TY - JOUR
AU1 - Walter, Lisa, M
AU2 - Tamanyan,, Knarik
AU3 - Weichard, Aidan, J
AU4 - Biggs, Sarah, N
AU5 - Davey, Margot, J
AU6 - Nixon, Gillian, M
AU7 - Horne, Rosemary S, C
AB - Abstract Study Objectives Sleep disordered breathing (SDB) in children has significant effects on daytime functioning and cardiovascular control; attributed to sleep fragmentation and repetitive hypoxia. Associations between electroencephalograph (EEG) spectral power, autonomic cardiovascular control and cerebral oxygenation have been identified in adults with SDB. To date, there have been no studies in children. We aimed to assess associations between EEG spectral power and heart rate variability as a measure of autonomic control, with cerebral oxygenation in children with SDB. Methods One hundred sixteen children (3–12 years) with SDB and 42 controls underwent overnight polysomnography including measurement of cerebral oxygenation. Power spectral analysis of the EEG derived from C4-M1 and F4-M1, quantified delta, theta, alpha, and beta waveforms during sleep. Multiple regression tested whether age, SDB severity, heart rate (HR), HR variability (HRV), and cerebral oxygenation were determinants of EEG spectral power. Results There were no differences in EEG spectral power derived from either central or frontal regions for any frequency between children with different severities of SDB so these were combined. Age, HR, and HRV low frequency power were significant determinants of EEG spectral power depending on brain region and sleep stage. Conclusion The significant findings of this study were that age and autonomic control, rather than cerebral oxygenation and SDB severity, were predictive of EEG spectral power in children. Further research is needed to elucidate how the physiology that underlies the relationship between autonomic control and EEG impacts on the cardiovascular sequelae in children with SDB. EEG spectral power, cerebral oxygenation, autonomic control, pediatric, sleep disordered breathing, obstructive sleep apnea Statement of Significance Sleep disordered breathing (SDB), a common condition in children. This is the first study to investigate the relationship between age, SDB severity, cerebral oxygenation, and heart rate variability as a measure of autonomic control on EEG power spectral analysis in children with SDB and nonsnoring controls. SDB severity and cerebral oxygenation were not predictive of EEG spectral power. Age, heart rate, and heart rate variability were significant determinants of the EEG power spectra. This study highlights the importance of the early detection of SDB in children before the adverse cardiovascular outcomes associated with SDB impact on brain activity. Introduction Sleep disordered breathing (SDB) is a common condition with a reported prevalence of 4%–11% in children [1]. We and others have previously shown that children with all severities of SDB have adverse neurocognitive [2], behavioral [3], and cardiovascular [4–6] outcomes. These sequelae are believed to result from repeated hypoxia and reperfusion and surges in heart rate (HR) and blood pressure associated with the obstructive respiratory events that characterize the condition. Power spectral analysis of the EEG obtained during an overnight polysomnography (PSG) study quantifies the delta, theta, alpha, and beta waveforms that occur during sleep, in the frequency domain (EEG spectral power), and provides a measurement of sleep micro-architecture [7–10]. We have previously reported differences in the distribution of EEG spectral power over the night between SDB severity groups in elementary school-aged children with SDB [10]. Furthermore, studies have shown that spectral characteristics of the EEG in adults [11], adolescents [9], and children [12] exhibit regional differences. Studies in infants [13–15] and adults [16] have identified associations between EEG activity and cerebral blood flow and cerebral oxygenation. In adults, cerebral blood flow was correlated with slow wave activity, a marker of sleep intensity, during NREM sleep [16]. In preterm infants during quiet sleep (the precursor to NREM sleep), spontaneous bursts of EEG activity were coupled to a stereotypical hemodynamic response measured by near infrared spectroscopy (NIRS) [14]. Furthermore, more mature EEG activity in premature infants has been associated with increased [15] and less variable [13] fractional tissue oxygenation extraction (FTOE), indicating an association between EEG activity and cerebral oxygenation. Similar studies examining the relationship between cerebral oxygenation and EEG activity have not been conducted in children. During sleep there are coordinated cortical and cardiac oscillations which reflect the communication and interdependency between the central (CNS) and autonomic (ANS) nervous systems [17]. An association between EEG spectral analysis (reflecting CNS) and autonomic cardiovascular control (reflecting ANS), as measured by heart rate variability (HRV), has been identified in adults, with and without SDB [18–21]. However, we have recently identified that SDB significantly disrupts the positive association between autonomic cardiovascular control and cerebral oxygenation found in nonsnoring children [22]. Whether the diminished relationship between cerebral oxygenation and autonomic cardiovascular