SEARCH

SEARCH BY CITATION

Keywords:

  • eyelid movements;
  • dream recall;
  • pont-geniculo-occipital waves;
  • electroencephalogram arousal

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

The present study aimed to test whether spontaneous eyelid movements (ELMs) during stage 2 and rapid eye movement (REM) sleep are related to more frequent and vivid reports of visual mentation on awakening. Participants were awakened 15 s after an ELM was observed during ongoing REM and stage 2 sleep and immediately asked for a mentation report and to rate the visual vividness of any imagery they could remember. These reports were compared with control reports collected after a period of ELM quiescence before awakening (noELM). Significantly greater frequencies of imagery reports were collected after ELM awakenings compared with noELM awakenings from stage 2, but not REM sleep. When imagery was reported, imagery ratings were not significantly different between ELM and noELM conditions, regardless of sleep stage. The average amount of electroencephalogram (EEG) arousal 15 s after stage 2 awakenings was significantly higher in the ELM compared with noELM conditions. In addition, within the stage 2 ELM condition, EEG arousal was significantly higher when visual imagery was reported compared with reports without imagery; suggesting that the observed increase in imagery reporting from the stage 2 ELM condition could have been mediated by the level of brain arousal. Such arousal possibly provides better conditions to attend and recall previous mental activity from NREM sleep. However, there was no ELM/arousal effect within REM sleep, possibly because this state is already at maximum sleeping levels of arousal, attention and resulting dream recall.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

Almost 40 years ago, a slow corrosion of the once-held belief that rapid eye movement (REM) sleep was the exclusive domain of dreaming began (Foulkes, 1967). Based on the acceptance that dreaming was not exclusive to REM sleep, researchers started searching for new ‘phasic’ physiological indicators of dreaming outside of REM sleep, with the hope of finding more reliable neurobiological correlates of dreaming. (For a review see Pivik, 1991, 1994.) More recently, Nielsen (2003) proposed a ‘Covert REM’ model of dreaming, where Covert REM sleep is a period of NREM sleep for which some REM sleep processes are present, but for which REM sleep cannot be scored with standard criteria. This has now led to resurgence in the search for such ‘Covert REM’ or phasic sleep measures.

The ponto-geniculo-occipital (PGO) wave is a physiological event occurring during sleep that is controversially claimed to form the physiological basis of visual imagery in models such the activation synthesis hypothesis and the activation, input and modulation model of dreaming (Hobson and McCarley, 1977; Hobson et al., 2003). Although PGO waves occur primarily in REM sleep, they also appear as isolated or short bursts of activity intermittently throughout non-REM sleep (Callaway et al., 1987). Thus, PGO activity remains as a potential physiological marker underlying ‘phasic’ or ‘Covert REM’ sleep processes. In order to adequately test PGO models of dreaming, the relationship between some measure of PGO activity in humans and dream reporting needs to be investigated.

At present, it is not possible to directly observe PGO activity in humans, as PGO waves can only be studied using indwelling recording electrodes. PGO activity has been found to correlate well with eye movements (EMs) in cats (Callaway et al., 1987) and has also been postulated to correlate well with EM activity in humans (McCarley et al., 1983). However, as EMs rarely occur during NREM sleep, this measure cannot provide a useful PGO correlate during this sleep state (Pivik, 1991). Thus, other hypothesized PGO analogues such as periorbital integrated potentials (PIPs; Bliwise and Rechtschaffen, 1978) and middle ear muscle activity (MEMA; Ogilvie et al., 1982), have been investigated. However, reviews by Pivik (1991, 1994 and Rechtschaffen (1973) critically examining data regarding PIP and MEMA as indicators of sleep mentation have argued that these PGO analogues have failed to provide a strong relationship to dream mentation recalled from sleep.

In mammals, extraocular muscles consist of two types of skeletal fibers: a slow contracting form and a fast ‘twitch’ form (Hess and Pilar, 1963; Campbell et al., 1999). When early investigators of PGO analogues measured PIPs, slow frequencies were filtered out (Rechtschaffen, 1973). This was carried out to capture the fast ’twitch’ activity of the extraocular muscles. However, such measures of highly amplified and integrated signals from EOG electrodes with slow frequencies (<10 Hz) filtered out, also obviously recorded electrical activity from levator palpebrae and orbicularis eyelid muscles along with REMs. At the time, eye blinks were regarded as an orbicularis muscle component ’confounding’ the measure (Rechtschaffen, 1973). It is most likely that what these researchers measured as PIPs was extraocular EM muscle activity mixed with a combination of eyelid blinks and twitches. Ironically, the electric measurement of PIPs is not too dissimilar to the methods used in modern studies of eye-blink reflex. Although the placement of electrodes differs slightly, the fundamental measure of a highly amplified integrated signal taken across the eye musculature, remains the same (Filion et al., 1998). Thus, there is considerable evidence to suggest that a component of the correspondence of PIPs to PGO activity in human sleep could have been attributed to eyelid activity generated by levator palpebrae superioris musculature (Conduit et al., 2002).

