Vegetative background of sleep: spectral analysis of the heart rate variability.

Phasic events during sleep, either arousal or REM-burst related, and the associated transient cardiovascular responses have been the subject of intensive research in previous studies. However, nontransient (stationary) fluctuations in heart rate have been studied less extensively in the past. They allow a differentiation of the sympathetic and parasympathetic activation, which are related to a low-frequency (LF) and a high-frequency (HF) component of the heart rate variability (HRV) signal, respectively. The resulting LF/HF ratio is a quantitative index of the sympatho-vagal balance. Sleep polygrams from 20 healthy volunteers were recorded in a sleep laboratory. Standard vegetative tests (orthostatic and Valsalva tests) were evaluated. A segmentation procedure, performed on the HRV signal, separated 70-130 transients during the night from the records of stationary heart rate fluctuations. From these periods the sympatho-vagal balance, quantified by the LF/HF, was computed by means of spectral analysis. The more synchronized the sleep was, the more the LF/HF decreased, whereas the LF/HF was significantly increased during REM sleep, indicating a sympathetic predominance during this period. Such an increase was also evident during the last 15 min before REM sleep onset. Results suggest that spectral analysis of the HRV provides additional information of the ultradian rhythmic behavior of the autonomic nervous system function beyond the traditional cardiovascular measurements (mean heart rate, blood pressure, etc.). In contrast to these measurements, which generally show a continuously decreasing cardiovascular activity, as the night proceeds, the results of this study reveal a high sympathetic peak activity during the later REM sleep periods, which is comparable in magnitude to that found in the upright position in wakefulness. This activity may be associated to the well-known incidence peak of ischemic events in the early morning hours.

The future of computer-assisted investigation of the polysomnogram: sleep microstructure.

Previous attempts at automated analysis of sleep were mainly directed towards imitating the Rechtschaffen and Kales rules (RKR) in order to save scoring time and further objectify the procedure. RKR, however, do not take into consideration the sleep microstructure of REM, stage 2, and SWS. While the microstructure of stage 2 has been analyzed in the past decade, the microstructure of REM and SWS are virtually unknown. In stage 2 the amount and distribution of spindles, K complexes, and arousal reactions have been studied. At least two types of spindles (12/s and 14/s) with different dynamics and locations have been identified. Two different shapes for K complexes have been described: one related to external sensory stimuli with similarities to evoked potentials and another one more related to sinusoidal slow wave activity seen in SWS. These two different K complex shapes have different distributions and, obviously, different functions. The authors also suggest that one should differentiate between arousal reactions and true arousals. Recent investigations suggest two types of delta waves in SWS. The more sinusoidal 1-3/s delta waves with a frontal maximum are already seen with lower amplitude in late stage 2 and increase their amplitude and incidence towards stage 3 and Stage 4. The other delta-wave type is slower (< 1/s), polymorphic, and has varying amounts of theta and higher frequency waves superimposed. During REM sleep it seems to be important to separate phases with rapid eye movements from those with none (REM sine REM), and count the amount and distribution of sawtooth activity. Background activity during REM and REM sine REM, as well as intra- and interhemispheric coherence should be analyzed separately. Only if the microstructure of the sleep EEG can be analyzed automatically using newer techniques such as transformation into wavelets and pattern classification with neuronal networks, and only if we learn more about the importance of microstructure elements, can automated sleep analysis go beyond the limited information obtained from scoring according to RKR.

Bivariate global frequency analysis versus chaos theory. A comparison for sleep EEG data.

Various quantitative descriptors for EEG data will be compared taking sleep as an example. In this contribution, Hjorth's mobility and complexity measures will be used to classify sleep stages. The results will be compared with those of a dimensionality analysis. Several authors have shown that the correlation exponent can describe the complexity of sleep EEG data and is able--with the exception of REM sleep--to distinguish significantly between sleep stages. The discriminative power of a bivariate global frequency analysis appears to be superior to that of the correlation exponent. Furthermore a very high statistical correlation between the estimator of fractal dimension and Hjorth's mobility was obtained.

Pilot controlled double-blind study of the hypnotic effects of zolpidem in patients with chronic 'learned' insomnia: psychometric and polysomnographic evaluation.

