SEARCH

SEARCH BY CITATION

Keywords:

  • cognition;
  • event-related potential;
  • excessive daytime sleepiness;
  • narcolepsy;
  • P300;
  • sleep inertia

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

Sleep propensity and sleep inertia were assessed in 43 patients with excessive daytime sleepiness (EDS) and 21 sleep-deprived controls, using a forced awakening test under continuous electroencephalographic (EEG) recording. Event-related potentials (ERPs) were first obtained in waking, while participants performed a target detection auditory task. Subjects were then allowed to take a nap with lights off and sleep latency was calculated. After 3 min of continuous sleep, frequent and rare tones were suddenly presented again (and ERPs recorded) in a forced awakening condition, which was repeated a second time if patients fell asleep. ERPs in pre-nap wakefulness did not differ in patients and controls. On forced awakening, almost half (48%) of EDS patients retained morphologically normal ERPs, but showed a significant delay of P300 relative to waking. In the other half of the patients (and none of the controls), the N200/P300 complex to targets was replaced on forced awakening by high-amplitude negative waves (‘sleep negativities’). Single subject analysis showed that 65% of patients had abnormal responses during forced awakening (significant P3 delay or sleep negativities), while only three of them (7%) had abnormal ERPs on wakefulness. The presence of sleep negativities was associated with shorter sleep latencies and increased target detection errors on forced awakening. Sleep negativities were more prevalent in narcolepsy and idiopathic hypersomnia than in EDS associated to psychiatric disorders. By combining sleep latency and ERP measures, the forced awakening test provided a robust and relatively rapid tool (45–60 min) to evaluate both sleep propensity and sleep inertia within a single recording session. The test allows each subject to act as his/her own control, thus increasing sensitivity. In the present series, it proved to be much more discriminative than waking ERPs alone to demonstrate specific abnormalities in patients complaining of excessive daytime sleepiness.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

Excessive daytime sleepiness (EDS) due to a variety of aetiologies is a major public health problem leading to severe behavioural consequences, including work and driving accidents (Dement and Mitler, 1993; Lyznicki et al., 1998). Its quantification remains a critical medical problem as the techniques currently available are time-consuming, and sometimes not accurate enough to discriminate patients from normal subjects. The most commonly used test of sleepiness evaluation is the Multiple Sleep Latency Test (MSLT) (Carskadon et al., 1986; Richardson et al., 1978), which measures iteratively the speed with which a person falls asleep throughout the day, usually referred to as ‘sleep latency’. Although the MSLT's ability to screen pathological sleepiness in large populations is beyond doubt, it is also admitted that reduced sleep latencies in individual subjects do not unequivocally indicate abnormal sleep propensity. Sleep latency can, for instance, be significantly reduced by increasing motivation to sleep (Harrison et al., 1996), and a proportion of normal-sleeping subjects can relax efficiently enough as to exhibit ‘abnormal’ sleep latencies without actual sleepiness or sleep-debt (Harrison and Horne, 1996). Independently of sensitivity problems, using sleep propensity as the only key symptom for diagnosis of EDS fails to cover the whole spectrum of pathological sleepiness, which also includes enhanced hypnopompic disorientation, more often called ‘sleep inertia’. This symptom has been referred to as a state of altered vigilance and disrupted behaviour observed immediately upon awakening (Dinges, 1989; Lubin et al., 1976), and reflecting a ‘struggle to move from sleep toward full alertness’ (Balkin and Badia, 1988). Although EDS patients often report spontaneous sleep inertia phenomena, either on morning awakening or after diurnal sleep attacks (Valley and Broughton, 1983), studies on this problem are scarce and have yielded contradictory conclusions. For example, in narcoleptic patients, behavioural performances were reported as improved (Billiard, 1976) or worsened (Godbout and Montplaisir, 1986) after a diurnal nap.

Sleep inertia has, to our knowledge, never been studied with electrophysiological measures such as event-related potentials (ERPs). Although several groups have recorded ERPs in patients suffering from pathological sleepiness, with various fortune (review in Bastuji and García-Larrea, 1999), these studies were always performed during wakefulness, and their aim was to demonstrate a basic cognitive impairment in EDS patients, rather than to quantify EDS intensity or participate to diagnosis. In the present study, we describe and apply a test to investigate and quantify a number of parameters commonly encountered in EDS, including acute sleep inertia (i.e. inertia occurring during the first minutes following the awakening signal) (Broughton, 1989; Chugh et al., 1996; Mullington and Broughton, 1994; Valley and Broughton, 1983). Our purpose was to obtain a valid estimation of both sleep propensity and sleep inertia in the context of a short afternoon nap. Thus, in the forced awakening test described herein, the subject learns to perform a simple discrimination task just before taking a nap, and has to resume the task after being suddenly awakened following a short sleep period. ERPs and behavioural measures are recorded first during wakefulness, and then during the state of acute sleep inertia that follows sudden awakening (Dinges, 1990; Ferrara et al., 2000; Van Dongen et al., 2001). As each subject is recorded during two different contextual conditions, the patient may act as his/her own control for both behavioural and ERP measures. In addition to sleep inertia, the forced awakening test permits to estimate some sleep parameters such as sleep latency and speed of sleep progression, and correlate these variables with both ERPs and behaviour during sudden awakening.

Although both sleep propensity and sleep inertia have different meaning in ‘open field’ (clinical) and in controlled (laboratory) conditions, our working hypothesis was that the electrophysiological signs of acute sleep inertia should be more pronounced in EDS patients than in normal subjects, even under a ‘non-ecological’ recording context. To test this hypothesis, we applied the forced awakening test to a sample of patients complaining of excessive daytime sleepiness from different aetiologies, and to a control group of healthy subjects with partial sleep deprivation. Among the ERP components that change behaviour during sleep and somnolence, the N2-P3 (or N200-P300) complex is particularly prominent (Bastuji et al., 1995; Colrain et al., 2000a; Gora et al., 2001; Harsh et al., 1994; Nielsen-Bohlman et al., 1991; Niiyama et al., 1994, 1996; Salisbury et al., 1992). These components, which reflect the time needed to evaluate and classify task-relevant sensory stimuli, are usually delayed during the wake–sleep transition, and then replaced by high-amplitude fronto-central negative waveforms that are unique to slow-wave sleep (N350, N550, reviews in Bastuji and García-Larrea, 1999; Campbell, 2000; and Halász, 1998). N350 is known to occur during sleep onset and all along slow-wave sleep, while the N550 fully develops in slow-wave sleep only. These two peaks are also known as ‘sleep negativities’, being usually separated by a small positivity (P450). We hypothesized that similar changes could develop in EDS patients (but not in controls) during a forced awakening test, reflecting a greater difficulty to pass from sleep to the waking state, and thus witnessing the ‘struggle to move from sleep toward alertness’ that characterises sleep inertia (Balkin and Badia, 1988). We further assumed that ERP changes would be related to the severity of pathological sleepiness, and possibly also to the aetiology of EDS.

Patients

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

We recorded 43 patients, mean age 28.3 ± 10.5 years (22 women) in whom the forced awakening test was included in the routine screening of their EDS. The test was performed between 14.00 and 15.30 hours, with the subjects lying on a bed in a dimly illuminated room.

Each patient's final diagnosis was determined after completion of clinical and paraclinical investigations, i.e. between 1 week and 2 months following the forced awakening test. Thus, precise diagnosis was not known at the moment of the forced awakening test. After complete evaluation, 14 patients (32.5%) were found to have typical clinical and polysomnographic features of narcolepsy, including two or more paradoxical [rapid eye movement (REM)] sleep onset in 24-h polysomnography, and/or cataplexy. Fourteen patients (32.5%) were classified as having idiopathic hypersomnia; of them, eight completely fulfilled International Classification of Sleep Disorders (ICSD, 1990) criteria for this condition, (i.e. complaint of prolonged sleep episodes, excessive sleepiness, or excessive deep sleep and prolonged nocturnal sleep), while six others presented with short and refreshing daily sleep attacks, short sleep latency without clinical sleep inertia in the morning, and no signs of narcolepsy nor of respiratory disorders. Four patients (9.3%) had sleep apnoea syndrome (SAS) with apnoea–hyponoea index between 20 and 40/h. One patient had neurological hypersomnia related to a tumour of the pineal area. The remaining 10 patients were classified as ‘other hypersomnia’; this group included nine patients with EDS associated to psychiatric disorders (anxiety or depression) and one subject with sleep deprivation and inverse sleep/wake cycle in a context of night work.

Recordings were performed before treatment in 36 patients (84%), and after 24-h discontinuation of modafinil in the other seven cases (16%). The choice of this time lapse was based on pharmacokinetic data on the drug (Civil et al., 1995).