Early research investigating the elicitation of an electrically induced blink reflex in humans during sleep, found spontaneous eye-twitch activity occurring during REM sleep (Ferrari and Messina, 1972). Soon after, Orem and Dement (1974) investigated spontaneous eyelid behavior in sleeping cats and its relationship to PGO activity measured at the laternal geniculate nucleus. Phasic eyelid twitches were observed in every REM period. These twitches appeared throughout REM, with 80–90% associated with EMs. Twenty-eight REM periods were analyzed in three cats to determine the relationship between lid twitches and PGO waves. A positive case was defined as the presence of PGO spikes either 1 s before or after the onset of a twitch. Using this criterion, an average of 89% of twitches was related to PGO waves.

Bowker & Morrison (1976) later found that PGO waves could be elicited by presenting startling tones during sleep. As the intensity of tones was increased, eye-blinks, neck electromyogram (EMG) twitches, body twitches, electroencephalogram (EEG) desynchronization and arousal were also induced with PGO waves. From this initial work, extensive research was conducted investigating the proposal that PGO activity is related to spontaneous ‘alerting’ in both awake and sleeping animals (Bowker & Morrison, 1976; Sanford et al., 1992, 1993, 1994; Hunt et al., 1998). As eye-blinks are a widely used and accepted measure of the startle response in awake human subjects (Filion et al., 1998), this provides evidence to suggest that ELMs measured during sleep might also occur with spontaneous alerting and thus be coincident with PGO activity in humans.

Currently it is proposed that spontaneous eye lid movements (ELMs) in sleeping subjects are related to arousal, and decreased ELMs in awake subjects are related to decreased vigilance and sleep onset (Cantero et al., 2002). Cantero et al. (2002) have proposed that activity of the upper eyelid may be related to the activity of the reticular formation. This relationship is based on the reasoning that the upper eyelid is contracted by the levator superioris, which is innervated by the oculomotor nucleus, which, in turn is innervated from the midbrain reticular system and the pontine reticular formation (Spencer and McNeer, 1991). However, this is also the same pathway that is implicated in the activation of rectus muscles in the generation of eye movements during waking (Spencer and McNeer, 1991) and presumably REM sleep (Hobson and McCarley, 1977).

Rectus extraocular muscles have previously been shown to be closely related to PGO activity in cats and rats (Rechtschaffen et al., 1972; Fredrickson et al., 1972). In addition, the levator superioris is an extraocular muscle innervated by similar oculomotor pathways to that of the superior rectus muscle (Spencer and McNeer, 1991). Together, these findings also suggest that an external measure of upper eyelid activity might provide a useful indicator of PGO activity in humans. If this were the case, it would be expected that ELMs could occur during REM sleep. Hobson and colleagues (Hobson and Stickgold, 1994; Ajilore et al., 1995; Stickgold and Hobson, 1994; Pace-Schott et al., 1994; Cantero et al., 2002; Rowley et al., 1998) have monitored both ELMs and REMs using a small adhesive-backed piezo-electric film placed directly on the eyelid. However, if one is interested in whether ELMs occur during REM, this data then becomes problematic because differentiation of ELMs and REMs during REM sleep using this placement method is technically difficult.

Studies of human sleep from this laboratory have utilized a piezo-ceramic vibration sensor placed at the supraorbital ridge of the orbit (Conduit et al., 2002). Such placement avoids recording artifacts from EMs, and has revealed that ELMs co-occur with other proposed PGO analogues such as PIPs and MEMA at a rate far greater than expected by chance (Conduit et al., 2002). In addition, ELM activity increases in NREM during REM deprivation and increases with subsequent REM rebound (Conduit, 1999; R. Conduit and G. Coleman, submitted for publication) in a manner similar to that observed of PGO activity in animals (Duysan-Peyrethon et al., 1967; Vimont-Vicary et al., 1966). Thus, in light of converging evidence suggesting that ELMs might be related to PGO activity in humans, another convergent approach to testing the PGO hypothesis of dreaming was now possible by examining the relationship between ELM activity and sleep mentation.

The aim of the present study was to investigate whether spontaneous ELMs were related to reports of visual mentation on awakening from sleep in humans. Participants in an experimental condition were awakened 15 s after an ELM from ongoing stage 2 and REM sleep. Corresponding control conditions were also implemented, where a period of at least 90 s of ELM quiescence during stage 2 and REM was observed before awakening subjects for mentation reports.

