sleep was little affected by the 12-h SD, but the amount of SWS and the EEG-amplitude of NREM-sleep were increased (Figs. 4, 6). After a 24-h SD period terminating before light-onset, NREM-sleep was reduced and REM-sleep was markedly enhanced (Figs. 4, 6; Table 1). Both the duration and frequency of REM-sleep episodes were increased, and episodes of total sleep prolonged (Table 2). The amount of SWS was significantly more increased after 24-h SD than after 12-h SD, whereas the EEG-amplitude of NREM-sleep was enhanced to a similar extent after both SD-schedules (Tables 1, 3 Fig. 6). 4. After a 24-h SD period terminating before dark-onset, sleep (particularly REM-sleep) was enhanced in the first hours of the dark-phase, yet the usual high activity bouts prevailed in the later part of the dark-phase (Figs. 7, 8; Table 1). The extent and time-course of REM-sleep rebound was similar after the two 24-SD schedules, whereas SWS-rebound was different: SWS exhibited a one-stage rebound when recovery started in the light-phase, and a two-stage rebound when recovery started in the dark-phase (Fig. 9). 5. A comparison of the effects of 12-h SD performed with the usual and with the double cylinder rotation rate, showed only small differences, indicating that forced locomotion was a minor factor in comparison to sleep-deprivation (Fig. 10; Table 1). 6. The daily pattern of SWS on control days, and the marked increase of SWS after SD correspond to the results from other animal and human studies. It is proposed that due to the existence of an intensity dimension, NREM-sleep is finely regulated around its baseline level, and thus may be readily and accurately adjusted to current ‘needs’, whereas REM-sleep, lacking an apparent intensity gradient, is regulated around a level which is considerably below baseline. Thus, in contrast to NREM-sleep, REM-sleep compensation can occur only by an increase in the time devoted to this state, thereby curtailing the time available for other activities.

Morphine self-administration and EEG power spectra in the rat

Power spectral analyses were used to study changes in cortical EEG during morphine self-administration in freely-moving dependent rats prepared with chronic cortical and muscle electrodes and with permanent indwelling IV cannulae. As time progressed from a morphine self-injection toward another injection, a significant spectral shift of the EEG to lower frequencies occurred during successive REM sleep episodes. Each morphine self-injection reinstated the predominance of higher frequencies in the EEG spectra. These EEG changes which preceded lever pressing may reflect changes in morphine plasma levels and in the state of the CNS that precede drug-seeking behavior.

Effects of the acute administration of ethanol on the sleep of the rat: A dose-response study

Seven-hr sleep recordings were performed on rats following intraperitoneal injection of saline or one of four doses of ethanol (1.1, 1.5, 2.0 or 2.5 g/kg). Total minutes of REM sleep and percentage REM sleep were decreased in a dose-dependent manner. Percentage nonREM sleep increased with progressively higher doses. The decrease in REM sleep appeared to be related to a decrease in the number of REM sleep episodes and an increase in the length of the REM-nonREM cycle. Other variables such as mean length of REM sleep episodes and REM sleep efficiency were unchanged. An analysis of the first and second 3.5 hr of the recording showed that ethanol continued to have marked effects on REM and nonREM sleep during the second 3.5 hr, when blood levels were declining. Ethanol produced decreases in sleep latency, but total sleep time was unchanged.

Baseline sleep-wake patterns in the pointer dog

The 24-hr electrographic patterns (EEG, EOG, EMG) of six normal pointer dogs were recorded in a laboratory setting. Two states of sleep (slow-wave and rapid eye movement) and wakefulness (alert and drowsy) were identified. The total recording period comprised 44% of alert wakefulness, 21% of the drowsy state, while slow-wave sleep occupied 23% and REM sleep 12% of the time. The mean length of a REM sleep episode averaged 6 min and the mean REM sleep cycle was 20 min. The mean polycyclic sleep-wake cycle was 83 min. Sleep episodes averaged 45 min and the mean waking episode was 38 min. There was an average of two REM sleep episodes per sleep-wake cycle. The dog has a propensity to sleep over a 16-hr interval from 1300 to 0500 but the most sleep occurred between 2100 and 0400 hr during darkness.

