REM and NREM power spectral analysis on two consecutive nights in psychophysiological and paradoxical insomnia sufferers☆,☆☆
Introduction
Nearly 10% of the general population is suffering from chronic insomnia (Morin et al., 2006). In addition, a complaint of insomnia is expressed by one fifth of those who consult a generalist (Vgontzas and Kales, 1999). Associated daytime consequences are often the reason for medical consultation and treatment seeking (Morin et al., 2006). Chronic insomnia is defined by a complaint of difficulty falling asleep, awaking at night or too early in the morning, or of poor sleep quality for more than a month that conducts to significant distress or alters social or professional functioning (American Psychiatric Association. and American Psychiatric Association. Task Force on DSM-IV, 2000, American Sleep Disorders Association. Diagnostic Classification Steering Committee, 1990). The International Classification of Diseases (Organisation Mondiale de la Santé, 1992) adds the criterion of three nights of sleep difficulty per week. Many studies have investigated the sleep macrostructure of insomnia sufferers although less work has been dedicated to the microstructure of their sleep.
Studies have found several differences in terms of physiological and cortical activation between chronic primary insomnia sufferers (PI) and good sleeper controls (GS). PI usually present greater metabolic rate as well as faster cardiac rhythm on a 24 h basis compared to GS (Bonnet and Arand, 1995, Monroe, 1967). Polysomnography (PSG) studies suggest that PI's sleep macrostructure is impaired, presenting longer sleep-onset latency (SOL), more stage 1 sleep and less stage 3 and 4 sleep (Merica et al., 1998, Gaillard, 1976, Frankel et al., 1976). Sleep microstructure, mainly assessed through the activity of different frequency bands at central sites with power spectral analysis (PSA), presents also particularities in PI compared to GS. At sleep onset, less delta power and more beta and alpha spectral powers are observed (Merica and Gaillard, 1992, Lamarche and Ogilvie, 1997, Staner et al., 2003). Altogether, results in non rapid eye movement (NREM) sleep suggest higher beta1, beta2 and gamma spectral power (Merica et al., 1998, Perlis et al., 2001a). Furthermore, elevated powers in alpha, sigma and theta frequencies have been reported (Krystal et al., 2002, Nofzinger et al., 1999). In rapid eye movement (REM) sleep, some studies have also reported increased activity in the beta frequency band (Merica et al., 1998, Freedman, 1986, Perlis et al., 2001b). These increased activities in faster frequency bands, suggesting increased cortical activation, support theories stating that insomnia is linked to emotional and physiological hyperactivation as predisposing, precipitating, and maintaining factors (Bonnet and Arand, 2010).
PI sufferers represent a heterogeneous group of individuals. Sub-classifications of the diagnosis of chronic insomnia exist (Organisation Mondiale de la Santé, 1992) and include, among others, psychophysiological insomnia sufferers (PsyI) and paradoxical insomnia sufferers (ParI). The former relates to a condition where the difficulties are caused by a conditioned aroused state at bedtime and the latter where objective reports do not corroborate the patient's complaint of sleep difficulty. Edinger et al. (2004) proposed that the privileged interpretation of ParI would be the absence of sleep pathology while a complaint is expressed. The following criteria were suggested: it is necessary that the absence of sleep pathology be shown as expressed by a total sleep time (TST) of more than 6.5 h and sleep efficiency (SE) of at least 85% at night PSG recordings. Moreover, important sleep difficulties should be reported. As such, it may be the report of a subjective impression of little or no sleep most of the time and rare normal nights. Otherwise, the average sleep time, in at least a week of sleep logs, is well below normative values, often with several sleepless nights and no daytime naps. In addition, a consistent and marked difference between PSG data and subjective sleep estimates is present. Since these guidelines are recent, few research protocols have used them to examine possible discrepancies between types of insomnia. Although seemingly thorough, these guidelines are nonetheless lacking operationalization. We intend to remedy to this lack of operationalization in the present paper.
