2024; 22(1): 95-104  https://doi.org/10.9758/cpn.23.1063
Hyperarousal-state of Insomnia Disorder in Wake-resting State Quantitative Electroencephalography
Gyutae Jang1,*, Han Wool Jung1,*, Jiheon Kim1,2, Hansol Kim1, Ji‑Hyeon Shin3, Chan-Hyung Kim4, Do-Hoon Kim1,2, Sang-Kyu Lee2, Daeyoung Roh1,2
1Mind-Neuromodulation Laboratory, Hallym University College of Medicine, Chuncheon, Korea
2Department of Psychiatry, Hallym University College of Medicine, Chuncheon, Korea
3Department of Otolaryngology-Head and Neck Surgery, College of Medicine, The Catholic University of Korea, Seoul, Korea
4Department of Psychiatry and Institute of Behavioural Science in Medicine, Yonsei University College of Medicine, Seoul, Korea
Correspondence to: Daeyoung Roh
Department of Psychiatry, Hallym University Chuncheon Sacred Heart Hospital, 77 Sakju-ro, Chuncheon 24253, Korea
E-mail: omydoc@hallym.ac.kr
ORCID: https://orcid.org/0000-0001-7242-9496

*These first authors contributed equally to this work.
Received: February 10, 2023; Revised: March 15, 2023; Accepted: March 16, 2023; Published online: May 26, 2023.
© The Korean College of Neuropsychopharmacology. All rights reserved.

This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
Objective: Insomnia is associated with elevated high-frequency electroencephalogram power in the waking state. Although affective symptoms (e.g., depression and anxiety) are commonly comorbid with insomnia, few reports distinguished objective sleep disturbance from affective symptoms. In this study, we investigated whether daytime electroencephalographic activity explains insomnia, even after controlling for the effects of affective symptoms.
Methods: A total of 107 participants were divided into the insomnia disorder (n = 58) and healthy control (n = 49) groups using the Mini-International Neuropsychiatric Interview and diagnostic criteria for insomnia disorder. The participants underwent daytime resting-state electroencephalography sessions (64 channels, eye-closed).
Results: The insomnia group showed higher levels of anxiety, depression, and insomnia than the healthy group, as well as increased beta [t(105) = −2.56, p = 0.012] and gamma [t(105) = −2.44, p = 0.016] spectra. Among all participants, insomnia symptoms positively correlated with the intensity of beta (r = 0.28, p < 0.01) and gamma (r = 0.25, p < 0.05) spectra. Through hierarchical multiple regression, the beta power showed the additional ability to predict insomnia symptoms beyond the effect of anxiety (ΔR2 = 0.041, p = 0.018).
Conclusion: Our results showed a significant relationship between beta electroencephalographic activity and insomnia symptoms, after adjusting for other clinical correlates, and serve as further evidence for the hyperarousal theory of insomnia. Moreover, resting-state quantitative electroencephalography may be a supplementary tool to assess insomnia.
Keywords: Spectral power analysis; Pittsburgh Sleep Quality Index; Hamilton Depression Rating Scale; Hamilton Anxiety Rating Scale
INTRODUCTION

Insomnia is the most common, distressing sleep-wake problems. According to the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5), the diagnosis of insomnia is based on the patient’s subjective complaint regarding sleep quality, difficulty in sleeping despite the adequate opportunity to sleep, and impairment in daytime functioning due to the difficulty in sleeping [1,2]. Its main symptoms are difficulty in the initiation or maintenance of sleep and early awakening. It triggers or aggravates fatigue and exhaustion in daily life, increases the risk of cardiovascular disease [3], and causes immune dysfunction. Furthermore, it increases the risk of numerous neuropsychiatric disorders, such as anxiety disorder [4], major depressive disorder [5], and dysfunction of cog-nitive abilities such as episodic memory, problem solving, and working memory [6].

The hyperarousal theory of insomnia is one of the most widely known theories explaining sleep disturbance [7, 8]. It emphasizes arousal as a central aspect of insomnia. According to this theory, insomnia is defined as a hyperarousal state of the central nervous system. The cycle of hyperarousal for insomnia they present with is simple but compelling. First, daily stressful events trigger a form of arousal state. Once this daytime arousal continues into the night, it functions as a sleep-related stimulus, finally leading to a learned sleep-preventing association (i.e., conditioned hyperarousal). The learned sleep-preventing association not only deprives the current sleep but also increases affective insomnia symptoms such as worry or ruminative thoughts, which worsen insomnia symptoms and reform arousal states. Eventually, it develops into chronic insomnia [7].

This model explains the phenomenon of insomnia well, especially considering high-frequency electroencephalogram (EEG) power [9,10]. For decades, studies on brain activity related to insomnia have consistently supported the hyperarousal model. These studies encompass not only EEG but also diverse neurophysiological methods such as polysomnography (PSG), event-related potentials, neuroimaging, various tests of the autonomic nervous system (e.g., heart rate, temperature, metabolic rate), and neuroendocrine system/hormone tests [11].

