Sleep Quality and Electroencephalogram Delta Power

Delta activity on electroencephalogram (EEG) is considered a biomarker of homeostatic sleep drive. Delta power is often associated with sleep duration and intensity. Here, we reviewed the literature to explore how sleep quality was influenced by changes in delta power. However, we found that both the decrease and increase in delta power could indicate a higher sleep quality due to the various factors below. First, the differences in changes in delta power in patients whose sleep quality is lower than that of the healthy controls may be related to the different diseases they suffered from. We found that the patients mainly suffered from borderline personality disorder, and Rett syndrome may have a higher delta power than healthy individuals. Meanwhile, patients who are affected by Asperger syndrome, respiratory failure, chronic fatigue, and post-traumatic stress disorder have lower delta power. Second, if the insomnia patients received the therapy, the difference may be caused by the treatment method. Cognitive or music therapy shows that a better therapeutic effect is associated with decreased delta power, whereas in drug treatment, there is an opposite change in delta power. Last, for healthy people, the difference in delta change may be related to sleep stages. The higher sleep quality is associated with increased delta power during the NREM period, whereas a deceased delta change accompanies higher sleep quality during the REM period. Our work summarizes the effect of changes in delta power on sleep quality and may positively impact the monitoring and intervention of sleep quality.

Delta activity on electroencephalogram (EEG) is considered a biomarker of homeostatic sleep drive. Delta power is often associated with sleep duration and intensity. Here, we reviewed the literature to explore how sleep quality was influenced by changes in delta power. However, we found that both the decrease and increase in delta power could indicate a higher sleep quality due to the various factors below. First, the differences in changes in delta power in patients whose sleep quality is lower than that of the healthy controls may be related to the different diseases they suffered from. We found that the patients mainly suffered from borderline personality disorder, and Rett syndrome may have a higher delta power than healthy individuals. Meanwhile, patients who are affected by Asperger syndrome, respiratory failure, chronic fatigue, and post-traumatic stress disorder have lower delta power. Second, if the insomnia patients received the therapy, the difference may be caused by the treatment method. Cognitive or music therapy shows that a better therapeutic effect is associated with decreased delta power, whereas in drug treatment, there is an opposite change in delta power. Last, for healthy people, the difference in delta change may be related to sleep stages. The higher sleep quality is associated with increased delta power during the NREM period, whereas a deceased delta change accompanies higher sleep quality during the REM period. Our work summarizes the effect of changes in delta power on sleep quality and may positively impact the monitoring and intervention of sleep quality.
Keywords: delta power, sleep quality, EEG, sleep stage, sleep disorder Delta oscillation reflects the burst-pause firing pattern of the hyperpolarized thalamic cortex and corticothalamic neurons in synchronization (Steriade et al., 1993). Delta activity can be measured by delta power by Fourier analysis. It is a biomarker of homeostatic sleep drive, as evidence shows that delta power is enhanced after prolonged wakefulness, but declines as sleep deepens (Feinberg, 1974;Borbély et al., 1981;Achermann and Borbély, 2011). One study showed that after 40 h of sleep deprivation, both young and old subjects showed a significant increase in the power spectral density in the delta band (Münch et al., 2004). The finding that delta power is high during wake time but decreases during sleep (Esteban et al., 2005) was supported by healthy participants from various age groups as well as sleep-deprived people. A study of infants, adolescents, and adults showed that both children and adults experienced a gradual decrease in delta power and slow-wave sleep throughout the night (Bes et al., 1991). Sleep analysis of sleep-deprived young adults showed the gradual decline of electroencephalogram (EEG) power density in the low-frequency range as an indicator during sleep (Borbély et al., 1981).
In addition, delta power is often accompanied by sleep duration and intensity (Davis et al., 2011). For example, after a lack of sleep, EEG delta power and non-rapid eye movement (NREM) duration during NREM episodes will increase (Pappenheimer et al., 1975;Achermann, 2004;Davis et al., 2011). Thus, delta power is a crucial part of the sleep process. We review some literature related to the change in delta power influencing sleep quality. However, current research on the relationship between delta power and sleep quality is inconclusive (Gao et al., 2020;Johnson and Durrant, 2021).
There is a lack of analysis on which factors lead to these opposite aspects. We list out the studies related to the change in delta power, including the advantages of decreased delta power and increased delta power (Tables 1, 2). We summarize the factors that may influence the adverse delta change. A wide range of factors, including the type, age, number, and sex of participants, the period of delta power changes, EEG derivation, and the power calculation method, may result in different conclusion.