Overall, it was hypothesized that ELMs, as an indicator of PGO activity, would be related to an increased frequency of visual mentation reports compared with control awakenings void of ELM activity, and these reports would be rated as more perceptually vivid than those from control awakenings.

Participants

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

Twelve female and 12 male subjects participated in the experiment. All participants were normal, healthy adults aged 18–30 years. Participants were told that their sleep patterns and eyelid muscle activity was being monitored, and the aim of the experiment was to investigate the relationship between these measures and dream recall.

Apparatus and Materials

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

Sleep was monitored using Grass model 8–16 polygraph (Grass Instrument Co., Quincy, MA, USA). EEG, EMG, electrooculogram, and ELM made up the recording montage.

EEG placements were made to C4-A1 and C3-A2 according to the international 10–20 placement system (Jasper, 1958). EMG recording electrodes were attached at the chin muscles (mentalis) and under the chin. EOG locations were to the right and left outer canthus (ROC and LOC) and right and left upper sides of the nose (right and left inner cantus; RIC and LIC). Bipolar referencing was used (ROC–RIC and LOC–LIC). ELMs were measured using a piezo-ceramic vibration sensor (model EB-T-320; NTK, Tokyo, Japan), attached with double-sided tape and medical tape just below the eyebrow on the supraorbital ridge of the right orbit (Conduit et al., 2002).

All subjects were situated in a sound-attenuated sleep laboratory. The experimenter conducted all monitoring and awakenings from an adjacent room. There was no visual contact between the subject and experimenter. All communication was via intercom. All mentation reports were tape-recorded.

Subject preparation

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

Participants arrived at the sleep laboratory approximately 1 h prior to their normal sleeping time. After signing a declaration of informed consent, the subjects were then connected for sleep recording. To calibrate ELM, subjects were asked to blink lightly. The ELM sensitivity was then adjusted to ensure the trace showed at least a 1-cm pen deflection. Participants were then asked to look left, right, up and down to ensure that clear EOG traces were present and minimal artifacts were present on the ELM channel (less than 5 mm pen deflection). Once the calibration procedures were completed, subjects were left to sleep.

Sleep scoring and the scoring of ELM events

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

Sleep scoring was carried out manually in 30-s epochs according to standard procedures (Rechtschaffen and Kales, 1968). Previous investigation has shown that the technique of detecting ELMs used in this study typically records small amplitude pulse artifacts at the orbital ridge (Conduit et al., 2002). Therefore, in order to maintain a strictly conservative count of ELM activity, an ELM event had to show a minimum increase in activity of 10 times baseline amplitude (Conduit et al., 2002). This is therefore a minimum of 1 cm or greater pen deflection, which is an eyelid movement approximately equivalent to a blink or greater, as per the calibration procedures. Piezosensor activity less than the set criteria (1 cm or 10× baseline) was not included in the analysis, as it is technically difficult to ensure that it is not the result of artifacts such as eye movement, pulse, or other muscle twitch activities (Conduit et al., 2002).

Awakening procedure

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

Participants were awakened for the collection of mentation reports from two experimental (ELM) conditions and two control (noELM) conditions during stage 2 sleep in an alternating pattern across the night. The order of experimental (ELM) and control (noELM) conditions was counterbalanced across subjects, with half of the subjects having an experimental condition as their first awakening and the other half a control condition as their first awakening. Once the first four stage 2 awakenings were conducted, an additional ELM and noELM awakening was conducted from REM sleep. These were also implemented in a counterbalanced manner across subjects. Only two awakenings occurred in REM sleep, as a total of six awakenings across a single night was considered to be the upper limit of what was practically achievable for all participants. As previous investigations using the same methods of quantifying dream recall have found baseline REM awakenings to be at almost ceiling levels (Conduit et al., 1997), the emphasis in this study was towards investigating stage 2 sleep.

Mentation report collection

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

The experimenter called the participant's name though an intercom situated above the bed, and asked if they were awake. Once the subject responded by stating that they were awake, the experimenter then played a tape-recording of the experimenter's voice, which asked: ‘Could you please describe any thoughts or images that were going through your mind just before I woke you?’ If visual imagery was mentioned, a second tape was then played which then asked: ‘Could you please rate the vividness of this imagery on a scale from 1 to 10, with 1 being ‘least vivid vague or hardly rememberable’, to 10 being ‘‘most vivid like real’’’. This response was recorded as an imagery rating.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

Two ELM and noELM awakenings from stage 2 sleep and one each from REM sleep were achieved for each of the 24 subjects. The average time into each stage awakenings occurred was as follows: stage 2 ELM (mean = 3.6 min, SD = 1.22), stage 2 noELM (mean = 3.3, SD = 0.94), REM ELM (mean = 2.6, SD = 0.79) REM noELM (mean = 2.9, SD = 1.18). There were no significant differences in mean awakening time between REM (t23 = 0.74, P > 0.05) and stage 2 conditions (t23 = 1.05, P > 0.05).