Control subjects

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

Results were compared with those of a control group of 21 subjects (14 women) of matched age (23.9 ± 3.8 years), who participated in the experiments after providing informed consent. They were all MDs or undergraduate students free of neurological symptoms, not addict to drugs and naive to the experimental procedure. Participants to the study were told that the ERP recording session would include recordings during somnolence and sleep, and arrived to the laboratory partially sleep-deprived (2 h lesser than their usual time in bed). Time of testing, subjects’ instructions and recording procedure were identical for controls and patients.

Stimulation and recording

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

Auditory stimuli were sinusoidal tone bursts of 60 ms duration (10 ms rise and fall time, 40 ms plateau) delivered binaurally at 70 dB hearing level (HL) through inserted earphones (EuropSonic®, Roanne, France). Ninety per cent of stimuli were ‘non-target’ 1000 Hz tones, and 10% were ‘target’ tones of 2000 Hz. Inter-stimulus interval (ISI) varied randomly between 1125 and 1375 ms (mean 1250 ms). Each recording run lasted about 3 min and comprised 15–20 target stimuli.

Brain activity was recorded with 20–32 scalp tin electrodes fixed to an elastic helmet (electrode cap) according to the international ‘10–20’ System (Jasper, 1958; Steinmetz et al., 1989), and referenced to the nose. The electro-oculogram (EOG) was monitored by a silver/silver chloride cup electrode attached to the infero-lateral margin of the right orbit, also referenced to the nose. A ground electrode was placed midway between the Fz and Fpz positions. Electrode/skin impedances were kept below 2 kΩ. The electroencephalogram was amplified differentially 30 000 times (bandpass 0.3–30 Hz) and digitized with a time resolution of 4 ms over an analysis time of 1024 ms, including 80 ms of pre-stimulus delay used for baseline computation. ERPs were obtained online with a Brain Atlas® (Bio-Logic Sys. Corp., Chicago, IL, USA) in the first 35 patients and 16 controls. In the remaining subjects, the continuous EEG (Ceegraph®; Bio-Logic Sys. Corp.) was segmented offline, averaged according to stimulation type, displayed and analysed using a BPM® System (Orgil Corp., Orgil Medical Equipment, En Ayalla, Israel) modules. An automatic artefact-rejection system excluded from the average any trial containing transients exceeding ± 65 μV at any recording channel, including the EOG. For topographic analysis data were exported to Brain Atlas® after averaging.

Experimental procedure

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

Subjects arrived to the laboratory in the early afternoon, and the whole recording session took place between 14.00 and 16.00 hours in all patients and control subjects. ERPs to auditory tones were recorded during a ‘counting oddball’ discrimination paradigm where the subjects had to detect and count silently the number of infrequent high-pitched tones (‘targets’) presented with 10% probability among a flow of low-pitched (non-target) sounds. The procedure of the test was as follows:

  • (a)
     After describing the principle of the detection task, ERPs were first recorded with the subjects awake (wake-1 condition). The number of correct detections was verified after the run, by comparing the number of targets detected with that actually delivered.
  • (b)
     Then, the subjects were prompted to take a nap and were covered with a blanket to ensure comfort. They were told that if they did not manage to sleep during naptime, the test would be over after 20 min of continuous recording with lights off. They were also explicitly instructed to restart the counting task of target tones should they hear them again during the nap period. They were also systematically told that the test would be equally useful if they did not hear anything during their nap. Neither patients nor controls knew in advance that the experimental procedure included obligatorily the presentation of auditory tones during the nap.
  • (c)
     Both EEG and EOG were continuously monitored, and sleep latency was calculated as the time elapsed from lights off to sleep onset. The latter was defined as the onset of stage I, according to the criteria of Rechtschaffen and Kales (1968). Two first-line and two secondary criteria were employed to define sleep I onset. The first-line criteria were (1) replacement of occipital alpha rhythm by low amplitude theta and occasional delta waves, and (2) predominance of theta over alpha activity on central, parietal and occipital leads. Second-line criteria were the presence of (3) vertex sharp waves and (4) runs of centro-frontal alpha activity. Presence of two first-line criteria, or one first-line and one second-line criteria were sufficient to define stage I. Once the subject was asleep, microarousals of 10 s or less did not prevent from quoting sleep as ‘continuous’, unless they occurred more than twice in a minute. Conversely, sleep was considered as interrupted if waking EEG activity reappeared for more than 10 s.
  • (d)
     After 3 min of continuous sleep, subjects were stimulated again by a new series of high- and low-pitched tones (forced awakening-1). After completion of the stimulation run the subjects were allowed to continue their nap. If a subject did not have a full epoch of 3–4 min stable sleep within the 20 min allowed, he was not stimulated and the test was considered as normal. If sleep was obtained anew, the procedure was repeated, i.e. the subject was stimulated again after 3 min of continuous sleep (forced awakening-2). After this second run, the nap was interrupted and the subjects immediately questioned about their memories of the stimulation, i.e. whether they remembered to have been stimulated during the nap, and, in that case, which was the number of target stimuli they were able to remember. Subjects were also debriefed as to the subjective quality of their sleep and the content of any mental/oniric activity during the nap.
  • (e)
     After complete debriefing and with lights open, a last recording run was obtained with the subjects fully awake, 5 min after the end of the nap period (wake-2 condition).

A graphical description of the procedure is presented in Fig. 1. The whole recording session (including electrode positioning) lasted about 1 h.

image

Figure 1. Experimental procedure. Event-related potentials (ERPs) were first recorded during wakefulness (wake-1 condition). Then, with lights off, the subjects were prompted to take a nap and after 3 min of continuous sleep were suddenly awakened by a new series of stimuli (forced awakening-1). After completion of the stimulation run, if sleep was obtained anew, the procedure was repeated (forced awakening-2). Then, with lights on, a last recording run was obtained during wakefulness, 5 min after the end of the nap period (wake-2 condition).

Download figure to PowerPoint

Criteria for ERP components

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

The N1, N2 and P3 responses to target stimuli were analysed. According to Guidelines of the International Federation of Clinical Neurophysiology (Heinze et al., 1999), the N1 and N2 components were defined as the negative waves of fronto-central topography peaking, respectively, in the 80–120 ms (N1) and the 160–250 ms (N2) latency range. The P3 (or P300) was considered as the maximal positive activity peaking at within the 280–470 ms interval, with central–parietal midline predominance. When two distinct peaks appeared in the interval the mid-latency between them (obtained by extrapolation of ascending and descending branches of the waveform) was taken as the component latency (Goodin et al., 1994). Amplitudes were measured from pre-stimulus baseline, at Fz and Cz electrode sites for N1 and N2 and at Fz, Cz and Pz for P3; if two peaks were present in the interval, the amplitude of the highest was taken.

In a number of patients, the morphology of ERPs changed on forced awakening, the N2-P3 waves being replaced by two distinct high-amplitude negativities of fronto-central distribution, that have been described during light sleep by different authors (Bastuji et al., 1995; Colrain et al., 2000a; Harsh et al., 1994; Nielsen-Bohlman et al., 1991; Ujszászi and Halász, 1986, 1988). These ‘sleep negativities’ were labelled respectively ‘sleep-N2’ (or ‘N350’ peaking at 300–450 ms over Cz) and ‘sleep-N3’ (or ‘N550’ at 450–650 ms over Fz). Responses containing sleep negativities were categorized as ‘sleep pattern’ ERPs, as opposed to those showing the usual N2-P3 sequence, or ‘wake pattern’ ERPs (see examples in Fig. 2). The group of patients presenting a sleep pattern ERP on at least one forced awakening and that of patient always presenting a waking pattern were analysed separately.

image

Figure 2. Event-related potentials (ERPs) of three patients representative of the different abnormal patterns observed on forced awakening condition. For each patient, the responses to deviant stimulus on wakefulness (thin traces) and forced awakening (thick traces) are superimposed; the percentage of hits indicated corresponds to that of forced awakening. (a) Waking pattern with P300 delay; (b) sleep pattern with a N350 component (vertex sharp wave); and (c) sleep pattern with a N350 and a N550 (K complex).

Download figure to PowerPoint

Other variables

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

The percentage of errors after each run was calculated by dividing the number of targets omitted by the number of targets actually delivered.

The dynamic changes in vigilance during the forced awakening condition were scored and classified in four different groups: (1) subject awaken by the stimuli and remaining awake; (2) subject awaken by the stimuli then waxing and waning of vigilance between wake and stage I; (3) subject awaken by the stimulus for less than 1 min, the returning to sleep; and (4) subject remaining asleep despite the stimuli.

In addition to classical sleep scoring (Rechtschaffen and Kales, 1968), the power spectrum of the EEG during the 10 s preceding the first auditory stimulus was computed in patients and controls for each of the nap session. Mean spectral power was first calculated individually for each band, on the electrode showing maximal values [i.e. on Cz for delta (0.5–4 Hz) and on Pz for theta (4–7.5 Hz), alpha (8–11.5 Hz) and beta (12–25 Hz)] bands. Then, the alpha + beta/theta + delta ratio of mean power was compared in patients and control subjects with a t-test and correlated with the ERP results.