However, one limitation of classic research was that it had mainly focused on brain/neural activities when sleeping or at most, ‘on the bed.’ As explained above, the first stage of the hyperarousal cycle begins with daytime events; therefore, identifying activities in the waking state should be considered theoretically as important as those in the sleeping state. Nevertheless, most of the research methods mentioned above (e.g., PSG, sleep cycle, rapid eye movemen [REM]) do not seem to reflect daytime activities well. Moreover, although researchers have started to consider the daytime measures for insomnia, it seems to have been derived from clinical reasons rather than theoretical ones because patients with insomnia have consistently indi-cated cortical hyperarousal not only in the sleeping state but also during the daytime [12,13]. Nevertheless, the numerous advantages of measuring daytime resting-state brain activity have now been recognized. First, it provides new information unidentifiable with conventional sleep tests. In general, resting-state EEG can be obtained in 5−10 minutes. In addition, as it is measured simultaneously with the patient’s self-reported symptoms, the validity of the specific physiological data collected is ensured [12]. Finally, it allowed us to obtain daytime stress monitoring data for insomnia [9].

Recent studies have demonstrated that patients with insomnia have unique electroencephalographic patterns in both sleep and waking states, which can be interpreted as hyperarousal. In the wake-resting state EEG, patients with insomnia tend to show increased high-frequency power spectra and decreased low-frequency power spectra [13]. Specifically, less power in the narrow upper alpha band and more power in a broad beta power range were observed [14]. Evidence suggests that such characteristics can explain hyperarousal, and a decrease in theta power and an increase in beta power are known to be associated with hyperarousal [15].

However, these studies are insufficient to provide strict evidence for the hyperarousal theory of insomnia. As described above, since affective symptoms play an important role in this theory, a more in-depth examination of the role of affective symptoms between EEG power/neurological arousal and sleep disturbance among patients with insomnia is needed. However, most studies to date tend to not adequately consider affective symptoms, which might be a crucial problem because the high-frequency EEG power among patients with insomnia may be interpreted as a mere byproduct of their affective (especially, anxious) symptoms, rather than as an actual cause of sleep disturbance. Although affective symptoms are broadly included in the domain of insomnia, it is necessary to distinguish between arousal as an affective symptom and the one that directly affects sleep, especially considering that objective sleep disturbance and affective symptoms are known to be controlled by different brain networks in the mechanism of insomnia. In particular, a recent systematic review presented two intertwined but distinct brain networks that cause insomnia. According to this review, the default mode network is composed of the connectivity between the posterior cingulate cortex and the ventromedial prefrontal cortex and is linked to objective sleep disturb-ance. On the other hand, the salience network consists of the connection between the anterior cingulate cortex, amygdala, and insula and is relevant to affective symptoms and self-reported alertness [16]. For the progress of hyperarousal theory, we need to further investigate whether insomnia is associated (or more associated) with affective symptoms such as depression/anxiety or the arousal derived from a separate origin.

Although previous studies have adopted a process that excludes those with depression or anxiety disorders through the initial screening or pre-interview [4,17], we decided to pursue a further investigation to reveal the more scrupulous mechanisms of insomnia. In this study, the measurement of affective symptoms will be included in the study models, and such effects will be separated from objective sleeping symptoms during the analysis to realize the antecedents of insomnia. In other words, as an extension of recent studies regarding the wake-resting state EEG of insomnia, this study aimed to test the predictive ability of such brain activity measured during the daytime after the effects of affective symptoms are controlled. In this study, we expected that the spectral power indicating hyperarousal, which is observed in the wake-resting state, is closely related to insomnia, even after the symptoms of depression and anxiety are controlled.

METHODS

Participants

The participants were recruited online or through local community advertisements. For the initial screening and group assignment, each participant attended an interview conducted by a trained clinical psychologist or psychiatrist. During this stage, they performed the Mini-International Neuropsychiatric Interview (MINI) [18] and the structured clinical assessment of depression and anxiety (i.e., the Hamilton Depression Rating Scale [HAM-D] [19] and Hamilton Anxiety Rating Scale [HAM-A] [20]). Sleep symptoms were assessed using the Pittsburgh Sleep Quality Index (PSQI) [21]. The HAM-D, HAM-A, and PSQI were also used for the main analysis (see below). This study was approved by the Institutional Review Board of Chuncheon Sacred Heart Hospital, Hallym University, Republic of Korea. All the participants provided written informed consent.

Participants with (a) hearing problems; (b) current psychiatric disorders such as mood disorders, anxiety disorders, substance use disorders, or psychotic disorders; (c) psychotic disorders such as schizophrenia, delusional disorder, or personality disorders; (d) current or historic epileptic symptoms; (e) scalp-related problems; (f) other severe disorders in the central nervous system; (g) sleep disorders other than insomnia (e.g., obstructive sleep apnea syndrome, periodic limb movement disorder, restless legs syndrome, and narcolepsy); (h) having atypical sleep schedules (e.g., shiftwork, an overseas trip across time zones within 6 months); and (i) sleep-wake cycle problems were excluded from the initial interview. Among the participants who passed the initial screening, the insomnia disorder (ID) group comprised those who (1) complained of poor sleep quality or quantity, (2) met the diagnostic criteria of insomnia disorder in the DSM-5, and (3) scored 8 or higher on the PSQI (see below). The healthy control (HC) group consisted of those who did not meet all the three criteria. Those who met only some of the criteria were excluded from this study. Finally, 107 participants were included in the study, wherein 58 were in the ID group, and 49 were in the HC group.