Characteristics of the experimental reflex sleep induced by vago-aortic nerve stimulation

1. 1. Electrical stimulation of the baroceptive fibres of the vago-aortic trunks (VA St) reflexly induces, in the Encéphale isolé (EI) cat preparation, a complete sleep cycle characterized by a progression of all stages of sleep previously described in chronic preparations (slow wave sleep, phasic slow wave sleep, paradoxical or REM sleep). Some qualitative and quantitative characteristics of this “vagoaortic” (VA) sleep are analysed and compared with those obtained in EI cats previously deprived of REM sleep. 2. 2. VA sleep is not simply the consequence of chance occurrences, but the consequence of VA St. It is stimulus-bound, even if the interval between successive stimuli is as short as 7.5 min. Thus it is possible to impose an artificially regular cycle much shorter than the normal one. 3. 3. VA St. precipitates the onset of the different sleep cycle episodes in the non-REM-deprived EI cat as well as in the REM-deprived EI cat. 4. 4. REM sleep begins rapidly after stimulus onset, so that an experimental reflex REM narcolepsy may be obtained. This effect is particularly noteworthy in REM-deprived cats. In this situation, REM sleep episodes with cataplexy (muscular atonia and areflexia) can immediately follow the onset of stimulation.

Sleep satiation in narcoleptic patients

Polygraphic sleep recording was performed in 20 narcoleptics with one or more of the auxiliary symptoms, 4 narcoleptics with only sleep attacks and 10 normals during one night and into the following day. Total sleep time in the narcoleptics did not differ significantly from that in the normals. Sleep of the narcoleptics with auxiliary symptoms was unstable with frequeny awakening. The temporal organization of the REM—NREM sleep cycle was irregular in the narcoleptics with auxiliary symptoms, compared with those in the other two groups. Percentage of Stage 1 was significantly larger in the narcoleptics with auxiliary symptoms than in the other two groups and percentage of Stage 2 was smaller. Percentage of Stage 4 was smaller in the narcoleptics with auxiliary symptoms than in those with only sleep attacks. Percentages of Stages 3 and REM did not differentiate the three groups. Spindle density during Stage 2 did not differentiate the three groups. Sleep onset Stage REM was frequently observed exclusively in the narcoleptics with auxiliary symptoms. Excluding the instances showing sleep onset Stage REM, mean latency of initial episodes of REM sleep in the three groups was shorter after daytime sleep onset than after nighttime sleep onset. In the narcoleptics with auxiliary symptoms, no significant correlation was found between the percentage of Stage REM and clinical findings. Conclusions: the sleep attacks in narcoleptics are due to an ill-timed, strong tendency to fall asleep (including both REM and NREM sleep), and it is therefore suggested that they are manifestations of their disturbed circadian rhythm of sleep-wakefulness cycle.

Effects on sleep of restricted sleep. A cat case study

One male cat was adapted to different schedules of restricted sleep. The cat was allowed to go to sleep during a certain number of hours per day. During the rest of the 24 h period, wakefulness was enforced by means of a treadmill. The following schedules of restricted sleep were run: 12 h sleep–12 h treadmill (12S–12T), 8S–16T, 4S–20T. The cat was also adapted to a 36 h day: 12S–24T. The sleep was investigated after at least 2 weeks on each schedule and compared to ad lib. sleep (24S-OT). As available sleep time became shorter, the composition of the sleep changed. LSWS (in % of available sleep time) decreased, while DSWS % and REM sleep % increased. The length of the DSWS and REM sleep episodes increased with decreasing sleep time, as did sleep cycle length.

Duration of EEG sleep stages in different types of disturbed night sleep.