Statistical analysis

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

Amplitudes and latencies of ERP components were submitted to a two-way, repeated-measures analysis of variance (anova) with two factors: (a) experimental condition (pre-sleep wakefulness versus forced awakening-1 versus forced awakening-2 versus post-sleep wakefulness) and (b) electrode position (Fz versus Cz for N1 and N2, or Fz versus Cz versus Pz for P3). Separate anova were performed for controls, patients with ‘waking pattern’ and patients with ‘sleep pattern’ ERPs on forced awakening. Only results reaching significance at P < 0.05 (after Geisser-Greenhouse correction when needed) are presented. Post hoc bilateral paired t-tests were performed when significant results emerged from the anova.

T-tests and Mann–Whitney test were also used to compare behavioural data in the two ERP groups of patients, those with a sleep pattern and those with a waking ERP pattern. The possible association between ERP patterns on forced awakening and the aetiology of EDS was tested using chi-square statistics.

On the basis of data analysis in control subjects (see Results), the forced awakening test was considered normal if:

  • 1
    no sleep episode occurred during 20 min after lights off, or
  • 2
    sleep did occur, and P3 latency changes on forced awakening did not exceed 40 ms (corresponding to 99.9% confidence limits for P3 changes in controls).

The test was considered abnormal if:

  • 1
    changes in P3 latency on forced awakening (relative to waking) equalled or exceeded 40 ms (99.9% upper confidence limit in controls), or
  • 2
    ERPs on forced awakening changed morphology relative to waking, with the appearance of N350 and/or N550 components typical of sleep.

Behavioural data

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

Table 1 summarizes behavioural data of patient and control groups. At least one nap period was obtained after lights off in 40 of 43 patients (93%), and in 15 of 21 sleep-deprived control subjects (71%). A second episode of sleep (after the first forced awakening) was obtained in 81% of patients, but only 48% of controls. Differences were significant at the P < 0.01 level (chi-squared test). The mean sleep latency for the first sleep period in EDS patients was 6.2 ± 5.2 min. While none of the control subjects fell asleep in less than 10 min, 27 patients (63%) had abnormally reduced sleep latency relative to standard normal limits (<8 min, Van den Hoed et al., 1981). Also, patients were significantly more numerous than controls to reach stage II during the 3-min period that preceded forced awakening (53% versus 13% during the first nap, and 74% versus 14% during the second nap, P < 0.01).

Table 1.  Behavioural results in patients and control subjects. For each group and on each forced awakening condition (forced awakening-1 and forced awakening-2), the percentage of subjects who fell asleep, the mean sleep latency, the percentage of subjects reaching stage II, the mean percentage of errors on the counting task are indicated
SubjectsSleep obtained (%) Sleep latency (min)Sleep stage II obtained (%)Percentage of errors
Patients (n = 43)
 Forced awakening-1936.2 ± 5.25311.5 ± 19.6
 Forced awakening-2813.1 ± 3.77445.8 ± 44.4
Controls (n = 21)
 Forced awakening-171>10131.5 ± 2.7
 Forced awakening-2489.5 ± 6141.5 ± 2.4

During the waking periods (before and after the nap) the patients’ counting performance was strictly comparable with that of controls (average <4% of errors in both groups, non-significant difference). Conversely, the percentage of errors in the patients group increased significantly relative to controls during the first forced awakening (1.5 ± 2.7% in controls versus 11.5 ± 19.6% in patients, Mann–Whitney z = −2.1, P < 0.05). The difference became still greater on awakening from the second sleep period (1.5 ± 2.4% versus 45.8 ± 44.4%, z = −4.4, P < 0.001).

ERPs during wakefulness

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

Event-related potentials obtained in the waking state were of similar morphology in controls and patients. No significant difference in latency or amplitude was observed for the N1 to target stimuli. The N2-P3 complex was very slightly delayed (by less than 10 ms in the average) in patients relative to controls; differences were not significant for P3 (325.7 ± 22.6 ms versus 318 ± 23 ms, t = −1.35, P = 0.18), but reached significance for N2 (210 ± 13 ms versus 201 ± 16.7 ms, t = −2.4; P < 0.02). P3 amplitude during wakefulness was equivalent in patients and controls (19.6 ± 9.8 μV versus 18.4 ± 8.8 μV), while N2 amplitude was significantly reduced in the patient group (−3.1 ± 5.8 μV versus −6.8 ± 4.9 μV; t = −2.5 P < 0.02). When compared with our ERP laboratory databank, only three patients (6.9%: one narcolepsy, one idiopathic hypersomnia and one mental disorder) could be considered as ‘abnormal’ (i.e. exceeding normal N2-P3 limits in age-matched controls by more than 2.5 SD). All control subjects fell within normal limits of latency and amplitude.

Control subjects (Fig. 3)

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References
image

Figure 3. Grand-averaged event-related potentials (ERPs) in the 10 control subjects in whom two forced awakenings were obtained and presented for illustrating purpose. From top to bottom: responses obtained during pre-sleep, forced awakening-1, forced awakening-2 and post-sleep waking. On the left, the traces recorded on Fz, Cz and Pz in response to frequent (thin traces) and to target tones (thick traces); on the right the corresponding maps of P3. The amplitude decrease of P300 during forced awakening is the more striking result.

Download figure to PowerPoint

Fifteen of the 21 control subjects had enough sleep to allow performing at least one recording on forced awakening. Of them, 10 were able to fall asleep twice (i.e. have a second nap after a first awakening), and could therefore undergo two forced awakening tests. anova was accordingly performed using three conditions (pre-sleep, forced awakening-1 and post-sleep) in 15 subjects, and four conditions (including two forced awakenings) in 10. As the anova results were the same in both cases, only those concerning three conditions (pre-sleep waking, forced awakening-1 and post-sleep waking) will be presented here.

As shown in Fig. 3, illustrating grand averaged traces across control subjects in each condition, the overall ERP morphology did not change in control subjects during forced awakening relative to the waking conditions. ERP latencies were not significantly modified by the test, although there was a non-significant trend towards P3 latency increase after sudden awakening (+10 ms in the average). There was a significant effect of vigilance condition on P3 amplitude [F(2,14)=4.56; P < 0.02]. On post hoc analyses, P3 amplitude was significantly attenuated on forced awakening relative to pre-sleep waking at Fz (t = 2.8; P < 0.02) and Cz (t = 2.3; P < 0.05) and to post-sleep waking at all electrodes (Fz: t = 3.2; P < 0.01; Cz: t = 2.2; P < 0.05; Pz: t = 2.4; P < 0.05) (see Fig. 4, left part). There was no significant effect of the electrode position on P3 amplitude, nor interaction between conditions and electrodes.

image

Figure 4. Mean latencies (top) and amplitudes (bottom) (±SEM) of P300 on Fz, Cz and Pz during each experimental conditions (pre-sleep waking, forced awakening-1, forced awakening-2, post-sleep waking). Values obtained in controls (left), patients with a waking pattern (middle) and patients with a sleep pattern ERPs (right). Asterisk represents a significant difference (P < 0.05) of the values as compared with pre-sleep waking, one on post hoct-test.

Download figure to PowerPoint

EDS patients

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

In about half (n = 19, 48%) of EDS patients who fell asleep, the overall morphology of target ERPs was the same on forced awakening and in wakefulness (as was the case of control subjects), i.e. composed by the common N1-P2-N2-P3 sequence (see Figs 2a, 5 and 6a). Conversely, in the other rough half of patients (n = 21, 52%), the morphology of the evoked response changed and included a high-amplitude negative wave with vertex maximum, peaking over Cz at 373 ± 60 ms, and followed sometimes by a later negative wave of frontal predominance, peaking over Fz at 556 ± 50 ms (Figs 2b,c, 5, 6b and 7). These two responses were collectively labelled ‘sleep negativities’ by analogy with the negative responses to auditory stimuli that have been previously described during sleep (see Methods), and to which they appeared to correspond. Twelve patients showed only the earlier sleep negativity N350, and five others only the second N550, while the two appeared together in four further cases (individual examples and variants of this ERP pattern are illustrated in Fig. 7). Eleven patients exhibited this particular pattern during the two consecutive episodes of forced awakening, while the other 10 showed it during only one forced awakening episode. Patients with sleep negativities on at least one forced awakening were said to have a ‘sleep pattern’ of ERPs, while the others were considered as having a ‘waking ERP pattern’ (see Methods). Examples of the ‘waking’ and ‘sleep’ ERP patterns on forced awakening are illustrated in Figs 2, 6 and 7. As these two ERP patterns were morphologically very different, it was impossible to include both of them in the same parametric analysis. Statistical tests were, therefore, applied separately to each group of patients, according to their ERP pattern on forced awakening.

image

Figure 5. Grand-averaged event-related potentials (ERPs) to target tones obtained in the two sub-groups of excessive daytime sleepiness (EDS) patients, presented for illustrating purposes (see Fig. 4 for actual amplitude and latency values measured individually). The upper row illustrates the ‘waking ERP pattern’ on forced awakening; the lower row the ‘sleep’ ERP pattern (with sleep negativities). Left, middle and right columns depict responses obtained in (respectively) the first forced awakening, the second forced awakening, and the waking post-nap session, in each case superimposed to traces obtained prior to the nap (Waking-1). In the group with waking pattern ERPs, P300 latency was significantly delayed during the two forced awakenings as compared with wakefulness. Although there seems to be some amplitude reduction in grand averages, this effect was not significant (see Fig. 4). In the group with sleep ERP pattern, a N350 appears within the P300 latency range, followed by a delayed P3 on the forced awakening-1 condition, and by a ‘sleep N550’ in forced awakening-2. On post-sleep wakefulness the P300 still shows slight but statistically significant latency delay and amplitude decrease.