Clinical Measures

PSQI

The PSQI [21] is a self-report tool that measures sleep quality and problems. In this study, a validated Korean version of the PSQI (PSQI-K) [22] was used. This version has an internal consistency (Cronbach’s alpha) of 0.84. Usually, a PSQI-K total score of 5 or more is diagnosed as insomnia in most clinical settings; however, to only include patients with clear insomnia in research settings, 8 points instead of 5 were used as a cut-off in this study. Psychometric research, including patients with less severe insomnia symptoms, may cause validity issues [23]. The validation study of the Korean version also recommended 8 to 9 points because the optimized sensitivity and specificity were found to be 8.5 [22].

HAM-D/HAM-A

The Hamilton Depression Rating Scale (HAM-D or HDRS) is a structured clinical interview tool, which consists of 17 questions, used to measure depressive symptoms [19,24]. Similarly, the Hamilton Anxiety Rating Scale (HAM-A or HARS) is a 14-item structured interview tool used to diagnose anxiety [20]. The HAM-D and HAM- A have long been known as the ‘gold standards’ for both clinical diagnosis and psychological measures of the most common neurotic disorders of depression and anxiety and are integral parts of the objective measurement of the severity of symptoms or effects of treatments [25,26]. The validated Korean versions of the HAM-D [26] and HAM-A [25] were used in this study.

EEG Analysis

Data collection

A 5-minute resting-state EEG (64 channels) session was conducted for each participant. Only the eye-closed state was measured because, in the case of the eye-open state, EEG related to visual processing may interfere with the output [27]. All participants were not taking medication pills for a week before the EEG measurement. The participants were seated on a comfortable chair in a sound-attenuated room without performing any other activities or tasks. The data were recorded with a 1−100-Hz band- pass filter at a sampling rate of 1,000 Hz, using NetStation 4.4.2 (EGI Software), an EGI NetAmps 300 amplifier with a 24-bit analog-to-digital converter, and the HydroCel Geodesic Sensor Net. The device had 19 electrodes, and the scalp channels were initially referenced to Cz. Imped-ance was maintained below 50 kW during the entire recording session. All EEG data were recorded between 10:00 and 16:00.

Preprocessing

The EEG data were preprocessed using EEGLAB 2021.0 (Swartz Center for Computational Neuroscience), MATLAB R2021a (MathWorks), and Neuroguide 2.8.5 (Applied Neuroscience). Online signals were re-referenced to the linked ear. Common artifacts (e.g., eye blinks, pulse, myogenic artifacts, eye/tongue movement artifacts) were checked and removed through visual inspection, sweeps, and independent component analysis with EEGLAB and by a trained person without prior information about the participants and data. Artifacts exceeding ± 100 mV were excluded from all channels. For some participants, bad channel data that provided invalid information were excluded from the analysis.

For spectral power analysis, 15 artifact-free epochs, each of them having 4,096 ms in length, were randomly selected from each participant’s original session. Using a fast Fourier transform with a 1−50-Hz filter, the signal data were divided into the following frequency bands: delta (1−4 Hz), theta (4−8 Hz), alpha (8−12 Hz), beta (12−30 Hz), and gamma (30−40 Hz). High gamma frequencies of 40−50 Hz were excluded from the analysis. For EEG analysis, we only used 19 standard 10−20 electrodes considering user-friendly and easy setup for real-world applications. The relative power for each electrode was calculated from the absolute power value as the percentage of the total power (1−50 Hz) contained in each band (e.g., 1−4 Hz). The regional relative power for each frequency band (delta, theta, alpha, beta, and gamma) was calculated as the average of the valid relative power data from a set of electrodes categorized according to the following criteria: left frontal (Fp1, F3, F7), right frontal (Fp2, F4, F8), left temporal (T3, T5), right temporal (T4, T6), centro-parietal (C3, C4, Cz, P3, P4, Pz), occipital (O1, O2), and uncategorized/mid-frontal (Fz). Finally, the total (overall) relative power of each frequency band was calculated as the average relative power of all valid channels.

Statistical Analysis

Statistical analyses were performed using MATLAB R2021a and IBM SPSS Statistics 26 (IBM Corp.). Demo-graphic data, descriptive statistics, and correlations between all major measures were provided. The group differences between ID and HC in terms of demographics and clinical measures were analyzed using ttests or χ2 tests. The group differences in each of the specified brain regions (i.e., left/right frontal/temporal, centroparietal, and occipital) and the overall relative power of each band were analyzed using ttests, and the localized group differences were presented visually. Finally, the effect of the overall relative EEG power on sleep symptoms (PSQI), along with the variability of HAM-D/HAM-A, was estimated using the hierarchical multiple regression model. Statistical significance was determined based on an alpha value of 0.05, but for the regional analysis, the adjusted pvalues based on the Benjamini-Hochberg method (with a false discovery rate of 0.05) was reported to correct for the multiple statistical testing.