Summary The duration of each episode of any one electrophysiological sleep stage or any episode of intervening wakefulness was determined in three kinds of disturbed sleep, namely, in naturally impaired sleep of late middle-aged, normal people, in sleep after caffeine administration, and in sleep after hypnotic drug withdrawal. When compared with the sleep of young people, the sleep of late middle-aged subjects was characterized by an increase in longer episodes of intervening wakefulness plus drowsiness, and by a decrease in longer episodes of sleep stages 3 + 4 and REM. The sleep after caffeine, compared with the baseline sleep, contained an increased proportion of longer episodes of intervening wakefulness plus drowsiness but no significant change in the episode duration of any of the sleep stages. In the case of the drug withdrawal sleep, compared with the sleep on hypnotic drug, there was no change in the episode duration of intervening wakefulness plus drowsiness but there was a significant shortening of episode duration for sleep stages 2 and 3+4, with a similar trend for REM sleep episodes. Thus a sleep disturbance might be characterized by increased stability of wakefulness, by decreased stability of sleep or both. Different types of sleep disturbance might require different types of treatment.

Time trends and periodic cycles in rem sleep eye movements

Eye movements during REM sleep episodes were tabulated in 16 young adults. REM episodes were then broken down into four ranges according to length in min: (1) 11.0–21.3; (2) 21.7–29.7; (3) 30.0–42.3; (4) 42.7 or longer. These data were then analyzed for linear and quadratic trends. Eight episodes had a significant linear trend, 10 had a significant quadratic trend, 7 had both linear and quadratic trends, while 12 had no trend. The residuals from the best-fitting polynomial curve were then subject to a spectral analysis. In addition, 2 long periods of pre-sleep wakefulness (approximately 2 h each) were also analyzed. In general, the spectral analysis revealed the dominant presence of a slow cycle (period of 10 min to about 30 min) the exact period of which varied according to the length of the REM episode. A binomial probability test indicated that the presence of slow cycles was significant in REM episodes except for those in the 21–30 min range. For the episodes of wakefulness, a dominant slow cycle was found in both cases. The results give the impression of similarity in the periodic organization of eye movements during REM sleep and waking. The data also indicated that an ultradian (70–150 min) cycle was present in eye movements during sleep and waking. Further, the finding of a decrease in eye movements before sleep onset, coupled with previous reports of an increase in eye movement after sleep onset, indicate the presence of a circadian cycle.

The number and duration of the episodes of the various EEG stages of sleep in young and older people

Episodes of any one electrophysiological sleep stage and of intervening wakefulness were examined as to their numbers and durations in sleep of ten young normal subjects (mean age 22) and in significantly more disturbed sleep of fourteen late middle age normal people (mean age 55). No significant difference between the two age groups was found for any of the stages in the mean number of episodes taken from 4 nights. For any of the stages, the individual numbers of episodes accumulated over the 4 nights varied in a wide range and showed significant consistency with similar data obtained in another set of 4 nights. There were significant differences between the two age groups in episode duration. The sleep of the older group contained significantly higher proportions of longer episodes of wakefulness and drowsiness and significantly lower proportions of longer episodes of slow wave sleep and REM sleep. Thus, the initiation of episodes of EEG stages was on average similar in the better sleep of young and impaired sleep of older subjects, but the maintaining or termination of the episodes were different.