Among studies that did categorize PI in PsyI and ParI, although ParI were then called ‘sleep state misperception’, larger absolute spectral amplitude in NREM in frequency bands from 14 to 54 Hz (beta and gamma bands) in ParI than GS (Perlis et al., 2001b) was reported. On the other hand, Krystal et al. (2002) observed larger absolute spectral amplitude in frequency bands ranging from 8.5 to 30 Hz (sigma, alpha, and beta bands) in stages 2 and 4 of ParI compared to PsyI and GS. These interesting results suggest higher activation in ParI than PsyI. These authors have also found trends for relations between NREM spectral power in ParI and subjective rest, TST and SE, three variables that can be related to sleep quality. As such, the differences in absolute spectral power observations between PsyI and ParI are not necessarily or only attributable to group belonging, but could also reflect a difference in the perceived sleep quality. This information is meaningful in view of the many studies that have used only one night of PSG recording, this procedure leading to the possible observation of a first night effect where participants' sleep does not represent their usual sleep because of the unfamiliar experimentation setting (Hauri and Olmstead, 1989). Moreover, previous studies did not assess the natural fluctuation of sleep difficulties in insomnia, instability of sleep quality (Vallieres et al., 2005), while collecting PSA data. Consequently, laboratory settings and sleep variations are to be considered.
In addition to the above mentioned variations in studies' protocols, some authors have reported absolute, and others relative PSA measures. Such inconsistencies in PSA literature make it even more laborious to compare and relate the conclusions of each study so to brush a comprehensible picture of cortical activation in insomnia. Whereas absolute PSA is the measure of the actual power in the designated frequency bands, relative PSA refers to the proportion of the power in a frequency band in relation with the power in all other frequency bands. Thus, these two measures offer different viewpoints of the activity in frequency bands and are both of interest. As relative power calculation reduces the variance between subjects due to individual anatomical differences and tissue conductance found in absolute power, it has the inconvenience of distorting the data interpretation by creating dependencies between bands as an increase of one frequency band is understood as a decrease of other bands. Still, the evaluation of relative power spectra is indicated for the detection of subtle shifts in brain function over time by normalizing fluctuations in total power seen across individuals or within one individual across several recordings (Chen and Black, 2005). Unfortunately, cortical activation at other sites than central, such as frontal and parietal regions, has been poorly explored. The comparison of multiple sites in statistical models certainly contributes to burden the interpretation of the results and seems unappealing. On the other hand, if not studied, it will limit the cortical mapping of activity of insomnia and its cognitive underpinnings. To our knowledge, only one study has compared multiple sites while studying ParI (n = 10) and GS (n = 10) (Marzano et al., 2008). Although this study targeted sleep-onset only, 1 Hz binned preliminary results showed local functional impairment in the synchronization process of ParI compared to GS. In this regard, ParI tended to display more beta and less delta and sigma relative power on anterior scalp locations (mainly centro-parietal). Although these preliminary findings are interesting, we still do not know about nightly sleep/awakenings. Furthermore, it is now recognized that frontal regions are associated with planning, judgement and memory, among others, and parietal regions with sensory information processing (Purves et al., 1999), which could explain the subjective recall of the night, e.g. in sleep diaries.
Therefore, the aim of this study is to examine the sleep microstructure (absolute and relative PSA) differences between GS, PsyI and ParI on the second and third night of a three night protocol at multiple scalp sites (frontal, central and parietal). It is hypothesized that previous results will be replicated i.e. elevated activity in alpha, sigma and beta frequency bands in ParI compared to GS and PsyI will be found at central sites. As for frontal and parietal sites, far less explored in PSA studies, the objective is to provide a description of the cortical activity in those regions although they should also show signs of elevated activity in ParI. Finally, it is expected that more PSA variability between nights in insomnia sufferers (INS) will be observed since they might also display a variation in their objective sleep quality and report a variation in their subjective sleep quality over the course of the two nights of recording, as for diagnostic criteria permit such observation in sleep difficulties.
Section snippets
Participants
Volunteers from 25 to 55 years of age were recruited through local newspaper advertisement. On the basis of the three group criteria, 26 PsyI (14 females, 12 males), 20 ParI (14 females, 6 males) and 21 GS (12 females, 9 males) participated in this study. Participants had a mean age of 40.21 (9.38) years.