RESULTS

Demographics and Descriptive Statistics

Table 1 shows the demographics and clinical measures of the two groups and their differences. Generally, the two groups did not seem to violate the homogeneity assumption based on the demographic variables of age and sex; however, a slight difference was observed in the sex ratio. Unsurprisingly, the ID group reported higher HAM-D, HAM-A, and PSQI scores than the HC group. Considering that this study aimed to determine the predictive ability of EEG power after controlling for the influences of affective disorders, caution must be observed when interpreting the results in terms of group differences because the two groups have different baseline depression and anxiety scores.

Group Differences

Table 2 presents the differences between the ID and HC groups in the overall relative power of each frequency band. As expected, the ID group reported significantly higher beta and gamma power than the HC group. In addition, although insignificant, the HC group had a tendency to have higher alpha power than the ID group. Generally, the ID group reported a stronger power of high-frequency bands, although few differences were observed in the low-frequency bands.

Figure 1 shows the visualized group differences in localized relative spectral power for each band. Testing the beta band with multiple test adjustments, the higher power for the ID group was found in the right frontal [Fp2, F4, F8; t(103) = −2.33, adjusted p = 0.029], left temporal [T3, T5; t(103) = −2.48, adjusted p = 0.029], right temporal [T4, T6; t(104) = −2.30, adjusted p = 0.029], centro- parietal [C3, C4, Cz, P3, P4, Pz; t(104) = −2.57, adjusted p = 0.029], occipital [O1, O2; t(103) = −2.28, adjusted p = 0.029], and mid-frontal [Fz; t(102) = −2.32, adjusted p = 0.029] areas. The difference in the left frontal (Fp1, F3, F7) area was not significant [t(103) = −1.64, adjusted p = 0.103].

Testing the gamma band with multiple test adjustments, the higher power for ID was found in the right frontal [t(101.84) = −2.31, adjusted p = 0.048], left temporal [t(102) = −2.46, adjusted p = 0.048], centro-parietal [t(105) = −2.24, adjusted p = 0.048], and mid-frontal [Fz; t(99.60) = −2.71, adjusted p = 0.048] areas. The differences in the rest areas were not significant (adjusted ps > 0.10).

For the rest bands, no significant difference was found in the regional analysis. In general, the differences in band power appeared globally rather than locally; such tendencies can also be observed visually in Figure 1.

Correlations

Table 3 presents the correlation coefficients of the variables measured in this study. As expected, the overall relative powers for beta and gamma were significantly positively correlated with sleep symptoms (PSQI). In addition, alpha power showed a marginally significant negative correlation with PSQI (r = −0.189, p = 0.051). The PSQI also showed significant correlations with the HAM-D and HAM-A, implying its continuity with depression and anxiety.

Hierarchical Multiple Regression

Table 4 presents the hierarchical multiple regression model for predicting PSQI, including HAM-A and overall beta EEG power as independent variables. The increase in the proportion of variance explained between Steps 1 and 2 indicates that overall beta power has additional predictive ability beyond the effect of anxiety (ΔR2 = 0.041, p = 0.018). The results also remained significant after HAM-D was included in the model (ΔR2 = 0.033, p = 0.032), indicating its predictive ability beyond the effects of anxiety and depression combined.

DISCUSSION

This study aimed to identify the relationship between EEG power spectra and insomnia after considering affective symptoms, such as depression and anxiety. Wake- resting-state EEG data were obtained during the daytime from patients with ID, in which comorbid neuropsychi-atric disorders were excluded. The ID group showed higher levels of depression, anxiety, and sleep disturbances than the HC group. Multiple regression analysis revealed that high-frequency EEG power had a significant ability to predict insomnia, beyond the effects of affective symptoms, supporting the hyperarousal theory of insomnia.

In this study, patients with ID had significantly higher levels of depression and anxiety than their healthy coun-terparts. Unsurprisingly, the present findings are consistent with results from previous studies that have reported a significant impact of depression/anxiety on in-somnia. Most importantly, the ID group showed a higher relative spectral power for overall beta and gamma power than the control group. Increased beta and gamma power are associated with increased wakefulness and cognitive process of anxiety [28,29], and these beta and gamma powers may induce worry, mind wandering, or rumination, if excessive. These EEG results (i.e., high-frequency power) may coincide well with the characteristics experienced by patients with insomnia [30]. In addition, EEG power was positively correlated with PSQI: This result is consistent with previous studies indicating that high-frequency EEG power is a good marker for hyperarousal in patients with insomnia [14,15]. In particular, such tendency occurred globally (i.e., throughout the brain) rather than locally, which is also consistent with existing studies, suggesting that hyperarousal is a phenomenon that spans the entire brain-nervous system rather than a focal one that appears only in a specific area [14].