Differential effects of pentylenetetrazol on REM sleep in naive and cobalt-epileptic rats

Chronic experimental epilepsy was produced in adult female Sprague-Dawley rats by the implantation of cobalt wire into the right parietal cortex. Rats treated similarly with glass rods were used as controls. All animals were simultaneously prepared with permanent cortical and temporalis muscle electrodes for EEG and EMG recording. Administration of pentylenetetrazol i.p. at a dose of 15 mg/kg every 15 min until the appearance of generalized convulsions resulted in the lowering of the chemical seizure threshold expected for the cobalt-treated rats, with only 2 injections of the drug required in contrast to the 3 to 4 injections needed for the control rats. Analysis of the EEG recordings collected over the 24 h period after the first injection revealed the presence of a more pronounced suppression of REM sleep in the cobalt-epileptic rats. This effect was found to be due to a reduction in the total number of REM sleep episodes, while the duration of the individual episodes remained unchanged. The latencies to REM onset in these rats, moreover, were markedly reduced. These results in a chronic seizure model further support the usefulness of cobalt experimental epilepsy in the rat for the study of human seizure disorders in comparison with acute seizure models.

Δ9-Tetrahydrocannabinol and the sleep-wakefulness cycle in rabbits

The effects of 0.5 and of 1.0 mg/kg of intravenously administered Δ 9-tetrahydrocannabinol ( Δ 9-THC) on the sleep-wakefulness cylce were studied in six freely moving rabbits with chronically implanted electrodes. On three consecutive days 5-hr polygraphic recordings were made in each animal; the first record was begun immediately following administration of the drug, the second and third were made 24 and 48 hr later. At both dose levels, Δ 9-THC evoked biphasic effects on the first day; an initial phase of extremely low amplitude fast activity of the alert state was followed by a period of predominantly slow wave sleep. During the initial alert phase spike discharges appeared in the motor cortex and hippocampus and behavioral alterations characterized chiefly by loss of motor coordination were observed. A decrease in the number of REM sleep episodes was mainly responsible for a significant reduction in the total time spent in REM sleep on the first day. On the second day the percent time of REM sleep was still below the control value although the difference was statistically insignificant, but by the third day it had returned to the control level.

Basic rest-activity cycle rhythms in the human auditory evoked response

Two experiments are reported in this study. In the first, the N 2 latency and P 2N 2 amplitude of the vertex-recorded averaged auditory evoked response (AER) to click stimuli were studied in 14 normal children for a total of 23 nights. Changes in amplitude and latency measurements were studied as a function of elapsed time during REM sleep, and during Stage 2 sleep as a function of proximity to an episode of REM sleep. While no significant group changes were noted, it was observed that in many cases there were regular increases or decreases in AER latency or amplitude during episodes of REM sleep or during Stage 2 sleep as a function of proximity to REM sleep. It was hypothesized that regular changes might be occurring in the amplitude and latency characteristics of the AER independent of stage of sleep. The second experiment was designed to test this hypothesis. One normal infant aged 3 months, 8 normal children aged 35–56 months, and 1 normal adult aged 34 years, were studied. Infant and child subjects were continuously studied during 7–9 h of the night and the adult subject was studied for a 24 h period. AERs were continuously recorded throughout each experiment and tagged as to their time in relation to the beginning of the experiment and as to the state of the subject at the time they were recorded. N 2 latency and P 2N 2 amplitude were measured in each AER. The data were ordered as a function of time and regressed on to a series of sine-cosine wave forms whose periods ranged from 10 to 280 min. The percent of the variance in latency or amplitude measurements accounted for by each sine-cosine pair was calculated and graphed. The results showed that the amplitude and latency characteristics of the AER oscillated rhythmically throughout each experiment, the period of the oscillations matching that of the basic rest-activity cycle (BRAC) previously described as characteristic of nervous system function by Kleitman. The period of the BRAC varies from 40–60 min in infants to 90–110 min in adults. F values generated by the regression analyses were used to estimate the probability that cyclical increases and decreases had occurred in the AER measurements. Such increases and decreases in amplitude and latency were found in all 10 cases at a significant level. The “latency cycle” accounted for 17.6–39.3% of the latency variance in each case, while the “amplitude cycle” accounted for 7.2–30.0% of the variance in amplitude measurements.