Download figure to PowerPoint

image

Figure 6. Event-related potentials (ERPs) of two individual patients, each representative of one abnormal ERP pattern observed on forced awakening. Responses to deviant stimulus on wakefulness (thin traces) and forced awakening (thick traces) are superimposed; the percentage of hits indicated corresponds to that of forced awakening. (a) ‘Waking ERP pattern’ with P300 delay without amplitude decrease; (b) ‘sleep ERP pattern’ showing a prominent N350 component (in phase inversion with the waking P3) over the midline (Fz-Cz). Note that delayed P3 still remained on posterior areas on forced awakening.

Download figure to PowerPoint

image

Figure 7. Midline event-related potentials (ERPs) of six different patients showing the ‘sleep ERP pattern’ during the second forced awakening condition. Responses to deviant stimulus obtained during wakefulness (thin traces) and forced awakening (thick traces) are superimposed; the percentage of hits indicated corresponds to that of forced awakening. Please note that during forced awakening the N350 sleep negativity appears in the same latency range as the waking P300. Note also that the P3s recorded on forced awakening are of very different latency and amplitude from one patient to another.

Download figure to PowerPoint

EDS patients with waking ERP pattern (Figs 4 and 5) In patients with a ‘waking’ ERP pattern on forced awakening, the experimental condition had a significant effect on both N2 and P3 latencies [respectively F(3,12)=3.40; P = 0.05], and [F(3,12)=6.13; P < 0.02]. On post hoc analyses, P3 latency was significantly delayed during both forced awakening conditions as compared with waking (see Fig. 4, middle), while N2 latency was significantly delayed on the second forced awakening only (t = 2.71; P < 0.05). Furthermore, while the vigilance condition had no significant effect on N1, N2 and P3 amplitudes, there was a significant effect of electrode position [F(2,12)=21.05; P < 0.001] on P3 amplitude, which was always smaller on Fz as compared with Cz and Pz (P < 0.001). There was no significant interaction between factors on P3 amplitude.

A separate anova was performed on fronto-parietal amplitude ratios, to test for a possible topographic change of P3 across conditions. As in controls, the ratio was normalized, of the form (a − b)/(a + b), where a and b were, respectively, P3 amplitudes at Pz and at Fz on each of the experimental conditions. The design of this anova included a specific contrast between the two waking and the two forced awakening conditions. A very significant effect of the experimental condition was obtained [F(3,11)=4.73; P = 0.017], as well as a very significant effect of the planned contrast between ‘waking’ and forced awakening situations [F(1)=13.28, P = 0.003], suggesting that the forced awakening test modified significantly the topographical amplitude gradients of P3 in our patients, by decreasing the amplitude at frontal sites relative to that at parietal sites (see Fig. 4).

EDS patients with sleep ERP pattern (Figs 4–8) In this group the classical waking N2-P3 complex of target ERPs disappeared on forced awakening (see above section) and, therefore, anova effects could not be adequately studied on N2 or P3. In order to quantify latency and amplitude changes of components, statistical analyses were however applied, labelling N2 and P3 the consecutive negative and positive waves that occurred within the predetermined latency range of N2 and P3 (see Methods). On these premises, a significant effect of the vigilance condition was evidenced on the latencies of N1, N2 and P3 [respectively, F(3,19) = 3.46, P < 0.05; F(3,19) = 3.75, P < 0.05; and F(3,19) = 9.74, P < 0.001]. Post hoc analysis showed that the latency of each component was delayed on forced awakening-1 as compared with W1 (N1: t = −2.32; P < 0.05; N2: t = −2.31; P < 0.05; P3: t = −4.7; P < 0.001) and only that of N2 and P3 on forced awakening-2 as compared with W1 (N2: t = −2,8; P < 0.02; P3: t = −4.3; P < 0.001). The vigilance condition had no significant effect on N1 and N2 amplitudes. A significant effect of the experimental condition was observed on P3 amplitude [F(3,19)=3.55; P < 0.05], and on post hoc analyses P3 was smaller on forced awakening-1, forced awakening-2 and wakefulness after the nap (W2) as compared with that prior to the nap (W1). Finally, there was also a significant ‘electrode’ effect on P3 latency [F(2,19)=5.59; P < 0.02] and amplitude [F(2,19)=34.44; P < 0.001], which were, respectively, shorter and smaller at Fz relation to the other midline electrodes (see Fig. 4, right).

image

Figure 8. Sub-groups of excessive daytime sleepiness (EDS) patients showing ‘sleep’ event-related potential (ERP) pattern on forced awakening. Left: superimposition of grand-averaged ERPs to target tones obtained in waking-1 (thin traces) and forced awakening-1 (thick traces); superimposition of responses at Fz, Cz, Pz. Right: maps of P300 during waking-1 and of ‘sleep N350’ on forced awakening-1.

Download figure to PowerPoint

Sleep latency, sleep stability and errors in patients

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

Patients with a ‘sleep’ ERP pattern on at least one episode of forced awakening had signs of increased sleep pressure as compared with those who always retained a ‘waking’ ERP pattern. This was reflected by significantly decreased sleep latencies for stage I and stage II, in both nap periods (Fig. 9, left). Also, 100% of the patients with ‘sleep ERP pattern’ were able to fall asleep again after a first forced awakening, while only 68% of those with a ‘waking pattern’, and 48% of controls, did so. The ability to maintain vigilance during presentation of stimuli was also different in the two groups of patients. None of the ‘sleep pattern’ group was able to remain awake during the whole run, while 34% of the ‘waking pattern’ ERP group did. Half (50%) of the sleep pattern group remained awakened less than 1 min after the first stimulus and then returned to sleep, while this was the case for only 18% patients with waking pattern. Twenty-nine per cent (29%) of the ‘sleep pattern’ group was awakened by the first stimuli of the run, and then presented waxing and waning of waking and sleep, mainly stage I; this was observed in 43% of the waking pattern group. Finally, while 21% of patients with sleep pattern remained asleep without being awakened by the stimuli, this occurred in only 5% of the waking pattern group.

image

Figure 9. Left: mean sleep latencies in the ‘waking type’ group of patients (empty histograms), as compared with the ‘sleep pattern’ group (filled histograms). Data are represented separately for forced awakening-1 and forced awakening-2. Patients with ‘sleep pattern’ event-related potentials (ERPs) had significantly shorter sleep latencies in both sessions. Right: percentage of errors on target detection according to the presence of a waking or a sleep pattern on the evoked response. Note that comparisons in the left part refer to patients, while in the right part they refer to ERPs. A patient was considered to be in the ‘sleep pattern group’ if ERPs contained sleep negativities in at least one forced awakening session (see Methods).

Download figure to PowerPoint

The percentage of counting errors associated to ERPs containing sleep negativities (i.e. ‘sleep pattern ERPs’) was significantly higher than that observed in association with waking ERP morphology (Fig. 9, right).

ERP patterns on forced awakening and aetiology of EDS

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

There was a significant association between the aetiology of EDS and the ERP pattern on forced awakening (P < 0.04, chi-squared). A ‘sleep’ ERP pattern was observed in the majority of patients with narcolepsy and idiopathic hypersomnia, as well as in one patient with SAS, in the patient with a pineal tumour and the one presenting inverse sleep/wake cycle and sleep deprivation. Conversely, patients with EDS related to psychiatric disorders, and those in whom investigations failed to confirm EDS, did not fall asleep or had a ‘waking’ ERP pattern on forced awakening (Fig. 10).

image

Figure 10. Three-dimensional diagram illustrating the relation between event-related potential (ERP) patterns on forced awakening and the aetiology of excessive daytime sleepiness (EDS). Abscissa: the different categories of EDS diagnosis; depth: the two ERP patterns observed on forced awakening; ordinate: number of patients. Most patients with idiopathic hypersomnia and narcolepsy had ‘Sleep Pattern’ ERPs (sleep negativities), while patients with EDS related to mental disorders retained ‘waking type’ ERPs on sudden awakening (P < 0.04 on chi-squared, corrected).