Frequency of imagery reports

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

An independent rater, blind to conditions, was used to judge audio-tape recordings of the 144 mentation reports as either (i) reporting at least one visualizable noun ‘imagery’, or (ii) not reporting a visualizable noun ‘no imagery’. Visualizable nouns were defined as nouns of objects that could be seen in waking life (Antrobus et al., 1991).

The number of visual imagery reports as judged by the rater divided by the number of awakenings for each subject in each condition was then calculated and is presented in Table 1. These results are summarized in graphical form as the percentage of imagery reports per condition in Fig. 1.

Table 1.  The number of reports with visual imagery (as judged by an independent rater, blind to experimental conditions) from two ELM and noELM awakenings from stage 2 sleep and one ELM and noELM awakening from REM sleep, for each participant
Number of mentation reports with imageryNREM sleepREM sleep
ELMNoELMELMNoELM
081444
11192020
251
image

Figure 1. The average percentage of imagery reports from ELM and noELM awakenings from REM and stage 2 sleep (as judged by an independent rater, blind to experimental conditions). Error bars represent standard error of the mean.

Download figure to PowerPoint

Because of the high frequency of ‘no imagery’ or zero responses particularly from stage 2 awakenings, the assumption homogeneity of variance was violated (Fmax4,23 = 3.64, P < 0.05). Therefore, nonparametric statistical tests were adopted (Kirk, 1982). There was an overall difference in the frequency of imagery reports across conditions (Friedman inline image = 39.78, P < 0.001) and between stage 2 and REM awakenings (Wilcoxon Z = 3.99, P < 0.001). The difference in frequency of imagery reports between ELM and noELM conditions within stage 2 sleep was also significant (Wilcoxon Z = 2.24, P < 0.05). There was no difference in average imagery report frequency between ELM and noELM conditions within REM (Wilcoxon Z = 0, P > 0.05). When the definition of a dream report was broadened to include any thought-like mentation (Foulkes, 1967), this added three reports to each ELM and noELM stage 2 condition, and one report to each REM condition, but did not change the over statistical difference in recall frequency (stage 2: Wilcoxon Z = 2.36, P < 0.05; REM: Wilcoxon Z = 0, P > 0.05).

Imagery ratings

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

Table 2 shows the average imagery ratings given by the subjects from sleep reports across all four conditions. When the rater judged a participant report to have ‘no imagery’, the imagery rating associated with this report was excluded from the analysis. This is because the comparison was regarding the visual quality of the imagery between the two conditions. It was not desirable that the vividness ratings be confounded by the frequency of imagery reports.

Table 2.  The average imagery ratings reported from subjects after ELM and noELM awakenings from stage 2 and REM sleep. When subjects did not report imagery from awakenings (as judged by an independent rater), the imagery rating was not included in the calculation of the mean
Sleep stageConditionAverage imagery ratingSDn
Stage 2No ELM4.91.4510
ELM4.51.8216
REMNoELM6.82.2320
ELM7.51.8920

A significant overall difference in imagery ratings between stage 2 (mean = 4.7, SD = 1.61) and REM (mean = 7.0, SD = 1.97) was present (t17 = 4.24, P < 0.01). However, there were no significant differences in the average imagery ratings between ELM and noELM conditions within stage 2 (ELM: mean = 4.9, SD = 1.97, noELM: mean = 4.4, SD = 0.98; t6 = 0.60, P > 0.05) or REM sleep (ELM: mean = 7.7, SD = 1.79, noELM: mean = 7.0, SD = 2.03; t17 = 1.20, P > 0.05). As imagery reports in both ELM and noELM conditions for stage 2 were produced by only seven subjects, it is acknowledged that the statistical insignificance of the t-test comparing these conditions could be the result of inadequacies of sample size rather than reflecting any true state of this relationship.

ELMs and concurrent phasic EM activity

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

As ELMs have been previously shown to occur with other phasic muscle activity (Conduit et al., 2002), ELMs during REM sleep were checked for co-occurrence with phasic eye movements (phasic REM), which has previously been related to increased dream imagery (Pivik, 1991). Twelve of 24 noELM REM awakenings and 19/24 ELM REM awakenings were during phasic REM. Thus, awakenings with ELMs were more likely to be during phasic REM than during tonic REM (79% versus 50%, χ2 = 4.46, P < 0.05). However, phasic REM awakenings were no greater than tonic REM awakenings in average imagery recall (phasic: 11/12 versus tonic: 9/12; Wilcoxon Z = 0.28, P > 0.1) or average imagery ratings (phasic mean = 7.0, SD = 2.42 versus tonic mean = 6.4, SD = 2.12; t18 = 0.67, P > 0.1). For the 19 ELM phasic REM awakenings, 18 had imagery with an average imagery rating of 7.8. Of the five remaining ELM tonic REM awakenings, two had imagery recall and the average imagery rating was 5.0. (Due to the low number of ELM REM tonic EM cases, statistical comparisons were not conducted.)