In this study, we attempted to delve deeper into the hyperarousal model of insomnia by controlling for depressive and anxiety symptoms in the wake-resting state EEG and adopting a multiple regression model predicting insomnia symptoms. Here, the effect of anxiety on insomnia is noteworthy; it still had a strong predictive ability even after participants diagnosed with anxiety disorder were excluded through the MINI interview. This result implies that anxiety is germane to arousal [31,32]. Indeed, the level of anxiety not only reflects hyperarousal but also strongly predicts insomnia and overuse of sleep aids [33]. Anxiety disorders precede insomnia, and anxiety disorders occur after insomnia [17,34]. However, it is difficult to consider that anxiety and insomnia only share those identical mechanisms. As explained above, anxiety and affective disorders are governed by different brain networks: Whereas affective symptoms are linked to the salience network centered on the amygdala and anterior cingulate cortex, objective sleep disturbance is usually controlled by the default mode network centered on the posterior cingulate cortex [16]. The relationship between insomnia and anxiety should be further investigated in future studies.

Nevertheless, the additional explanatory power revealed in the hierarchical multiple regression analysis indicates that EEG power has its own ability to predict the severity of insomnia beyond the mere byproduct of anxiety or depression. The effect size of this unique predictive ability was not large (rpartial = 0.230 and 0.211); however, the overall relative spectral power of beta frequencies accounted for significant amounts of insomnia symptoms in the regression model. This result is meaningful because the increase in beta power is consistent with the hyperarousal model of insomnia [8]. Although high-frequency activities such as beta power indicate arousal states by itself, to our knowledge, this study was a first attempt to reveal that such hyperarousal contains different attributes from the affective symptoms or anxiety. Overall, both affective symptoms and objective EEG spectrum predicted a relatively high amount of variance in insomnia symptoms (R2 = 0.276 or 0.293), suggesting the significance of both emotional arousal due to anxiety and (neurological) arousal of the central nervous system, which may originate from genetic vulnerability and physiological change [35].

It is also noteworthy that the strong high-frequency EEG power of patients with insomnia was observed in the waking state. Originally, the normal sleep process was characterized by a decrease in high-frequency power and an increase in low-frequency power [36]. However, previous studies have reported that patients with insomnia tend to show stronger high-frequency power during REM and non-REM sleep, which is significantly associated with subjective sleep difficulties [7,37]. Combining these studies with the current results, we can speculate that the increase in high-frequency power (hyperarousal) among patients with insomnia may appear 24 hours a day, not only at night [38]. As supported in a study using positron emission tomography, patients with insomnia have increased global cerebral glucose metabolism during the day as well as at night [39]. Such phenomena can also be explained by dysfunction of homeostasis. Cortical awake should be accelerated during daytime arousal but inhibited at night to maintain 24-hour sleep homeostasis [40]. However, if this homeostatic cycle is interrupted, it may hinder cortical excitability and sleep disturbance [12,41]. From this perspective, insomnia needs to be defined not only as a sleep disorder but also as a 24-hour hyperarousal disorder [42].

The novelty of this study was that we showed the existence of the hyperarousal distinguishable from the affective symptoms and detectable as a pattern of brain activity or spectral power. Perhaps the next step would be to figure out whether this distinct pattern can indicate the difference in the brain network system, as described above. Unfortunately, to better investigate the brain network, electroencephalography is not enough because it is not suitable to look into the activity of subcortical areas such as anterior/posterior cingulate cortices or limbic system, which plays key roles in the salience network and the default mode network. Therefore, future studies may adopt different measures such as functional magnetic resonance imaging (fMRI) or different analyses such as functional connectivity or transfer entropy.

This research is a meaningful advance from previous studies; however, some limitations remain. First, the present study suffered from the limitations of cross-sectional studies. We were not able to obtain detailed sleep and psychiatric histories; hence, it was impossible to establish a causative relationship between EEG markers and insom-nia. However, to our knowledge, this is the first study to evaluate and control for possible correlates (e.g., sex, depression, and anxiety) to maximize the likelihood of an unbiased estimate of the associations between insomnia and wake-resting-state EEG markers. Future prospective epidemiologic studies of sleep with sequential neurophysiological correlates, laboratory testing, and psychological interviews are needed. Second, PSG was not assessed in this study, although this study mainly focused on waking state activity. Adopting PSG along with the waking-state EEG may provide useful information about the cycle of sleep, as well as insomnia [43]. Finally, this study covered only some of the clinical symptoms related to insomnia, such as HAM-A and HAM-D. Insomnia may be related to many subtle episodes that cannot be fully covered through affective measures alone. Therefore, in future studies, it is necessary to adopt comprehensive methods with more diverse clinical components or measures, including various domains of cognitive function or episodic reports such as daytime diary [44], quality of life [45], the Daily Stress Inventory, or the Pre-Sleep Arousal Questionnaire [46].