Download figure to PowerPoint

On the basis of the normative data derived from control subjects, 65% of the patients’ sample could be classified as having abnormal responses during at least one forced awakening. In 21 of them (49%), this was because of the presence of a sleep ERP pattern on forced awakening (which was considered as abnormal as never occurring in control subjects), while in the other 16% abnormality was the result of excessive latency increase of P3 during forced awakening. In regard to diagnostic categories, 86% of narcoleptic patients, 79% of patients with idiopathic hypersomnia and 100% of the four patients with SAS had abnormal ERPs on at least one forced awakening. On the other hand, all the patients with psychiatric hypersomnia and the two misdiagnosed patients (in whom complete diagnostic evaluation failed to confirm EDS) had normal responses during forced awakening (see Table 2).

Table 2.  Results in the different diagnostic groups of patients according to the International Classification of Sleep Disorders (ICSD, 1990). For each group and on each forced awakening (forced awakening-1 and forced awakening-2), the percentage of patients who fell asleep, their mean sleep latency, the percentage of patients reaching stage II, the percentage of patients with abnormal event-related potential (ERP) on forced awakening, and the mean percentage of errors on the counting task are indicated
SubjectsSleep obtainedSleep latency (min)Sleep stage II obtainedAbnormal ERPPercentage of errors
Narcolepsy (n = 14)
 Forced awakening-1100%3.8 ± 2.650%57%15.9 ± 25.6
 Forced awakening-286%2.2 ± 2.671%71%56.3 ± 46.3
Idiopathic hypersomnia (n = 14)
 Forced awakening-1100%5.9 ± 4.357%64%9.4 ± 13.5
 Forced awakening-2100%2.4 ± 2.979%64%44.5 ± 43.8
Mental disorders (n = 9)
 Forced awakening-167%13.2 ± 6.30%0%0%
 Forced awakening-233%9.3 ± 60%0%0%
Sleep apneoa syndrome (n = 4)
 Forced awakening-1100%7.5 ± 5.750%75%8.7 ± 7.6
 Forced awakening-2100%5.0 ± 3.750%75%8.7 ± 9.6
Lesional hypersomnia (n = 1)
 Forced awakening-11/111/11/1100
 Forced awakening-21/111/11/1100
Sleep deprivation (n = 1)
 Forced awakening-11/111/11/1100
 Forced awakening-21/111/11/1100

Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

During the 3-min period of sleep that preceded forced awakening, EDS patients reached stage II significantly more often than did control subjects (70% versus 33%), and therefore the percentages of awakenings in stages I and II were not the same in both groups. As the sleep stage from which the subjects were awakened by the stimulus may have influenced the EP response, a separate analysis was conducted, limited to subjects aroused from stage I (i.e. excluding those who reached stage II during the test). Both conventional scoring (Rechtschaffen and Kales, 1968) and spectral analysis (alpha, theta and delta power) during the 10 s that preceded forced awakening ascertained the similitude of the sleep stage in this subset of controls and patients. As shown in Fig. 11, while there was no significant difference in ERPs between both groups latency during wakefulness, this latency was found increased in patients relative to controls on forced awakening [patients (n = 19) 360 ± 60 ms versus controls (n = 12) 317 ± 35 ms; P < 0.05]. Furthermore, among these patients awakened from stage I, five (26%) did present a ‘sleep pattern’ ERP. Even if these five patients were removed from the analysis, a significant difference between the two groups remained [patients (n = 14) 351.2 ± 42.4 ms versus controls (n = 12) 317 ± 35 ms; P < 0.05]. Thus, ERPs in patients and controls during forced awakening remained significantly different even when strictly matched according to the pre-awakening spectral power EEG.

image

Figure 11. Latency of P300 at Cz in controls and excessive daytime sleepiness (EDS) patients during wakefulness (left) and forced awakening (middle) occurring in a same sleep stage (stage I), as confirmed by similar spectral analysis during the 10 s that preceded forced awakening (alpha, theta and delta power, right). The latency of P300 on forced awakening was significantly delayed in EDS patients, even when awakened from a similar sleep stage as controls.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

The principal result of the present study is that the forced awakening test appeared as an adequate tool to reveal and quantify electrophysiological signs of acute sleep inertia in patients complaining of EDS. Sleep inertia has been defined as ‘the ubiquitous phenomenon of cognitive performance impairment, grogginess and tendency to return to sleep immediately after awakening’ (Van Dongen et al., 2001; see also Dinges, 1990), and is generally considered to be ‘acute’ when evidenced during the few minutes that is immediately follow abrupt awakening (Bruck and Pisani, 1999; Ferrara and De Gennaro, 2000). Several studies have shown that the greatest cognitive impairment occurs within 5–10 min following sudden arousal (Bruck and Pisani, 1999; Dinges, 1990; Ferrara et al., 2000; Naitoh et al., 1993), which include the time-course of ERP recordings in this study. Although sleep inertia has most often been assessed in the context of relatively long sleep periods, from 30 min to several hours (Ferrara et al., 2000, 2001; Jewett et al., 1999; Van Dongen et al., 2001), our results show that signs of acute sleep inertia, reflected by significant changes in cognitive ERPs, may be observed even after a short nap lasting less than 5 min. Furthermore, such signs were qualitatively different, and significantly more pronounced in EDS patients than in moderately (2 h) sleep-deprived control subjects. In what follows, we shall discuss in detail (a) the differential effects observed in controls and patients, (b) the relationship between sleep inertia and sleep propensity, and (c) the possible effects of sleep depth prior to awakening in ERP and behavioural changes.

ERP signs of acute sleep inertia in healthy subjects

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

In control subjects, sudden awakening after 3 min of sleep induced significant amplitude drop of the P300 (or P3) wave, with only a sub-significant trend towards longer latencies, and no effect on target detection performances. The amplitude of P300 is generally considered as an index of the capacity to process task-relevant stimuli (Wickens et al., 1984; reviews in Hansenne, 2000a,b; Kok, 1997; Picton, 1992). As P3 amplitude is also positively related to stimulus-induced arousal (Bahramali et al., 1997; Bastuji and García-Larrea, 1999), P3 attenuation upon forced awakening may reflect decreased allocation of attention to target because of inertia-related loss of phasic arousal. P300 attenuation without much latency increase is typically seen in the context of ‘dual task’ paradigms, where attention is distributed among several sources of demands, and where the amplitude of the P3 evoked by one task decreases as the difficulty of the second task increases (Isreal et al., 1980; Sirevaag et al., 1989; Wickens et al., 1983, 1984). In our subjects, this sign of decline in attentional capacities appeared in the absence of any ‘classical’ ERP sign of somnolence, notably in the preceding N1 component which remained normal. Thus, although N1 and P2 changes have been proposed as the earliest ERP signs of decreased vigilance in humans (de Lugt et al., 1996; Ogilvie et al., 1991), and were shown to reflect sleep inertia following 2 h of sleep (Ferrara et al., 2001), our results suggest that P3 attenuation, which can be observed after a nap of 3–4 min only, may represent a subtler sign of altered stimulus processing in healthy subjects. In favour of this stands the fact that N1-P2 changes have been associated with an increment of detection errors (Ogilvie et al., 1991), while P3 decrease still allowed quite accurate counting performances in our subjects (see Table 1). In the same line, Broughton et al. (1988) reported that the oddball P3 recorded in narcoleptics previous to a diurnal nap was significantly attenuated in the presence of correct performance, whereas N1 remained stable. P3 amplitude drop, as compared with N1-P2 changes, may therefore reflect a slimmer decrease in the level of alertness that remains compatible with the correct performance of a simple detection task.

Signs of sleep inertia in EDS patients

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

Both ERPs and behaviour on forced awakening changed drastically in patients, as compared with healthy, sleep-deprived subjects. On quantitative grounds, there was a significant latency delay of the N2 and P3 components, as well as a significant increase in error rates on both episodes of forced awakening. Qualitatively, almost half of the patients (52%) exhibited morphological changes in ERPs, which shifted from their typical ‘waking pattern’ (N1-P2-N2-P3) to a complex waveform including one or two high-amplitude ‘sleep negativities’. In healthy subjects, this polyphasic pattern of response has been described exclusively during sleep stages I and II, and convergent results from different laboratories indicate that it corresponds to K-complexes or vertex sharp waves evoked by sensory stimulation (Bastuji et al., 1995; Colrain et al., 2000a; Harsh et al., 1994; Niiyama et al., 1996; Ujszászi and Halász, 1986, 1988). Patients who exhibited sleep negativities had also the lowest performances in target detection. Ogilvie et al. (1991) have reported a close relationship between the decrease of behavioural responses during sleep onset and the emergence of ‘sleep negativities’ (N350) to auditory stimuli. The similarity between their results and ours supports the view that this ERP pattern reflects a persisting sleep state during part of the recording session, and is therefore a marker of increased sleep inertia (‘struggle to move from sleep toward alertness’) in this group, as compared with both control subjects and the other group of EDS patients who did not present this pattern along the test.