An analysis of the effect of reserpine upon ponto-geniculo-occipital wave activity in the cat*

The effect of reserpine upon ponto-geniculo-occipital wave activity has been studied, using implanted electrodes, in a series of experiments on unrestrained cats. Five stages of action of reserpine have been defined. Stage 1: Drug induced waves (PGO Res) resembling the waves normally present during REM sleep (PGO REM), appear and gradually increase in frequency, but may be suppressed by arousing stimuli. Stage 2: PGO Res occur with a very regular frequency of 15–25 waves min and cannot be suppressed by arousal. Animals are heavily sedated, but awake. Stage 3: PGO Res become modulated in frequency and amplitude. Both cortical synchrony and a low voltage cervical EMG are associated with enhanced wave activity. Stage 4: PGO Res gradually become inhibited during periods of behavioral arousal. Stage 5: REM sleep returns, but REM-type PGO waves continue to appear during slow wave sleep, several minutes before the onset of REM sleep. Regulation of PGO Res, during stages 3 and 4, can be accounted for by postulating that this activity is inhibited by arousal. Regulation of PGO Res during stage 5 suggests the existence of an additional influence, which is maximal at the end of each REM sleep episode and declines progressively to become minimal in the minutes immediately preceding the next period of REM sleep.

Sleep in the domestic pigeon ( Columba livia)

The EEG, EOG, neck EMG and head movements were recorded and behavioral observations made during undisturbed sleep and wakefulness in three domestic pigeons ( Columba livia). Three states of sleep and wakefulness were distinguished: (a) Wakefulness, characterized by behavioral activity, a low voltage-high frequency EEG, ocular movements and high tonic neck EMG activity. (b) Slow-wave sleep, characterized by behavioral inactivity, high voltage slow EEG activity, and reduced tonic neck activity. (c) REM sleep, characterized by successive downward and rotational jerks of the head, a low voltage-high frequency EEG, clusters of 3-8 eye movements and a marginal but inconsistent decrease in EMG activity. REMs persisted when the pigeon's head was immobilized. Forty-two percent of the 24-hr period was spent asleep. REM sleep episodes had a mean duration of 11 sec and constituted 7% of total sleep time and 3% of the 24-hr period.

Neurology and sleep research.

This brief review documents some of the important contributions of recent sleep research to the understanding of a number of neurological conditions. Narcoleptic attacks have been shown to be either episodes of NREM or REM sleep; and cataplexy, sleep paralysis and vivid hypnagogic hallucinations consist of dissociated or inappropriate REM sleep. Important relationships of the hypersomnias and various comas to sleep mechanisms are being increasingly elucidated. Various types of epileptic seizures have been found to be affected differentially by the two types of sleep and by arousal from them; and sleep deprivation may activate or perpetuate epilepsy. Finally, some miscellaneous conditions, such as dyskinesias, cerebrovascular accidents, migraine, and memory and repair functions have been considered. As well as being of pathophysiological interest, much of this new knowledge has a direct diagnostic and therapeutic relevance.

Relationship of intrauterine fetal activity to maternal sleep stage

An attempt was made to evaluate the possibility of a humoral, rather than neural, mediation of the epiphenomena of REM sleep. The exclusively systemic communication between a mother and her fetus provided a readily available preparation for an initial approach to this problem. With reference to the sleep of newborn infants, it was assumed that periods of active motility in the fetus of a sleeping mother reflected the occurrence of REM sleep in this organism. Maternal sleep was monitored by standard recording procedures, and fetal sleep was inferred from recordings of intrauterine activity. The following results were obtained from comparisons of all-night sleep records of the mother-fetus unit: (a) The fetus tends to develop a recurrent quiet-active cycle in the course of gestation. (b) Specific instances of high fetal activity frequently occur together with or in close proximity to the recurrent episodes of REM sleep in the mother. Such increases, when they occur, appear to be related to the appearance of maternal REM sleep; however, they are not a systematic feature of this sleep phase. (c) A number of potential mechanisms for a dynamic humoral communication between the two physiological units during REM sleep are discussed.