In a pilot double-blind trial in 21 patients with learned or idiopathic insomnia (DSM-IIIR), patients received placebo for 1 week (nights 1-7), either active (zolpidem, 10 mg) or placebo treatment for 2 weeks (nights 8-21) and then placebo for a further week (nights 22-28). Variables to measure efficacy, rebound and withdrawal were assessed daily from day 1 to day 28. Polysomnographic recordings together with sleep cycle analysis were performed on nights 7, 21 and 28. Patients treated with 10 mg zolpidem for 2 weeks had significantly improved sleep efficiency at the end of the randomised double-blind phase compared with the placebo group. Fractionated sleep-cycle analysis showed an increase in slow-wave sleep during the first 2-hour cycle after sleep onset. During the withdrawal placebo week, most of the main sleep variables remained relatively stable in the zolpidem group (nights 22-28), and deteriorated further in the placebo group. At the end of the withdrawal phase, there was a statistically significant difference between groups, in favour of the zolpidem treatment, in sleep efficiency, total sleep time, absolute and percentage of time awake, and percentage of REM sleep. REM sleep, which was normal in both groups at baseline, decreased significantly in the placebo group between nights 22 and 28 (during the withdrawal placebo week) compared with the zolpidem treatment group, and the number of periods of time awake increased. Minor subjective complaints were recorded under zolpidem and were comparable with those under placebo. Zolpidem seemed to improve some important sleep variables, when assessed both objectively and subjectively. The sleep cycle analysis suggested a possible shift of slow-wave sleep to an earlier period of the night, with a more physiological sleep structure. There was no evidence for withdrawal or rebound after stopping the 2 weeks of zolpidem treatment, but rather signs that the effect of zolpidem outlasted active treatment. The present pilot study justifies a prospective confirmatory comparison of zolpidem with benzodiazepines in an adequate number of patients and withdrawal after 6-8 weeks of treatment.

Spectral power and coherence analysis of sleep EEG in AIDS patients: decrease in interhemispheric coherence.

Fifteen patients aged between 26 and 55 years with the acquired immunodeficiency syndrome (AIDS) and various cerebral manifestations of the disease underwent an all-night sleep electroencephalogram (EEG) registration. The recordings of 15 age-matched volunteers were examined as controls. Sleep stages were determined visually and the following spectral analysis was based on corresponding artifact-free 40-second periods. The sampling rate was 64 second-1, the spectral resolution was 0.25 Hz and the frequency ranged from 0.25-24 Hz. The power density spectra of eight EEG derivations (left and right frontopolar, frontal, central and occipital; reference montage to the ipsilateral Cb electrodes) and the coherence spectra of interhemispheric (interfrontal, interoccipital) and intrahemispheric (frontooccipital, left and right) channel pairs were computed. The power density of the patients in the 11.5-13-Hz frequency range of nonrapid eye movement (NREM) sleep was considerably lower than that of the controls (p < 0.05 and p < 0.01 at left and right frontal derivations, two-tailed Mann-Whitney U test). The power density of rapid eye movement (REM) sleep showed no consistent differences between the two groups. The interfrontal coherence of the whole frequency range below 12 Hz was markedly lower in the patient group. This applied to NREM sleep and also to REM sleep (p < 0.01 and p < 0.001 for different frequency bands between 1 and 12 Hz in NREM and REM sleep). Possible relations to clinical features are discussed.

[Periodic discharges of sleep spindles with placebo and zopiclone]. / Periodische Entladungen von Schlafspindeln unter Plazebo und Zopiclone.

According to our experiences (parietal) sleep spindles often show a periodical appearance. Successive sleep spindles in series have a distance of about four seconds. In 95% the duration of such series of sleep spindles is not longer than 40 s, so that there appears no more than ten successive periodical sleep spindles. In the present study ten subjects aged between 57 and 77 years were given Placebo and 7.5 mg Zopiclone. Under the effect of verum we found a doubling of sleep spindle series. However, the distance between the periodical spindles and the length of the spindle series remained unchanged. The variance in the ability to generate sleep spindle series was very high across subjects. One subject for example produced only one sleep spindle series in the Placebo-night (three under verum) and another subject on the other hand 95 (under verum 130). The results of the present study well corresponding to the results of a former investigation comparing Placebo, Pentobarbital, Methaqualon, Carbromal, Flunitrazepam, Triazolam and Lormetazepam.