Patients with ‘sleep type’ ERPs who succeeded at sleeping twice tended to present a single, early sleep negativity (N350), on the first forced awakening, and either a double or a late negativity (N550) during the second awakening session (see Fig. 5). While the isolated sleep N350 is known to correspond to a vertex sharp wave (Colrain et al., 2000a; Näätänen and Picton, 1987), the late N550 matches up with the second portion of the K-complex (Colrain et al., 1999; Cotéet al., 1999; Halász, 1998; Ujszászi and Halász, 1986, 1988). The presence of this latter pattern was associated with worse behavioural responses than the former (‘vertex sharp wave’) configuration, suggesting that patients were probably awakened from a deeper sleep stage in the second than in the first forced awakening session (Colrain et al., 2000b; Ogilvie et al., 1989). This was in turn is probably favoured by the mediocrity of awakening in the first condition. As the presence of ‘sleep negativities’ was also associated with shorter sleep latencies, and with increased capability to have a second nap, these results suggest that the ‘sleep ERP pattern’ was associated with increased sleep pressure in this group of patients.

A P3 wave following N350 was commonly observed in ‘sleep pattern’ ERPs (see Figs 2c, 6b and 7). This component was significantly delayed and decreased relative to waking P300, as it has been previously described during sleep stages 1 and 2 in control subjects (see review in Bastuji and García-Larrea, 1999). The very important latency jitter of this positive wave across patients entails a drastic amplitude drop in grand averages, relative to its amplitude in individual traces. This is illustrated in Fig. 7, showing six individual examples of ‘sleep pattern’ ERPs where the ‘forced awakening-P3’ latency varies between 330 and 450 ms, making it virtually disappear in the grand averages (Fig. 5), but remains of sizeable amplitude in histograms of Fig. 4 (calculated after peak individual measurements). The question whether the ‘forced awakening-P3’ wave recorded in sleep pattern ERPs is or not a functional equivalent of the waking P300 remains a subject of controversy. On the one hand, the association between poor recall of targets and this delayed and small positivity, which may also be recorded in the absence of any cognitive task during stage II, strongly favours the hypothesis that this sleep P3 is not functionally equivalent to P300 as regards to consciousness of the stimulus. On the other hand, the fact that this component has been shown, under certain circumstances, to be sensitive to stimulus significance (Perrin et al., 1999) suggests that this sleep P3, distorted by the superimposed sleep negativities, could sometimes reflect a functional mechanism close to that of waking P300. In any case, it seems highly hazardous at this state of knowledge to consider this P3 as a true P300 (in the sense of waking P300), and to suggest that the mere presence of a P3, preceded and followed by sleep negativities, should reflect conscious stimulus detection.

In odd-ball paradigms, the detection of target tones is commonly assessed using either button pressing or mental counting. With the button pressing technique, evaluation of performance is more precise and avoids any approximate response, while mental counting implies some short-term memory process. In this study, we chose the counting method, because the instruction to press a button might have hampered sleep onset and the movement by itself could have arouse/awaken the patient, while what we sought to test was precisely the extent of sleep inertia. However, the possibility that some subjects could have learnt the approximate count from the waking condition, and then extrapolate it to the forced awakening test, cannot be excluded. This may have been the case of patients in whom sleep negativities were associated with (surprisingly) relatively preserved performances (see Figs 7e and 9), and who, confronted to the impossibility to produce a correct count, could have simply declared the number they had obtained previous to the nap. A further study using a button-pressing procedure would be indubitably useful to analyse with greater precision the relationship between ERP morphology and performances during the process of forced awakening.

The second sub-group of patients had preserved ERP morphology, but significant ERP delay on forced awakening. P3 latency exceeded 99.9% of normal confidence limits in 16% of them, in whom it was considered abnormal on single subject analysis. Latency increase of the P3 complex is thought to reflect a delay in the evaluation of the target stimulus, either because of increased difficulty of the task or cognitive impairment (Goodin and Aminoff, 1984; Picton, 1992; Polich et al., 1990). Persisting sleepiness in our suddenly awakened patients likely increased the difficulty of the discrimination task, as judged by the impairment of their performances. As P3 delay has been described in healthy subjects during sleep stage I (Bastuji et al., 1995; Gora et al., 1999; Harsh et al., 1994; Ogilvie et al., 1991), the latency increase of the component in our patients might reflect the persistence of a state close to stage I as a result of stronger sleep inertia than in controls. The pattern of abnormality in these EDS patients did not, however, simply mimic ‘normal’ sleepiness, as the P3, albeit delayed, remained of normal amplitude, contrary to what was is commonly seen in normal controls prior to sleep onset (Aguirre and Broughton, 1987; Bastuji et al., 1995; Harsh and Badia, 1989; Ogilvie et al., 1991; Sallinen and Lyytinen, 1997). We cannot unequivocally explain this result, especially as P3 was somewhat, albeit insignificantly, reduced. One well-known factor that increases P3 amplitude significantly is the subjects’ motivation to the test (Carrillo de la Peña and Cadaveira, 2000; Johnson, 1986). A strong motivational effect can be suspected in our patients, who wished to do their best during the test, and may have counterbalanced in part the tendency towards P3 amplitude decrease related with somnolence.

In this group of patients, although there was no overall decrease of P300 amplitude, the scalp distribution of this component varied on forced awakening, its relative decrease being maximal over frontal areas. Similar topographic P3 changes have been reported in healthy subjects during somnolence (Bastuji et al., 1995), sleep stage I (Cotéet al., 2002; Gora et al., 1999), and paradoxical sleep (Bastuji et al., 1995; Perrin et al., 1999), and may reflect modifications of the relative strength of different generators to the P300, those projecting to frontal areas being particularly sensitive to vigilance decrease. Although comparisons across different techniques are obviously delicate, it is noteworthy that a similar ‘sleep-related frontal deactivation’ has also been recently postulated on the basis of position effect tomographic-scan studies (Maquet et al., 1996).

c) Sleep depth and sleep inertia

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

As compared with control subjects, sleep latency was significantly decreased in EDS patients, 70% of whom reached stage II during their nap, versus only 33% of controls. One may therefore consider that the enhanced signs of sleep inertia on forced awakening (presence of sleep negativities and/or P3 latency increase) might simply have been due to the patients being awakened, on the average, from a deeper sleep than controls. Indeed, sleep depth previous to awakening is one major determinant of subsequent sleep inertia, as assessed both with behavioural (Ferrara et al., 2000; Mullington and Broughton, 1994) and EEG measures (Ferrara et al., 2001). We therefore tested whether ERP/behavioural signs of sleep inertia could also be observed in a subset of patients matched to controls with respect to sleep depth prior to forced awakening. As illustrated in Fig. 11, both behavioural and ERP results under these strict conditions remained significantly altered in EDS patients, who showed P3 latency delay and/or replacement by sleep negativities. Thus, the enhanced signs of sleep inertia in patients do not appear to merely reflect a ‘normal’ awakening from a deeper sleep than controls, but, on the contrary, reveal a genuine ‘difficulty for arousal’ with tendency to return to sleep, which perturbed behavioural and electrophysiological markers even when the sleep level before awakening was matched with that of controls. If confirmed, this may have important implications, as it suggests that enhanced inertia per se might be a distinctive component of syndromes leading to excessive daytime sleepiness. Even if sleep inertia and sleep propensity were generally associated in our patients, dissociations were also observed, as for instance in 26% of patients awakened from superficial (stage I) sleep, who had ‘sleep negativities’ on forced awakening. Possible dissociation between these two axes suggests that a better diagnostic classification and assessment of EDS patients may be achieved if sleep inertia evaluation is added to that of sleep propensity.

d) Time course of sleep inertia in patients and controls

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

ERP signs of sleep inertia persisted 5 min after the nap's end in both groups of patients, and were specially pronounced in patients with ERP sleep negativities. This differs from results in controls, whose ERPs invariably returned to normal 5 min after the end of the test. Behavioural studies have shown signs of sleep inertia, reflected by performance decrement, lasting minutes to hours after morning awakening in non-sleep-deprived healthy subjects (Achermann et al., 1995; Dinges, 1989; Jewett et al., 1999). In a previous study, we observed that morning awakening after a whole night of sleep was associated in healthy subjects with a progressive normalization of P3 latencies, from stage I values down to wakefulness norms, over a period that could last over 10 min (Bastuji et al., 1995). ERP signs of sleep inertia after a very short afternoon nap in our EDS patients were therefore comparable in intensity to those observed in healthy controls after much longer naps, or even all-night sleep. This might help to explain in part the cognitive difficulties encountered by these patients, when they just get out of an involuntary short sleep attack (Broughton, 1989; Mullington and Broughton, 1994).