We summarize the factors that may have influenced the contrary results. First, we infer that the category of the disease may influence the quality of sleep from various aspects, which may have different impacts on EEG power. We found that the patients in Table 1 mainly suffered from borderline personality disorder (BPD), Rett syndrome (RTT) (Lindberg et al., 2003;Philipsen et al., 2005;Ammanuel et al., 2015). A study of female patients with BPD who often have trouble sleeping showed that the delta power of BPD patients increased during NREM and REM sleep. Meanwhile, subjective scores, which mainly use the Pittsburgh Sleep Quality Index (PSQI) (Buysse et al., 1989), recorded a significant decline in sleep quality among BPD patients (Philipsen et al., 2005). It has also been suggested that sleep problems that present as reduced sleep duration, a longer incubation period, and interrupted sleep in girls (2-9 years old) with RTT were associated with higher delta frequency (Ammanuel et al., 2015). However, Table 2 mainly shows the patients who were affected by Asperger syndrome, respiratory failure, chronic fatigue, and post-traumatic stress disorder (PTSD) (Tani et al.'s, 2004;Guilleminault et al.'s, 2006;Kudchadkar et al., 2015;Wang et al., 2020). Tani et al.'s (2004) research mainly compared patients suffering from Asperger syndrome (27.2 ± 7.3) with healthy individuals and found a decreased delta power spectrum trend in patients during the SWS stage. The patients also had subjective impairment in the initiation and continuity of sleep. A study showed that children with respiratory failure who were associated with disruption of the normal sleep-wake rhythm had lower delta power than healthy children (Kudchadkar et al., 2015). Guilleminault et al.'s (2006) study compared the difference between chronic fatigue and normal and found that slow-wave sleep (SWS) percentage and sleep efficiency were lower, but there was a significant decrease in delta 2 relative power in the chronic fatigue group when compared to normal. Wang et al.'s (2020) study showed that compared with non-PTSD patients, PTSD patients exhibited reduced delta power during NREM sleep with reduced sleep depth. Second, the difference may be related to the method of treatment. Different therapies influence sleep quality in diverse ways. In studies of insomnia patients treated by cognitive behavioral therapy (CBT), the results showed that the continuous decrease in delta power was related to the treatment effect, and the rate of decrease in delta power was also used as the criterion of sleep drive (Feinberg, 1974;Krystal and Edinger, 2010). In addition, our previous study also found that SWS brain-wave music (Wu et al., 2009;Lu et al., 2012;Yao et al., 2020) leads to shorter sleep latency and higher sleep efficiency with decreased delta power (Gao et al., 2020). However, in Monti et al.'s (2000) study, power density in NREM sleep increased significantly in the low-frequency band (0.25-1.0 Hz) in the zolpidem group. We found that patients who received cognitive/music treatment and drug therapy had adverse delta changes (Monti et al.'s, 2000;Krystal and Edinger, 2010). Last, the period of delta change may influence the final results. We found from healthy participants that when the change in delta power occurred in the NREM period, better sleep quality was associated with increased delta power, which was proven in Kuula et al.'s (2020) study. They found that participants who received slow breathing training with higher sleep quality had higher delta power during N3, which is a deep sleep stage that lasts approximately 20-40 min. There is also evidence showing that increases in the subjective stress burden were associated with decreases in delta power during NREM (Hall et al., 2000). Research in Park et al.'s (2021) found that exercise can lead to a short rapid eye movement latency and significantly increased delta power in SWS with enhanced SWS stability in early sleep phases. However, Gabryelska et al.'s (2019) study showed different situations in which better sleep quality was significantly associated with decreased Delta 1 power density in REM. Considering this fact, we speculate that the period of delta change may influence the impact of delta change.
Besides, we also listed some parameters about the delta power spectrum processing (Tables 1, 2) and wondered if these factors include the EEG duration per epoch, filter frequency, and the fast Fourier transform (FFT) window length might affect the final results. However, we did not find the common characteristic in each condition. Therefore, these parameters do not influence the relationship between delta power change and sleep quality.
The delta power is like a marker whose change can detect the sleep quality. The current research we have conducted mainly speculated the possible reason for the increase or the decrease of delta power indicates the better sleep quality. However, they are some speculations that we conclude by comparing the different studies. In the future, we should further explore the physiology mechanism behind these factors influencing the change of delta power and further influence sleep quality, which will provide more reliable evidence for the current study and draw a more precise conclusion on how the shift in delta power influences sleep quality.
In summary, we summarize the studies related to the change in delta power associated with sleep quality. We inferred that some factors might influence the difference in the impact of changes in delta power. We hope our research may contribute to a deeper understanding of the impact of delta power on monitoring and intervention of sleep quality.

AUTHOR CONTRIBUTIONS
SL, RD, JL, and DY conceived the study and drafted the manuscript. JW and YY collected the information. All authors made critical revisions to the work, which ultimately approved the upcoming edition, and agreed to be responsible for all aspects of the work.

FUNDING
This work was supported by the Project of Science and Technology Department of Sichuan Province (Nos. 2021YFS0135 and 2020ZYD013) and the Research Unit of NeuroInformation, Chinese Academy of Medical Sciences (No. 2019RU035).