ELM and EEG arousal

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

Spontaneous arousal and awakening was frequently observed coincident with ELM events, with experimental conditions often not conducted after ELM events because of excessive arousal observed in the subsequent EEG and/or EMG recording. 26% of ELMs during observations from stage 2 sleep and 8% of ELMs during observations from REM sleep occurred with arousals according to Rechtschaffen and Kales (1978) criteria (>10 s alpha or higher frequency EEG). Of all the ELM events that were chosen for the ELM awakening condition, none were associated with enough EEG arousal to be considered an awakening according to Rechtschaffen and Kales (1978) criteria. Table 3 shows the average amount of manually scored EEG arousal time (alpha or higher frequency EEG) during the 15 s preceding awakening from ELM and noELM conditions in both stage 2 and REM sleep.

Table 3.  The average amount of manually scored alpha or higher frequency EEG arousal time during the 15 s before awakening from ELM and noELM conditions in REM and stage 2 sleep
Sleep stageConditionMean arousal time (s)SD
Stage 2NoELM1.50.90
ELM2.51.55
REMNoELM4.12.52
ELM4.22.62

Despite an experimental method to ensure minimal arousal related to the ELM condition, it was still associated with significantly more manually scored arousal than the noELM condition in stage 2 sleep (t23 = 4.43, P < 0.001). This was not the case for REM sleep (t23 = 0.08, P > 0.05).

EEG arousal, frequency of imagery reports and imagery ratings

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

Because of the possibility that elevated EEG arousal associated with ELMs might account for differences in the stage 2 mentation reports, the relationship between manually scored arousal and imagery reports within the ELM condition was investigated.

The average amount of manually scored stage 2 alpha or higher frequency EEG activity during the 15 s after the ELM and before waking, was found to be significantly higher for imagery reports (mean = 3.1 s, SD = 1.79) compared with reports with no imagery (mean = 2.3 s, SD = 1.50; t11 = 4.22, P < 0.01).

A correlation between the average imagery ratings from the stage 2 ELM condition with the average manually scored arousal from these awakenings (from each of the 16 qualifying participants with at least one imagery report from the stage 2 ELM condition) was not significant (Pearson r = 0.47, P > 0.05). This correlation was also not significant between mean imagery ratings from the stage 2 NoELM condition and mean manually scored arousal (from the 10 qualifying participants with at least one imagery report; Pearson r = 0.32, P > 0.1).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References

Visual imagery reports occurred more often after awakenings from REM compared to stage 2 sleep. In addition, when imagery was reported, imagery ratings were significantly higher after REM compared with stage 2 awakenings. These results are consistent with previous researchers who have described REM sleep as a state of high visual hallucinatory quantity and quality compared with NREM sleep (Hobson et al., 2003).

Awakenings after ELM events were hypothesized to be related to more frequent visual mentation reports compared with control awakenings void of ELM activity. This result was found within stage 2 sleep, but was not observed within REM sleep. This might have occurred because the frequency of imagery reporting from REM sleep was already at ‘ceiling’ levels, with extra phasic variations during REM sleep being of little consequence. This result is comparable with previous studies investigating other physiological events during REM sleep, which have found mixed results comparing dream reports from phasic and tonic REM sleep conditions (Pivik, 1991, 1994). However, because of the coincident activity of eye movements with ELM activity found previously (Conduit et al., 2002) and in the present study, one should consider any confounding influence of phasic EM activity when investigating REM sleep ELM correlates.

Another prediction of the present study was that imagery ratings would be higher in the ELM than the noELM condition. This result was not found for any comparison of ELM with noELM conditions. Hence, overall, the results suggest that ELM activity was related to increased occurrence of mentation reports from stage 2 NREM sleep, but not increased visual intensity of dream reports.