The abnormalities observed in EDS patients during forced awakening contrasted with recordings in wakefulness, where their P300 had similar amplitude and latency as that of control subjects in the same condition. Previous studies conducted during the waking state on narcolepsy and idiopathic hypersomnia patients (which formed 65% of our sample) have also reported normal auditory P3 latencies (Broughton et al., 1988; Sangal and Sangal, 1995), although Broughton et al. (1988) reported decreased amplitudes. This underscores that these patients do not suffer from gross neurocognitive disabilities in the waking state, and that the abnormalities reported herein are not the result of a background cognitive impairment.

Conclusion

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References

The forced awakening test appeared as a robust and relatively rapid tool to evaluate simultaneously, during daytime, three variables that may be relevant for EDS diagnosis and management, namely (a) sleep propensity (sleep latency), (b) sleep structure during a very short nap, and (c) acute sleep inertia. It was much more discriminative than standard waking ERPs to demonstrate abnormalities on patients complaining of excessive daytime sleepiness. Abnormal responses to the forced awakening test were more prevalent in patients with narcolepsy and idiopathic hypersomnia than in patients with EDS associated with psychological or psychiatric disorders. The results seem to warrant the use of the forced awakening test in the evaluation of selected patients suspected of these conditions; further research is needed, however, to evaluate the putative place of this test in patients with sleep apnoea syndrome.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Subjects and Methods
  5. Patients
  6. Control subjects
  7. Stimulation and recording
  8. Experimental procedure
  9. Data analysis
  10. Criteria for ERP components
  11. Other variables
  12. Statistical analysis
  13. Results
  14. Behavioural data
  15. ERPs during wakefulness
  16. Event-related potentials on forced awakening
  17. Control subjects ()
  18. EDS patients
  19. Sleep latency, sleep stability and errors in patients
  20. ERP patterns on forced awakening and aetiology of EDS
  21. Pre-awakening spectral analysis and comparison of P3 latency on forced awakening in patients/controls
  22. Discussion
  23. ERP signs of acute sleep inertia in healthy subjects
  24. Signs of sleep inertia in EDS patients
  25. c) Sleep depth and sleep inertia
  26. d) Time course of sleep inertia in patients and controls
  27. Conclusion
  28. Acknowledgements
  29. References
  • Achermann, P., Werth, E., Dijk, D. J. and Borbely, A. A. Time course of sleep inertia after nighttime and daytime sleep episodes. Arch. Ital. Biol., 1995, 134: 109119.
  • Aguirre, M. and Broughton, R. Complex event-related potentials (P300 and CNV) and MSLT in the assessment of excessive sleepiness in narcolepsy-cataplexy. Electroenceph. Clin. Neurophysiol., 1987, 68: 298316.
  • Bahramali, H., Gordon, E., Lim, C. L., Li, W., Lagopoulos, J., Leslie, J., Rennie, C. and Meares, R. A. Evoked related potentials associated with and without an orienting reflex. Neuroreport, 1997, 8: 26652669.
  • Balkin, T. J. and Badia, P. Relationship between sleep inertia and sleepiness: cumulative effects of four nights of sleep disruption/restriction on performance following abrupt nocturnal awakenings. Biol. Psychol., 1988, 27: 245258.
  • Bastuji, H. and García-Larrea, L. Evoked potentials as a tool for the investigation of human sleep. Sleep Med. Rev., 1999, 3: 2345.
  • Bastuji, H., García-Larrea, L., Franc, C. and Mauguière, F. Brain processing of stimulus deviance during slow-wave and paradoxical sleep. A study of human auditory evoked responses using the oddball paradigm. J. Clin. Neurophysiol., 1995, 12: 155167.
  • Billiard, M. Competition between the two types of sleep, and the recuperative function of REM sleep versus NREM sleep in narcoleptics. In: C.Guilleminault, W. C.Dement and P.Passouant (Eds) Narcolepsy. Advances in Sleep Research, vol. 3, Series editor. Spectrum Publication, New York, 1976: 7796.
  • Broughton, R. J. Sleep attacks, naps, and sleepiness in medical sleep disorders. In: D. F.Dinges and R. J.Broughton (Eds) Sleep and Alertness: Chronobiological, Behavioral and Medical Aspects of Napping. Raven Press, New York, 1989: 267298.
  • Broughton, R., Aguirre, M. and Dunham, W. A comparison of multiple and single sleep latency and cerebral evoked potential (P300) measures in the assessment of excessive daytime sleepiness in narcolepsy-cataplexy. Sleep, 1988, 11: 537545.
  • Bruck, D. and Pisani, D.L. The effects of sleep inertia on decision-making performance. J. Sleep Res., 1999, 8: 95103.
  • Campbell, K.B. Information processing during sleep onset and sleep. Can. J. Exp. Psychol., 2000, 54: 209229.
  • Carrillo de la Peña, M. T. and Cadaveira, F. Influence of motivational instruction on P3 amplitude. Neurophysiol. Clin., 2000, 30: 232239.
  • Carskadon, M. A., Dement, W. C., Mitler, M. M., Roth, T., Westbrook, P. R. and Keenan, S. Guidelines for the Multiple Sleep Latency Test (MLST): a standard measure of sleepiness. Sleep, 1986, 9: 519524.
  • Chugh, D. K., Weaver, T. E. and Dinges, D. E. Neurobehavioral consequences of arousals. Sleep, 1996, 19: S198S201.
  • Civil, R. H., Gorman, S. H., Stong, D. B. and King, S. P. Safety, tolerance and pharmacokinetics of single doses of modafinil in healthy subjects. APSS 9th Annual Meeting, Nashville, TN, 1995, A097: 49.
  • Colrain, I. M., Webster, K. E. and Hirst, G. The N550 component of the evoked K-complex: a modality non-specific response? J. Sleep Res., 1999, 8: 273280.
  • Colrain, I. M., Webster, K. E., Hirst, G. and Campbell, K. B. The role of vertex sharp waves and K-complexes in the generation of N300 in auditory and respiratory-related evoked potentials during early stage 2 NREM sleep. Sleep, 2000a, 23: 97106.
  • Colrain, I. M., Di Parsia, P. and Gora, J. The impact of prestimulus EEG frequency on auditory evoked potentials during sleep onset. Can. J. Exp. Psychol., 2000b, 54: 243254.
  • Coté, K. A., De Lugt, D. R., Langley, S. D. and Campbell, K. B. Scalp topography of the auditory evoked K-complex in stage 2 and slow wave sleep. J. Sleep Res., 1999, 8: 263272.
  • Coté, K. A., De Lugt, D. R. and Campbell, K. B. Changes in the scalp topography of event-related potentials and behavioral responses during the sleep onset period. Psychophysiology, 2002, 39: 2937.
  • Dement, W. C. and Mitler, M. M. It's time to wake up to the importance of sleep disorders. JAMA, 1993, 269: 15481549.
  • Dinges, D. F. Napping patterns and effects in human adults. In: D. F.Dinges and R. J.Broughton (Eds) Sleep and Alertness: Chronobiological, Behavioral and Medical Aspects of Napping. Raven Press, New York, 1989: 171204.
  • Dinges, D. F. Are you awake? Cognitive performance and reverie during the hypnopompic state. In: R. R.Bootzin, J. F.Kihlstrom and D.Schacter (Eds) Sleep and Cognition. American Psychological Association, Washington, DC, 1990: 159175.
  • Ferrara, M. and De Gennaro, L. The sleep inertia phenomenon during the sleep-wake transition: theoretical and operational issues. Aviat. Space Environ. Med., 2000, 71: 843848.
  • Ferrara, M., De Gennaro, L. and Bertini, M. Time-course of sleep inertia upon awakening from nighttime sleep with different sleep homeostasis. Aviat. Space Environ. Med., 2000, 71: 225229.
  • Ferrara, M., De Gennaro, L., Ferlazzo, F., Curcio, G., Barattucci, M. and Bertini, M. Auditory evoked responses upon awakening from sleep in human subjects. Neurosci. Lett., 2001, 310: 1458.
  • Godbout, R. and Montplaisir, J. All-day performance in narcoleptic subjects. Sleep, 1986, 9: 200204.
  • Goodin, D. and Aminoff, M.J. The relationship between the evoked potential and brain events in sensory discrimination and motor response. Brain, 1984, 107: 241251.