Topographical analysis of sleep spindle activity.

There is evidence for two types of sleep spindle activity, one with a frequency of about 12 cycles/s (cps) and the other of about 14 cps. Visual examination indicates that both spindle types occur independently, whereby the 12-cps spindles are more pronounced in the frontal and the 14-cps spindles in the parietal region. The purpose of this paper is to provide more information about the exact topography of these patterns. First the occurrence of distinct signals in anterior and posterior brain regions was verified using pattern recognition techniques based on matched filtering. Thus the existence of two distinct sources of activity located in the frontal and parietal region of the brain, respectively, was demonstrated using EEG frequency mapping. Evaluation of sleep recordings showed high stability both in the frequency and location of the presumed spindle generators across sleep. Pharmacological effects of lormetazepam and zopiclone on both spindle types were investigated. Both substances enhanced the sleep spindle activity recorded from the frontal and parietal electrodes, but this increase was more pronounced in the parietal brain region.

Pattern recognition by matched filtering: an analysis of sleep spindle and K-complex density under the influence of lormetazepam and zopiclone.

The evaluation of sleep EEG patterns is mostly accomplished by visual analysis. With modern personal computers however, it is possible to perform signal detection within a reasonable length of time automatically. This paper presents a method for signal processing based on matched filtering. This allows the detection of sleep spindles and K-complexes in a sleep EEG recording with a high degree of accuracy. First the technique is described, and the results of a validation study based on the comparison of visual evaluations and computer analysis are presented. Thereafter, results of an application study are presented. Sleep spindle and K-complex density under the influence of lormetazepam and zopiclone were examined. Under both medications sleep spindle density increased while K-complex density decreased. Computation of Pearson's correlation coefficients demonstrated that the interindividual sleep spindle and K-complex variations under both treatments are highly correlated. The data suggest that lormetazepam and zopiclone, although chemically different, have a similar mode of action and display comparable effects on the sleep EEG.

[Pattern recognition techniques in sleep polygraphy]. / Verfahren zur Mustererkennung in der Schlaf-Polygraphie.

The evaluation of EEG-patterns is usually accomplished by visual analysis. Nowadays however, even personal computers are fast enough for an efficient pattern recognition of EEG signals. Using sleep spindles and K-complexes as examples, our aim was to demonstrate how patterns can be detected in an EEG signal with a high degree of accuracy. Furthermore, recognition of K-complexes has been improved by applying an additional "adaptive algorithm" allowing individual adjustments to the signal's form and amplitude.

Short-term sleep EEG recordings after partial sleep deprivation as a routine procedure in order to uncover epileptic phenomena: an evaluation of 719 EEG recordings.

All night sleep deprivation prior to an EEG registration causes some inconvenience not only to the organization of the EEG department but presents a burden on the patients as well as their family members, and for these reasons is not suitable to be frequently employed as a routine procedure. As an alternative, we performed short-term sleep recordings in the early afternoon following a partial sleep deprivation of the patients during the preceding night. This method was well accepted by the patients and their family. Our only goal was to shorten the total time of night sleep using the following guideline: for very small children 22.00-06.00; for 4-14-year-old patients 24.00-06.00; and for patients older than that 01.00-06.00. 79.9%, out of 719 patients (573) who had been given the above instructions subsequently showed sleep patterns in their EEG. Additionally we had to administer an oral dose of promazine to only 67 patients. However, for the most part, patients showed only light sleep stages: 114 patients only reached sleep stage 1; 323 patients sleep stage 2; 88 patients sleep stage 3; and 48 patients sleep stage 4. As expected, REM sleep was never recorded. Nonetheless, in 32 out of 146 patients who were tired but unable to fall asleep, epileptic patterns could be provoked. In 636 patients, the EEG-recording after sleep reduction was ordered because of a suspected seizure disorder; in the remaining patients it was initiated solely because of sharp components in the routine-EEG. In 341 (53.6%) of the patients with suspected epilepsy, electroencephalographic activity indicative of a seizure disorder was activated. Such epileptic patterns were recorded almost exclusively in stages of waking, 1 and 2. Only in one out of the 124 patients who reached sleep stages 3 and 4 epileptic patterns were not seen until deep sleep was entered. We observed 2/s, 3/s and 6/s spike-and-wave complexes, sharp waves, spikes, polyspikes, groups containing remarkably sharp components and so called sharp vertex grapho-elements. Patients with suspected seizure disorders frequently show grapho-elements which can be interpreted as the expression of a disposition for epilepsy. These sharp vertex elements were evident in 54 out of 719 short term sleep recordings, more often in children than in adults. 49 times they coincided with typical epileptic discharges such as sharp waves, spikes or spike-and-waves in the same recording.