Previous researchers have suggested that PGO analogues could provide a better physiological indicator of dreaming than REM sleep itself (see Pivik, 1991, 1994). More recently, it has been suggested that indicators of ‘Covert REM sleep’ might provide a stronger indicator of dreaming during NREM sleep (Nielsen, 2003). However, PGO analogues such as MEMA and PIPs failed to provide a strong relationship to dream mentation recalled from sleep (Pivik, 1991, 1994; Rechtschaffen, 1973). According to the results of the current study, ELM appears to be no exception, although ELM activity was found to have a relationship with imagery reporting during NREM sleep. The only consistent difference between ELM and noELM conditions was a higher frequency of imagery reporting after stage 2 ELMs, with 44% of ELM awakenings and 23% of noELM awakenings providing reports containing imagery. No significant differences were found between REM sleep conditions. For imagery ratings, no differences were found between ELM conditions, regardless of sleep stage.

In this study, it was found that sleep stage was a better predictor of the occurrence of imagery mentation than ELM events. 83% of REM awakenings compared with only 34% stage 2 awakenings resulted in imagery reports. Sleep stages also provided more consistent differences in imagery ratings, with significantly higher imagery ratings from REM than stage 2 sleep.

On further exploring the ELM and noELM awakenings within stage 2 sleep, it was found that the ELM condition had significantly more manually scored arousal (during the 15 s before awakening) compared with the noELM condition. This result is consistent with previous findings of Cantero et al. (2002) who have shown ELMs in sleeping subjects are related to arousal, and decreased ELMs in awake subjects are related to decreased vigilance and sleep onset. Additionally, within the stage 2 ELM condition, significantly more arousal was present before imagery reports were reported compared with reports with no imagery. This suggests that elevated arousal associated with ELMs in stage 2 sleep might have accounted for the differences in mentation reporting. In addition, when average imagery ratings were correlated with the average amount of preceding arousal in the stage 2 ELM condition, a high but non-significant correlation was present (r = 0.49, P = 0.126), suggesting a relationship could be observed utilizing a larger sample size. However, as these are correlational relationships, the alternative argument remains that elevated visual dream imagery and/or the presence of imagery might produce elevated EEG arousal.

One question now remains: if both ELM activity and its associated EEG arousal are related to mentation reporting, is it the ELM, the arousal, or a third event related to both ELMs and arousal (e.g. PGO activity), involved in dream reporting? After previously reviewing similar mild, but significant relationships between other phasic events and sleep mentation, Pivik (2000), concluded:

This literature has made it clear, that phasic activity is not required for mental activity to be present during sleep, although phasic activity may, perhaps through an arousing effect, enhance the recall of mental activity from both REM and NREM sleep. (p. 497)