The current study defined insomnia disorder using validated and structured interviews and controlled for possible confounders to obtain unbiased estimates of the relationship between insomnia and resting EEG markers. We further elucidated the significant relationship between beta EEG activity and anxiety and insomnia symptoms, after adjusting for other clinical correlates. The present findings provide further evidence for the hyperarousal theory of insomnia and imply that a simple 5-minute, daytime, wake- resting EEG could be applicable as a time-and-cost efficient alternative tool to evaluate insomnia regardless of the implementation of PSG.

Conflicts of Interest

No potential conflict of interest relevant to this article was reported.

Author Contributions

Formal analysis: Gyutae Jang, Han Wool Jung, Jiheon Kim. Investigation: Hansol Kim. Resources: Jiheon Kim, Hansol Kim. Validation: Do-Hoon Kim, Sang-Kyu Lee. Methodology: Daeyoung Roh. Project administration: Daeyoung Roh. Funding acquisition: Daeyoung Roh. Conceptualization: Ji-Hyeon Shin, Chan-Hyung Kim, Daeyoung Roh. Supervision: Ji-Hyeon Shin, Chan-Hyung Kim. Writing―original draft: Gyutae Jang. Writing―review & editing: Han Wool Jung.

Figures
Fig. 1. Comparison of ID and HC groups in the standardized EEG relative power. Regions with signi-ficant differences in brain activity between the ID and HC groups are marked with dotted lines.
ID, insomnia disorder; HC, healthy control; EEG, electroencephalography.
Tables

Demographics and clinical measures

Variable Insomnia disorder (n = 58) Healthy control (n = 49) t or χ2 pvalue
Age 27.6 ± 9.87 26.0 ± 8.36 −0.90 0.372
Sex 45 (78) 31 (63) 2.65 0.104
HAM-D 11.97 ± 3.61 8.29 ± 4.73 −4.56** 0.000
HAM-A 12.91 ± 5.09 8.31 ± 5.37 −4.55** 0.000
PSQI 11.40 ± 2.38 5.96 ± 1.63 −13.96** 0.000

Values are presented as mean ± standard deviation or number (%).

HAM-D, Hamilton Depression Rating Scale; HAM-A, Hamilton Anxiety Rating Scale; PSQI, Pittsburgh Sleep Quality Index.

**p < 0.001.

Group differences in the EEG relative power

Band Insomnia disorder (n = 58) Healthy control (n = 49) t pvalue
Delta (1−4 Hz) 0.316 ± 0.105 0.299 ± 0.094 −0.87 0.387
Theta (4−8 Hz) 0.141 ± 0.031 0.149 ± 0.034 1.14 0.257
Alpha (8−12 Hz) 0.278 ± 0.151 0.322 ± 0.145 1.54 0.126
Beta (12−30 Hz) 0.199 ± 0.062 0.170 ± 0.052 −2.56* 0.012
Gamma (30−40 Hz) 0.049 ± 0.024 0.039 ± 0.019 −2.44* 0.016

Values are presented as mean ± standard deviation.

EEG, electroencephalography.

*p < 0.05.

Correlations between clinical measures and the EEG relative power

Variable PSQI Age HAM-D HAM-A Delta Theta Alpha Beta Gamma
PSQI
Age 0.214*
HAM-D 0.492*** 0.109
HAM-A 0.500*** 0.098 0.787***
Delta 0.145 0.021 0.232* 0.156
Theta −0.065 −0.131 0.078 0.033 0.307**
Alpha −0.189 −0.274** −0.326*** −0.213* −0.712*** −0.348***
Beta 0.275** 0.333*** 0.210* 0.157 −0.087 0.065 −0.406***
Gamma 0.246* 0.113 0.269** 0.222* 0.196* −0.037 −0.438*** 0.388***

EEG, electroencephalography; PSQI, Pittsburgh Sleep Quality Index; HAM-D, Hamilton Depression Rating Scale; HAM-A, Hamilton Anxiety Rating Scale.

p < 0.10, *p < 0.05, **p < 0.01, ***p < 0.001.

Hierarchical multiple regression analysis for PSQI (n = 107)

Regression model Predictor β rpartial R2 ΔR2 pfor ΔR2
Model 1 (anxiety) Step 1
HAM-A 0.486*** 0.486 0.236 0.236*** 0.000
Step 2
HAM-A 0.454*** 0.466 0.276 0.041* 0.018
Beta 0.204* 0.230
Model 2 (anxiety + depression) Step 1
HAM-A 0.292* 0.208 0.260 0.260*** 0.000
HAM-D 0.248 0.177
Step 2
HAM-A 0.296* 0.215 0.293 0.033* 0.032
HAM-D 0.206 0.150
Beta 0.185* 0.211

Dependent variable: Pittsburgh Sleep Quality Index.

HAM-A, Hamilton Anxiety Rating Scale; HAM-D, Hamilton Anxiety Rating Scale; β, standardized coefficient; rpartial, partial correlation.

p < 0.10, *p < 0.05, ***p < 0.001.