Periodicity analysis of sleep EEG in the second and minute ranges--example of application in different alpha activities in sleep.

To investigate the temporal organization of EEG sleep activity in the second and minute ranges we developed a method which, based on Fourier transformation, allows the presentation of periodic oscillations of spectral power and coherence. The application of this method is demonstrated in 3 subjects with different types of alpha activity during sleep: (a) alpha-sleep pattern (a physiological variant of NREM sleep activity); (b) abnormally increased arousal alpha activity. The results show that differences in the temporal organization of these alpha activities can be determined with the following parameters: period length, duration of sequences with periodic activity, number and rate of these sequences, and proportion of periodicities generated simultaneously in the left and right hemispheres. The physiologically modulated periodicities of the alpha-sleep pattern are contrary to a stereotyped 40-60 sec periodicity of abnormal arousal alpha activity. Such abnormal periodicity corresponds to periodicities occurring in association with other sleep disturbances, such as sleep apnea or periodic movements in sleep. Periodicity analysis gives additional criteria for a more refined evaluation of normal as well as abnormal sleep structure.

[Visual motor coordination of AIDS patients, HIV-positive asymptomatic probands and healthy persons during video-tracking]. / Visuomotorische Koordinationsleistung von AIDS-Patienten, HIV-positiven asymptomatischen Probanden und Gesunden im Videotracking.

Here we investigated the applicability of a computer-aided video-tracking as a method for evaluating potential deficits of neural information processing in patients with AIDS and those showing only positive HIV-seroreactivity. Video-tracking was accompanied with a simultaneous recording of EEG. Eight HIV-positive asymptomatic volunteers and eight AIDS-patients with cerebral manifestation of the disease participated in the pilot study. Two groups of eight normals each served as a control. Video-tracking performance of the HIV-positive volunteers and AIDS-patients significantly differed (p less than 0.05) from those of the healthy volunteers. Although the AIDS-patients' performance tended to be worse than that of the HIV-group, this difference was not significant. Power spectrum analysis of the EEG-data indicated that the diminished performance of the two test groups (AIDS and HIV-positives), accompanied by an increased spectral power across the entire frequency range measured in the study, could be an expression of an enhanced synchronization in cortical neuronal networks. The synchronization in turn could be a sign of possible organic brain damage resulting from HIV-infection. In conclusion, we suppose that video-tracking measures parameters which may indicate early deficits of information processing in CNS.

[Polygraphic sleep data in AIDS patients]. / Schlafpolygraphische Daten von AIDS-Patienten.

Somnopolygraphic recordings were registered from 29 patients with AIDS, their age ranged from 20 to 55 years (mean: 40.9; median: 44). The patients represent the full range of cerebral disintegration representing the picture of a progressing destruction of physiological sleep organization. Some disturbances begin quite early and progress successively, such as the reduction of REM- and delta-sleep as well as the reduction of sleep spindle- and K-complex-densities. Other changes are not manifest until intellectual capabilities break down; they occur massively such as the shortening of real sleep time and the reduction of sleep stage 2 with a simultaneous increase in waking time. It is remarkable that despite the enormous REM reduction there is no suppression of REM periods corresponding to "REM sine REM". REM periods become very short at an early stage; this is not only because the patients awake more frequently and earlier from their dream periods.

[Effects of flurazepam on disturbed sleep in patients with AIDS]. / Die Wirkung von Flurazepam auf den gestörten Schlaf von Patienten mit AIDS.