Several other researchers have also argued that brain arousal from sleep could play a major role in the process of dream reporting (Antrobus et al., 1995; Lehmann and Koukkou, 2003; Solms, 2003; Tyson et al., 1984). Ever since the discovery of REM sleep by Aserinsky and Klietman in 1953, the implicit and untested assumption of most researchers has been that mentation recalled from sleep is a direct reflection of prewaking mentation (Conduit et al., 2003). With still no scientific evidence to support this assumption, perhaps we should now start to consider the possibility that differences in dream recall may be just that, i.e. differences in recall. There is a real possibility that we describe more imagery after phasic events or REM sleep because more brain regions are aroused, allowing us to attend to mentation during sleep and therefore be better able to remember these events on awakening, not because these processes produce dream imagery during sleep.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Participants
  6. Apparatus and Materials
  7. Procedure
  8. Subject preparation
  9. Sleep scoring and the scoring of ELM events
  10. Awakening procedure
  11. Experimental (ELM) condition
  12. Control (noELM) condition
  13. Mentation report collection
  14. Results
  15. Frequency of imagery reports
  16. Imagery ratings
  17. ELMs and concurrent phasic EM activity
  18. ELM and EEG arousal
  19. EEG arousal, frequency of imagery reports and imagery ratings
  20. Discussion
  21. References
  • Ajilore, O. A., Stickgold, R., Rittenhouse, C. and Hobson, J. A. Nightcap: home-based evaluation of a portable sleep monitor. Psychophysiology, 1995, 32: 9298.
  • Antrobus, J. S., Fein, G., Jordan, L., Ellman, S. J. and Arkin, A. M. Measurement and design in research on sleep reports. In: S. J.Ellman and J. S.Antrobus (Eds) The Mind in Sleep: Psychology and Physiology, 2nd edn. J. Wiley, New York, 1991.
  • Antrobus, J., Kondo, T., Reinsel, R. and Fein, G. Dreaming in the late morning: summation of REM and diurnal cortical activation. Consc. Cog., 1995, 4: 275299.
  • Aserinsky, E. and Klietman, N. Regularly occurring periods of ocular motility during sleep. Science, 1953, 118: 273274.
  • Bliwise, D. L. and Rechtschaffen, A. Phasic EMG in human sleep III: periorbital potentials and NREM mentation. Sleep Res., 1978, 7: 58.
  • Bowker, R. M. and Morrison, A. R. The startle reflex and PGO spikes. Brain Res., 1976, 102: 185190.
  • Callaway, C. W., Lydic, R., Baghdoyan, H. A. and Hobson, J. A. Pontogeniculooccipital waves: spontaneous visual system activity during rapid eye movement sleep. Cell. Mol. Neurobiol., 1987, 7: 105149.
  • Campbell, S. P., Williams, D. A., Frueh, B. R. and Lynch, G. S. Contractile activation characteristics of single permeabilized fibres from levator palpebrae superioris, orbicularis oculi and vastus lateralis muscles from humans. J. Physiol. (London), 1999, 519: 615622.
  • Cantero, J. L., Atienza, M., Stickgold, R. and Hobson, J. A. Nightcap: a reliable system for determining sleep onset latency. Sleep, 2002, 25: 238245.
  • Conduit, R. A phenomenological investigation of the relationship between PGO waves and dream reporting in humans: towards an attention-based model of dreaming, Unpublished PhD dissertation. La Trobe University, Melbourne, Australia, 1999.
  • Conduit, R., Bruck, D. and Coleman, G. Induction of visual imagery during NREM sleep. Sleep, 1997, 20: 948956.
  • Conduit, R., Crewther, S., Bruck, D. and Coleman, G. Spontaneous eyelid movements during human sleep: a possible PGO analogue? J. Sleep Res., 2002, 11: 95104.
  • Conduit, R., Crewther, S. and Coleman, G. Shedding old assumptions and consolidating what we know: toward an attention-based model of dreaming. In: E. F.Pace-Schott, M.Solms, M.Blagrove and S.Harnad (Eds) Sleep and Dreaming: Scientific Advances and Reconsiderations. Cambridge University Press, Cambridge, UK, 2003.
  • Duysan-Peyrethon, D., Peyrethon, J. and Jouvet, M. Etude quantitative des phenomenes phasiques du sommeil paradoxal pendant apres sa deprivation instrumentale. [Quantitative study of phasic phenomena of paradoxical sleep during and after its instrumental deprivation]. Comptes Rendus des Seances de la Societe de Biologie, 1967, 161: 25302533.
  • Ferrari, E. and Messina, C. Blink reflexes during sleep and wakefulness in man. Electroenceph. Clin. Neurophysiol., 1972, 32: 5562.
  • Filion, D. L., Dawson, M. E. and Schell, A. M. The psychological significance of human startle eyeblink modification: a review. Biol. Psychol., 1998, 47: 143.
  • Foulkes, D. Nonrapid eye movement mentation. Exp. Neurol., 1967, 19(Suppl. 4): 2838.
  • Fredrickson, C. J., Whitehead, W. E. and Rechtschaffen, A. Phasic activity of the rat laternal rectus muscle during sleep. Psychophysiol., 1972, 9: 130.
  • Hess, A. and Pilar, G. Slow fibres in the extraocular muscles of the cat. J. Physiol., 1963, 169: 780798.
  • Hobson, J. A. and McCarley, R. W. The brain as a dream-state generator: an activation-synthesis hypothesis of the dream process. Am. J. Psychiat. 1977, 134: 13351348.