References
  1. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 5th ed. American Psychiatric Association Publishing;2013.
    CrossRef
  2. Grima NA, Bei B, Mansfield D. Insomnia theory and assess-ment. Aust J Gen Pract 2019;48:193-197.
    Pubmed CrossRef
  3. Javaheri S, Redline S. Insomnia and risk of cardiovascular disease. Chest 2017;152:435-444.
    Pubmed KoreaMed CrossRef
  4. Taylor DJ, Lichstein KL, Durrence HH, Reidel BW, Bush AJ. Epidemiology of insomnia, depression, and anxiety. Sleep 2005;28:1457-1464.
    Pubmed CrossRef
  5. Roberts RE, Shema SJ, Kaplan GA, Strawbridge WJ. Sleep complaints and depression in an aging cohort: A prospective perspective. Am J Psychiatry 2000;157:81-88.
    Pubmed CrossRef
  6. Fortier-Brochu E, Beaulieu-Bonneau S, Ivers H, Morin CM. Insomnia and daytime cognitive performance: A meta-analysis. Sleep Med Rev 2012;16:83-94.
    Pubmed CrossRef
  7. Perlis ML, Giles DE, Mendelson WB, Bootzin RR, Wyatt JK. Psychophysiological insomnia: The behavioural model and a neurocognitive perspective. J Sleep Res 1997;6:179-188.
    Pubmed CrossRef
  8. Riemann D, Spiegelhalder K, Feige B, Voderholzer U, Berger M, Perlis M, et al. The hyperarousal model of insomnia: A review of the concept and its evidence. Sleep Med Rev 2010;14:19-31.
    Pubmed CrossRef
  9. Alonso JF, Romero S, Ballester MR, Antonijoan RM, Mañanas MA. Stress assessment based on EEG univariate features and functional connectivity measures. Physiol Meas 2015;36:1351-1365.
    Pubmed CrossRef
  10. Jun G, Smitha KG. EEG based stress level identification. In: 2016 IEEE International Conference on Systems, Man, and Cybernetics (SMC) 2016 Oct 9-12:003270-003274.
    CrossRef
  11. Bonnet MH, Arand DL. Hyperarousal and insomnia: State of the science. Sleep Med Rev 2010;14:9-15.
    Pubmed CrossRef
  12. Colombo MA, Wei Y, Ramautar JR, Linkenkaer-Hansen K, Tagliazucchi E, Van Someren EJ. More severe insomnia complaints in people with stronger long-range temporal correlations in wake resting-state EEG. Front Physiol 2016;7:576.
    Pubmed KoreaMed CrossRef
  13. Kwan Y, Baek C, Chung S, Kim TH, Choi S. Resting-state quantitative EEG characteristics of insomniac patients with depres-sion. Int J Psychophysiol 2018;124:26-32.
    Pubmed CrossRef
  14. Colombo MA, Ramautar JR, Wei Y, Gomez-Herrero G, Stoffers D, Wassing R, et al. Wake high-density electroencephalographic spatiospectral signatures of insomnia. Sleep 2016;39:1015-1027.
    Pubmed KoreaMed CrossRef
  15. Wolynczyk-Gmaj D, Szelenberger W. Waking EEG in primary insomnia. Acta Neurobiol Exp (Wars) 2011;71:387-392.
    Pubmed CrossRef
  16. Khazaie H, Veronese M, Noori K, Emamian F, Zarei M, Ashkan K, et al. Functional reorganization in obstructive sleep apnoea and insomnia: A systematic review of the resting-state fMRI. Neurosci Biobehav Rev 2017;77:219-231.
    Pubmed KoreaMed CrossRef
  17. Neckelmann D, Mykletun A, Dahl AA. Chronic insomnia as a risk factor for developing anxiety and depression. Sleep 2007;30:873-880.
    Pubmed KoreaMed CrossRef
  18. Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, et al. The Mini-International Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry 1998;59 Suppl 20:22-33; quiz 34-57.
  19. Hamilton M. Hamilton anxiety scale [Internet]. Psychiatric University Hospital Zurich; 1959 [cited at 2021 Oct 13]. https://psycnet.apa.org/doi/10.1037/t02824-000.
    CrossRef
  20. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry 1960;23:56-62.
    Pubmed KoreaMed CrossRef
  21. Buysse DJ, Reynolds CF 3rd, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: A new instrument for psychiatric practice and research. Psychiatry Res 1989;28:193-213.
    Pubmed CrossRef
  22. Sohn SI, Kim DH, Lee MY, Cho YW. The reliability and validity of the Korean version of the Pittsburgh Sleep Quality Index. Sleep Breath 2012;16:803-812.
    Pubmed CrossRef
  23. Carpenter JS, Andrykowski MA. Psychometric evaluation of the Pittsburgh Sleep Quality Index. J Psychosom Res 1998;45:5-13.
    Pubmed CrossRef
  24. Sartorius N, Ban TA. Assessment of depression. Springer;1986.
    KoreaMed CrossRef
  25. Kim C. Psychiatric assessment instruments. Hana Medical Publisher;2001.