Polysomnograms were recorded of twelve patients with acquired immune deficiency syndrome (AIDS) during different stages in an open design. During the first night no hypnotics were administered, during the second night 30 mg flurazepam per os were given. Flurazepam affected mainly the NREM parameters. The times "awake" during the night were reduced, sleep stage 2 showed an increase, and the effective sleep time was also increased. The increase in sleep spindle density was remarkable, however, delta sleep and generation of K-complexes were not affected. Flurazepam did not affect REM sleep at all. The amount of REM sleep showed a slight increase. REM distribution during the night did not show the "bell shaped" increase and the decrease in the morning; the degree of the illness correlated with a flattening of REM distribution.

[Spectral power density and coherence in sleep EEG in patients with acquired immunodeficiency syndrome]. / Spektrale Leistungsdichte und Kohärenz im Schlaf-EEG bei Patienten mit dem erworbenen Immunschwächesyndrom.

15 male AIDS-patients from 26 to 55 years (mean 41.8 +/- 8.5) with various cerebral manifestations had a whole-night-sleep-EEG registration. As control the recordings of 15 age-matched volunteers (26-55 years, mean 41.8 +/- 9.8) were examined. Spectral characteristics of elementary EEG-epochs of 40 s length were computed, and sleep staging was performed visually for these intervals. The spectral power density of eight EEG-derivations (left and right frontopolar, frontal, central and occipital electrodes, reference montage to the ipsilateral Cb) were measured (sampling rate 64(-1) s, spectral resolution .25 Hz, frequency range from .25 to 24 Hz). Interhemispherical coherences of the frontal and occipital derivation pairs, and intrahemispherical fronto-occipital coherences of the left and right hemisphere, were computed. In the patients the frontal power density of NREM sleep showed lower values in the frequency range of 10 to 14 Hz. In central and in occipital derivations the power density between 12.5 and 15 Hz was lower in the patients, but the difference was less accentuated. The spectral power density of REM sleep showed similar characteristics in both groups. The interhemispherical frontal coherence of the whole frequency range below 13 Hz was markedly lower in the patient group. This was true for the NREM sleep, and, slightly less, for the REM sleep, too. The interoccipital spectral coherence was generally slightly lower in the patient group; the difference was most clearly in the 12.5 to 15 Hz range of NREM sleep.

Sleep EEG evaluation: a comparison of results obtained by visual scoring and automatic analysis with the Oxford sleep stager.

Oxford Medical has introduced an automatic sleep stager based on the stage-scoring criteria by Rechtschaffen and Kales. With our study we intended to examine whether the results of the stager (version 3.0) match those of the visual evaluation by two independent raters. We also wanted to test the reliability of this automatic sleep stage-scoring system. Ten somnopolygrams of subjects without sleep disturbances served as a basis for the comparison. Each sleep recording was scored twice automatically by the stager, twice visually by the first rater, and once by the second rater. The two automatic analyses of the somnopolygrams differed by 4.3% in a total of 13,850 epochs (1 epoch delta 20 s) regarding sleep stage scoring. The difference between the first and the second visual evaluation by the same rater amounted to 5.7%, whereas the results of the two independent raters deviated by 8.7%. Compared with the results of the visual analysis reached as a consensus by both raters--the so-called optimized visual analysis--the stager showed a 26.9% difference. The automatic analysis scored fewer epochs as stages wake, rapid eye movement (REM), and 2 and more as stages 1, 3, and 4. The sleep stager's frequent difficulty in identifying stage wake correctly as well as its incorrect allocation to other stages--mainly stage REM--could lead to misinterpretations of sleep recordings, whereas the increase in stages 1, 3, and 4, as compared with visual scoring, was negligible.

[The effect of age on sleep spindle and K complex density]. / Der Einfluss des Alters auf die Schlafspindel-und K-Komplex-Dichte.

The amount of sleep spindles and K-complexes shows a great interindividual variety of combinations, such as many or few sleep spindles and K-complexes respectively. There seems to be no direct correlation between the amount of sleep spindles and K-complexes intraindividually. In our unselected population - age range between 18 and 77 years - the mean sleep spindle density is at 2.59 +/- 1.85/min and the mean K-complex density at 1.96 +/- .96/min stage 2. The diffuse individual distribution, however, does not reflect the age factor involved. The sleep spindle and K-complex density were practically half the amount for the age group above 50 years as compared to the age group of less than 30 years.