  • Hobson, J. A. and Stickgold, R. Dreaming: a neurocognitive approach. Consc. Cog., 1994, 3: 115.
  • Hobson, J. A., Pace-Schott, E. F. and Stickgold, R. Dreaming and the brain: toward a cognitive neuroscience of conscious. In: E. F.Pace-Schott, M.Solms, M.Blagrove and S.Harnad (Eds) Sleep and Dreaming: Scientific Advances and Reconsiderations. Cambridge University Press, Cambridge, UK, 2003.
  • Hunt, W. K., Sanford, L. D., Ross, R. J., Morrison, A. R. and Pack, A. I. Elicited pontogeniculooccipital waves and phasic suppression of diaphram activity in sleep and wakefulness. J. Appl. Physiol., 1998, 84: 21062114.
  • Jasper, H. H. The ten-twenty electrode placement system of the International Federation. Electroenceph. Clin. Neurophysiol., 1958, 10: 418424.
  • Kirk, R. E. Experimental design (2nd ed) . Brooks Cole, Monterey, CA, 1982.
  • Lehmann, D. and Koukkou, M. All brain work including recall is state dependent. In: E. F.Pace-Schott, M.Solms, M.Blagrove and S.Harnad (Eds) Sleep and Dreaming: Scientific Advances and Reconsiderations. Cambridge University Press, Cambridge, UK, 2003.
  • McCarley, R. W., Winkleman, J. W. and Duffy, F. H. Human cerebral potentials associated with REM sleep rapid eye movement: links to PGO waves and waking potentials. Brain Res., 1983, 274: 308312.
  • Nielsen, T. A review of mentation in REM and NREM sleep: ‘covert’ REM sleep as a possible reconciliation of two opposing models. In: E. F.Pace-Schott, M.Solms, M.Blagrove and S.Harnad (Eds) Sleep and Dreaming: Scientific Advances and Reconsiderations. Cambridge University Press, Cambridge, UK, 2003.
  • Ogilvie, R., Hunt, H., Sawicki, C. and Samahalski, J. Psychological correlates of spontaneous MEMA during sleep. Sleep, 1982, 11: 1127.
  • Orem, J. and Dement, W. C. Spontaneous eyelid behavior in the sleeping cat. Exp. Neurol., 1974, 44: 145159.
  • Pace-Schott, E., Kaji, J., Stickgold, R. and Hobson, J. A. Nightcap measurement of sleep quality in self-described good and poor sleepers. Sleep, 1994, 17: 688692.
  • Pivik, R. T. Tonic states and phasic events in relation to sleep mentation. In: S. J.Ellman and J. S.Antrobus (Eds) The Mind in Sleep: Psychology and Physiology. 2nd edn. J. Wiley, New York, 1991.
  • Pivik, R. T. The psychophysiology of dreams. In: M. H.Kryger, T.Roth and W. C.Dement (Eds) Principles and Practice of Sleep Medicine, 2nd edn. W.B. Saunders Co., Philadelphia, 1994.
  • Pivik, R. T. Psychophysiology of dreams. In: M. H.Kryger, T.Roth and W. C.Dement (Eds) Principles and Practice of Sleep Medicine, 3rd edn. W.B. Saunders Co., Philadelphia, 2000.
  • Rechtschaffen, A., Michel, F. and Metz, J. T. Relationship between extraoccular and PGO activity in the cat. Psychophysiol., 1972, 9: 128.
  • Rechtschaffen, A. The psychophysiology of mental activity during sleep. In: F. J.McGuigan and R. A.Schoonove (Eds) The Psychophysiology of Thinking. Academic Press, New York, 1973.
  • Rechtschaffen, A. and Kales, A. A Manual for the Standardized Terminology, Techniques and Scoring System for Sleep Stages of Human Subjects. Public Health Services, US Government Printing Office, Washington, DC, 1978.
  • Rowley, J. T., Stickgold, R. and Hobson, J. A. Eyelid movements and mental activity at sleep onset. Consc. Cog., 1998, 7, 6784.
  • Sanford, L. D., Morrison, A. R., Ball, W. A., Ross, R. and Mann, G. L. Varying expressions of alerting mechanisms in wakefulness and across sleep states. Electroenceph. Clin. Neurophysiol., 1992, 82: 458468.
  • Sanford, L. D., Morrison, A. R., Ball, W. A., Ross, R. and Mann, G. L. The amplitude of elicited PGO waves: a correlate of orienting. Electroenceph. Clin. Neurophysiol., 1993, 86: 438445.
  • Sanford, L. D., Morrison, A. R. and Ross, R. J. Effects of auditory stimulation on phenomena of rapid eye movement sleep. In: Mallick, B. N. and Singh, R. (Eds). Environment & Physiology. Narosa Publishing, New Delhi, 1994.
  • Solms, M. Dreaming and REM sleep are controlled by different brain mechanisms. In: E. F.Pace-Schott, M.Solms, M.Blagrove and S.Harnad (Eds) Sleep and Dreaming: Scientific Advances and Reconsiderations. Cambridge University Press, Cambridge, UK, 2003.
  • Spencer, R. F., McNeer, K. W. The periphery: extraocular muscles and motor neurones. In: J.Cronly-Dillon (Ed.) Vision and Visual Dysfunction. Chapter 8: Eye Movements, R. H. S.Carpenter (Ed.) Macmillan Press, London, 1991.
  • Stickgold, R. and Hobson, J. A. Home monitoring of sleep-onset and sleep-onset mentation using the nightcap. In: R. D.Ogilvie and J.Harsh (Eds) Sleep Onset: Normal and Abnormal Processes. American Psychological Association, Washington DC, 1994: 141160.
  • Tyson, P. D., Olgilvie, R. D. and Hunt, H. T. Lucid, prelucid and nonlucid dreams related to the amount of EEG alpha activity during REM sleep. Psychophysiology, 1984, 21: 442451.
  • Vimont-Vicary, P., Jouvet, D. and Delrome, F. Effets EEG et comportementaux des privations du sommeil paradoxical chez le chat. [EEG and behavioral effects of deprivation of paradoxical sleep in cats]. Electroenceph. Clin. Neurophysiol., 1966, 20: 439449.