  • Goodin, D., Desmedt, J., Maurer, K. and Nuwer, M. R. IFNC recommended standards for long latency auditory event-related potentials. Rep. IFNC Committee, 1994, 91: 1820.
  • Gora, J., Colrain, I. M. and Trinder, J. Respiratory-related evoked potentials during the transition from alpha to theta EEG activity in stage 1 NREM sleep. J. Sleep Res., 1999, 8: 123134.
  • Gora, J., Colrain, I. M. and Trinder, J. The investigation of K-complex and vertex sharp wave activity in response to mid-inspiratory occlusions and complete obstructions to breathing during NREM sleep. Sleep, 2001, 24: 8189.
  • Halász, P. Hierarchy of micro-arousals and the microstructure of sleep. Neurophysiol. Clin., 1998, 28: 461475.
  • Hansenne, M. The P300 cognitive event-related potential. I. Theoretical and psychobiologic perspectives. Neurophysiol. Clin., 2000a, 30: 191210 (French).
  • Hansenne, M. The P300 cognitive event-related potential. II. Individual variability and clinical application in psychopathology. Neurophysiol. Clin., 2000b, 30: 211231 (French).
  • Harrison, Y. and Horne, J. A. Long-term extension to sleep – Are we really chronically sleep deprived? Psychophysiology, 1996, 33: 2230.
  • Harrison, Y., Bright, V. and Horne, J. A. Can normal subjects be motivated to fall asleep faster? Physiol. Behav., 1996, 60: 681684.
  • Harsh, J. and Badia, P. Auditory evoked potentials as a function of sleep deprivation. Work and stress, 1989, 3: 7991.
  • Harsh, J., Voss, U., Hull, J., Schrepper, S. and Badia, P. ERP and behavioral during the wake/sleep transition. Psychophysiology, 1994, 31: 244252.
  • Heinze, H. J., Münte, T. F., Kutas, M., Butler, S. R., Näätänen, R. and Goodin, D. S. Cognitive event-related potentials. In: G.Deuschl and A.Eisen (Eds) Recommendations for the Practice of Clinical Neurophysiology: Guidelines of the International Federation of Clinical Neurophysiology. Electroenceph. Clin. Neurophysiol., 1999, Suppl. 52 : 9195.
  • ICSD (International Classification of Sleep Disorders). Diagnostic and Coding Manuel. Diagnostic Classification Steering Committee, Thorpy MJ, Chairman. Rochester, Minnesota: American Sleep Disorders Association, 1990.
  • Isreal, J. B., Chesney, G. B., Wickens, C. D. and Donchin, E. P300 and tracking difficulty: evidence for multiple resources in dual-task performance. Psychophysiology, 1980, 17: 259273.
  • Jasper, H.H. The ten–twenty electrode system of the International Federation. Electroenceph. Clin. Neurophysiol., 1958, 10: 371375.
  • Jewett, M. E., Wyatt, J. K., Ritz-De Cecco, A., Khalsa, S. B., Dijk, D. J. and Czeisler, C. A. Time course of sleep inertia dissipation in human performance and alertness. J. Sleep Res., 1999, 8: 18.
  • Johnson, R. A triarchic model of P300 amplitude. Psychophysiology, 1986, 23: 367384.
  • Kok, A. Event-related-potential (ERP) reflections of mental resources: a review and synthesis. Biol. Psychol., 1997, 45: 1956.
  • Lubin, A., Hord, D., Tracy, M. L. and Johnson, L. C. Effects of exercise, bedrest and napping on performance decrement during 40 hours. Psychophysiology, 1976, 13: 334339.
  • de Lugt, D. R., Loewy, D. H. and Campbell, K. B. The effect of sleep onset on event related potentials with rapid rates of stimulus presentation. Electroenceph. Clin. Neurophysiol., 1996, 98: 484492.
  • Lyznicki, J. M., Doege, T. C., Davis, R. M. and Williams, M. A. Sleepiness, driving and motor vehicle crashes. JAMA, 1998, 279: 19081913.
  • Maquet, P., Peters, J. M., Aerts, J., Delfiore, G., Degueldre, C., Luxen, A. and Franck, G. Functional neuroanatomy of human rapid eye movement sleep and dreaming. Nature, 1996, 383: 163166.
  • Mullington, J. and Broughton, R. Daytime sleep inertia in narcolepsy-cataplexy. Sleep, 1994, 17: 6976.
  • Naitoh, P., Kelly, T. and Babkoff, H. Sleep inertia: best time not to wake up? Chronobiol. Int., 1993, 10: 109118.
  • Nielsen-Bohlman, L., Knight, R. T., Woods, D. L. and Woodward, K. Differential auditory processing continues during sleep. Electroenceph. Clin. Neurophysiol., 1991, 79: 281290.
  • Niiyama, Y., Fujiwara, R., Satoh, N. and Hishikawa, Y. Endogenous components of event-related potential appearing during NREM stage 1 and REM sleep in man. Intern. J. Psychophysiol., 1994, 17: 165174
  • Niiyama, Y., Satoh, N., Kutsuzawa, O. and Hishikawa, Y. Electrophysiological evidence suggesting that sensory stimuli of unknown origin induce spontaneous K-complexes. Electroenceph. Clin. Neurophysiol., 1996, 98: 394400.
  • Näätänen, R. and Picton, T. The N1 wave of the human electric and magnetic response to sound: a review and an analysis of the component structure. Psychophysiology, 1987, 24: 375425.
  • Ogilvie, R. D., Wilkinson, R. T. and Allison, S. The detection of sleep onset: behavioral, physiological, and subjective convergence. Sleep, 1989, 12: 458474.
  • Ogilvie, R. D., Simons, I. A., Kuderian, R. H., MacDonald, T. and Rustenberg, J. Behavioral, event-related potential, and EEG/FFT changes at sleep onset. Psychophysiology, 1991, 28: 5464.
  • Perrin, F., Garcia Larrea, L., Mauguière, F. and Bastuji, H. A differential brain response to the subject's own name persists during sleep. Clin. Neurophysiol., 1999, 110: 21532164.
  • Picton, T. W. The P300 wave of the human event-related potential. J. Clin. Neurophysiol., 1992, 9: 456475.
  • Polich, J., Ladish, C. and Bloom, F. E. P300 assessment of early Alzheimer's disease. Electroenceph. Clin. Neurophysiol., 1990, 77: 179189.
  • Rechtschaffen, A. and Kales, A. A Manual of Standardised Terminology, Techniques and Scoring for Sleep Stages of Human Subjects. US Government printing office, Washington DC (NIH Publication No. 204), 1968.
  • Richardson, G. S., Carskadon, M., Flagg, W., van den Hoed, J., Dement, W. C. and Mitler, M. M. Excessive daytime sleepiness in man: multiple sleep latency measurements in narcoleptic and controls subjects. Electroenceph. Clin. Neurophysiol., 1978, 45: 621627.
  • Salisbury, D., Squires, N. K., Ibel, S. and Maloney, T. Event-related potentials during Stage 2 NREM sleep in Humans. J. Sleep Res., 1992, 1: 251257.
  • Sallinen, M. and Lyytinen, H. Mismatch negativity during objective and subjective sleepiness. Psychophysiology, 1997, 34: 694702.
  • Sangal, R.B. and Sangal, J. M. P300 latency: abnormal in sleep apnea with somnolence and idiopathic hypersomnia, but normal in narcolepsy. Clin. Electroencephal., 1995, 26: 146153.
  • Sirevaag, E. J., Kramer, A. F., Coles, M. G. and Donchin, E. Resource reciprocity: an event-related brain potentials analysis. Acta Psychol., 1989, 70: 7797.
  • Steinmetz, H., Fürst, G. and Meyer. Craniocerebral topography within the international 10–20 system. Electroenceph. Clin. Neurophysiol., 1989, 72: 499506.
  • Ujszászi, J. and Halász, P. Late component variants of single auditory evoked responses during NREM sleep stage 2 in man. Electroenceph. Clin. Neurophysiol., 1986, 64: 260268.
  • Ujszászi, J. and Halász, P. Long latency evoked potential components in human slow wave sleep. Electroenceph. Clin. Neurophysiol., 1988, 69: 516522.
  • Valley, V. and Broughton, R. The physiological (EEG) nature of drowsiness and its relation to performance deficits in narcoleptics. Electroenceph. Clin. Neurophysiol., 1983, 55: 243251.
  • Van den Hoed, J., Kraemer, H., Guilleminault, C., Zarcone, V. P., Miles, I. E., Dement, W. C. and Mitler, M. M. Disorders of excessive daytime somnolence: polygraphic and clinical data for 100 patients. Sleep, 1981, 4: 2337.
  • Van Dongen, H. P., Price, N. J., Mullington, J. M., Szuba, M. P., Kapoor, S. C. and Dinges, D. F. Caffeine eliminates psychomotor vigilance deficits from sleep inertia. Sleep, 2001, 24: 813819.
  • Wickens, C., Kramer, A., Vanasse, L. and Donchin, E. Performance of concurrent tasks: a psychophysiological analysis of the reciprocity of information-processing resources. Science, 1983, 221: 10801082.
  • Wickens, C. D., Kramer, A. F. and Donchin, E. The event-related potential as an index of the processing demands of a complex target acquisition task. Ann. N. Y. Acad. Sci., 1984, 425: 295299.