Sociodemographic measures
The final sample comprised 26 PsyI, 20 ParI and 21 GS (mean age = 40 years, SD = 9.38). Statistical analyses showed that GS, PsyI and ParI were similar according to gender, χ2(2, N = 67) = 1.31, p = .52, and age, F(2, 64) = .48, p = .62. As expected, PsyI and ParI, compared to GS, had higher scores on the ISI, F(2, 61) = 65.14, p < .001, this higher score being indicative of greater severity of insomnia symptoms. Also, both groups of insomnia sufferers expressed more depressive and anxious symptoms than GS as
Discussion
The particularity of this study was the observation of absolute and relative power spectral analysis of ParI, PsyI and GS on two consecutive nights. This method was preferred over the analysis of a single PSG recording night to take into account not only the variability of sleep difficulties insomnia sufferers might present as a group, but also the fluctuation in individual patterns of sleep. Furthermore, efforts were made to categorize primary insomnia in two groups, PsyI and ParI, and
Conclusion
Finally, this study concludes to a decrease in absolute cortical activation in NREM in ParI compared to PsyI and GS as well as increased relative activation in REM. It also suggests a unique repartition of the cortical activation in GS, ParI and PsyI. The evaluation of the distinctive characteristics between PsyI, ParI and the non-clinical population would benefit from the use of other techniques. Parrino et al. (2007) have found interesting results while studying ParI and controls. Whereas
References (47)
- et al.
Validation of the Insomnia Severity Index as an outcome measure for insomnia research
Sleep Medicine
(2001) - et al.
Psychometric properties of the Beck depression inventory — 25 years of evaluation
Clinical Psychology Review
(1988) - et al.
Hyperarousal and insomnia: state of the science
Sleep Medicine Reviews
(2010) - et al.
Regional brain asymmetries in major depression with or without an anxiety disorder: a quantitative electroencephalographic study
Biological Psychiatry
(1997) - et al.
Quantitative analysis of the sleep encephalogram
- et al.
Cortical gamma responses: searching high and low
International Journal of Psychophysiology
(2011) Eeg power spectra in sleep-onset insomnia
Electroencephalography and Clinical Neurophysiology
(1986)- et al.
The Eeg of the sleep onset period in insomnia — a discriminant-analysis
Physiology & Behavior
(1992) - et al.
Epidemiology of insomnia: prevalence, self-help treatments, consultations, and determinants of help-seeking behaviors
Sleep Medicine
(2006) - et al.
Beta EEG activity and insomnia
Sleep Medicine Reviews
(2001)
Is quality of sleep related to the N1 and P2 ERPs in chronic psychophysiological insomnia sufferers?
International Journal of Psychophysiology
Diagnostic and statistical manual of mental disorders: DSM-IV-TR
The International classification of sleep disorders: diagnostic and coding manual
Signs of insomnia in borderline personality individuals
Journal of Clinical Sleep Medicine
Information processing varies between insomnia types: measures of N1 and P2 during the night
Behavioral Sleep Medicine
An inventory for measuring clinical anxiety — psychometric properties
Journal of Consulting and Clinical Psychology
Assessment of insomnia: validation of three questionnaires
Encephale-Revue De Psychiatrie Clinique Biologique Et Therapeutique
24-hour metabolic-rate in insomniacs and matched normal sleepers
Sleep
Muscle artifacts in the sleep EEG: automated detection and effect on all-night EEG power spectra
Journal of Sleep Research
EEG spectral analysis in primary insomnia: NREM period effects and sex differences
Sleep
Derivation of research diagnostic criteria for insomnia: Report of an American Academy of Sleep Medicine Work Group
Sleep
Interhemispheric asymmetry of human sleep EEG in response to selective slow-wave sleep deprivation
Behavioral Neuroscience
Recorded and Reported Sleep in Chronic Primary Insomnia
Archives of General Psychiatry
Cited by (0)
- ☆
All authors have indicated no financial conflicts of interest.