  26. Yi JS, Bae SO, Ahn YM, Park DB, Noh KS, Shin HK, et al. Validity and reliability of the Korean version of the Hamilton depression rating scale (K-HDRS). J Korean Neuropsychiatr Assoc 2005;44:456-465.
  27. Barry RJ, Clarke AR, Johnstone SJ, Magee CA, Rushby JA. EEG differences between eyes-closed and eyes-open resting conditions. Clin Neurophysiol 2007;118:2765-2773.
    Pubmed CrossRef
  28. Oathes DJ, Ray WJ, Yamasaki AS, Borkovec TD, Castonguay LG, Newman MG, et al. Worry, generalized anxiety disorder, and emotion: Evidence from the EEG gamma band. Biol Psychol 2008;79:165-170.
    Pubmed KoreaMed CrossRef
  29. Pavlenko VB, Chernyi SV, Goubkina DG. EEG correlates of anxiety and emotional stability in adult healthy subjects. Neurophysiology 2009;41:337-345.
    CrossRef
  30. Harvey AG, Tang NK, Browning L. Cognitive approaches to insomnia. Clin Psychol Rev 2005;25:593-611.
    Pubmed CrossRef
  31. Alkozei A, Creswell C, Cooper PJ, Allen JJ. Autonomic arousal in childhood anxiety disorders: Associations with state anxiety and social anxiety disorder. J Affect Disord 2015;175:25-33.
    Pubmed KoreaMed CrossRef
  32. Brown TA, Chorpita BF, Barlow DH. Structural relationships among dimensions of the DSM-IV anxiety and mood disorders and dimensions of negative affect, positive affect, and autonomic arousal. J Abnorm Psychol 1998;107:179-192.
    Pubmed CrossRef
  33. Pillai V, Cheng P, Kalmbach DA, Roehrs T, Roth T, Drake CL. Prevalence and predictors of prescription sleep aid use among individuals with DSM-5 insomnia: the role of hyperarousal. Sleep 2016;39:825-832.
    Pubmed KoreaMed CrossRef
  34. Johnson EO, Roth T, Breslau N. The association of insomnia with anxiety disorders and depression: Exploration of the direction of risk. J Psychiatr Res 2006;40:700-708.
    Pubmed CrossRef
  35. Basta M, Chrousos GP, Vela-Bueno A, Vgontzas AN. Chronic insomnia and stress system. Sleep Med Clin 2007;2:279-291.
    Pubmed KoreaMed CrossRef
  36. De Gennaro L, Ferrara M, Bertini M. The boundary between wakefulness and sleep: Quantitative electroencephalographic changes during the sleep onset period. Neuroscience 2001;107:1-11.
    Pubmed CrossRef
  37. Zhao W, Van Someren EJW, Li C, Chen X, Gui W, Tian Y, et al. EEG spectral analysis in insomnia disorder: A systematic review and meta-analysis. Sleep Med Rev 2021;59:101457.
    Pubmed CrossRef
  38. Corsi-Cabrera M, Rojas-Ramos OA, del Río-Portilla Y. Waking EEG signs of non-restoring sleep in primary insomnia patients. Clin Neurophysiol 2016;127:1813-1821.
    Pubmed CrossRef
  39. Nofzinger EA, Buysse DJ, Germain A, Price JC, Miewald JM, Kupfer DJ. Functional neuroimaging evidence for hyperarousal in insomnia. Am J Psychiatry 2004;161:2126-2128.
    Pubmed CrossRef
  40. Huber R, Mäki H, Rosanova M, Casarotto S, Canali P, Casali AG, et al. Human cortical excitability increases with time awake. Cereb Cortex 2013;23:332-338.
    Pubmed KoreaMed CrossRef
  41. Bastien CH, St-Jean G, Morin CM, Turcotte I, Carrier J. Chronic psychophysiological insomnia: Hyperarousal and/or inhibition deficits? An ERPs investigation. Sleep 2008;31:887-898.
    Pubmed KoreaMed CrossRef
  42. Cortoos A, Verstraeten E, Cluydts R. Neurophysiological aspects of primary insomnia: Implications for its treatment. Sleep Med Rev 2006;10:255-266.
    Pubmed CrossRef
  43. Reite M, Buysse D, Reynolds C, Mendelson W. The use of polysomnography in the evaluation of insomnia. Sleep 1995;18:58-70.
    Pubmed CrossRef
  44. Orff HJ, Drummond SP, Nowakowski S, Perils ML. Discrepancy between subjective symptomatology and objective neuropsychological performance in insomnia. Sleep 2007;30:1205-1211.
    Pubmed KoreaMed CrossRef
  45. McCall WV, Reboussin BA, Cohen W. Subjective measurement of insomnia and quality of life in depressed inpatients. J Sleep Res 2000;9:43-48.
    Pubmed CrossRef
  46. Morin CM, Rodrigue S, Ivers H. Role of stress, arousal, and coping skills in primary insomnia. Psychosom Med 2003;65:259-267.
    Pubmed CrossRef


This Article

Close ✕


Cited By Articles
  • CrossRef (0)
  • Scopus (0)
  • Download (369)

Author ORCID Information

Funding Information

Services
Social Network Service

e-submission

Archives