control in children with SDB affects the relationship between sleep micro-architecture and cerebral oxygenation in children remains to be elucidated. Thus, this study aimed to determine whether there were differences in sleep micro-architecture between children from different SDB severity groups and nonsnoring controls, determined at the frontal and central brain regions. Furthermore, we aimed to determine the association between sleep micro-architecture, autonomic cardiovascular control, and cerebral oxygenation in children with and without SDB, and whether there were regional differences in these associations. We hypothesized that sleep micro-architecture would be different between SDB severity groups. We further hypothesized that cerebral oxygenation and heart rate variability would be predictive of EEG spectral power, and would exhibit regional differences. Methods Ethical approval for this study was obtained from the Monash University and Monash Health Human Research Ethics Committees (14024B). Written consent was obtained from parents and verbal assent from children. Children (3–12 years) referred for clinical assessment of SDB at the Melbourne Children’s Sleep Centre between May 2014 and December 2016 were approached to participate. Children were otherwise healthy with no comorbidities, such as craniofacial syndromes, developmental disability, and genetic syndromes, and were not taking any medications known to affect breathing or sleep. Age-matched control children, with no previous history of snoring, were recruited from the community. All children underwent overnight attended PSG. Height and weight were measured and converted to a body mass index (BMI) z-score to adjust for gender and age [23]. Electrophysiological signals were recorded using a commercially available PSG system (E-Series, Compumedics, Melbourne, Australia) using standard pediatric recording techniques [24]. Electrodes for recording left and right electrooculogram, submental electromyogram, left and right anterior tibialis muscle electromyogram, and electrocardiogram (ECG) were attached. Thoracic and abdominal breathing movements were detected using respiratory inductance plethysmography (Pro-Tech zRIP Effort Sensor, Pro-Tech Services, Inc., Mukilteo, WA). Transcutaneous carbon dioxide (TcCO2, TCM4/40, Radiometer, Denmark, Copenhagen), nasal pressure, and oronasal airflow were also recorded. Oxygen saturation (SpO2) was measured using Bitmos GmbH (Bitmos, Dusseldorf, Germany), which uses Masimo signal extraction technology for signal processing and was set to a 2-s averaging time. In addition to the standard PSG electrodes, cerebral oxygenation was measured using a NIRO 200NX spectrophotometer (Hamamatsu Photonics KK, Tokyo, Japan) positioned on the forehead, avoiding any hair. The sensors were covered with cloth to shield from any external light, and held in place with an elastic net bandage (Tubular-Net, Sutherland Medical, Victoria, Australia). Near-infrared spectroscopy enables calculation of cerebral tissue oxygenation index (TOI) using continuous-wave light emission and detection measured over the frontal region of the brain, with the detection probe placed 4 cm away from the emission probe. TOI was computed continuously using a spatially resolved spectroscopy algorithm and represents mixed oxygen saturations of all cerebral vascular compartments. Cerebral fractional tissue oxygenation extraction (FTOE) was calculated using: FTOE = (SpO2 −TOI)/SpO2 [25, 26]. TOI and FTOE methods have been previously published [27]. TOI reflects the saturation of oxygen in veins (70%–80%), capillaries (5%), and arteries (20%–25%), and as such is an estimate of cerebral oxygen delivery [28]. The FTOE describes the balance between cerebral oxygen delivery and cerebral oxygen consumption from an estimate of the amount of oxygen extracted by brain tissue from the vascular compartment [28]. FTOE accounts for arterial oxygen saturation and therefore provides a ratio of cerebral oxygen consumption to delivery [25]. Pediatric sleep technologists sleep-staged and scored all PSG studies manually in 30 s epochs of NREM stages N1, N2, and N3, and REM sleep according to clinical practice at the time of the study [24, 29]. Obstructive apneas were defined as a >90% fall in airflow for ≥90% of event duration, with continued or increased respiratory effort. Mixed apneas consisted of a central component followed by an obstructive component. An obstructive hypopnea was associated with a ≥50% fall in airflow signal for at least ≥90% of the event, associated with an arousal, awakening or ≥3% desaturation. Respiratory event-related arousals (RERAs) were scored where there was a discernible decrease in amplitude and flattening of the nasal pressure trace, associated with snoring, noisy breathing, elevation of the end-tidal or transcutaneous pCO2 and/or visual evidence of increased work of breathing, leading to an arousal from sleep or ≥3% desaturation. A minimum of 4 h of sleep was required to assess SDB severity for children to be included in the study. Conventional sleep macro-architecture parameters from the PSG analyzed for this study were: the percentage of total sleep time (TST) spent in each sleep stage, wake after sleep onset (WASO) defined as the percentage of time awake during the sleep period time (SPT), with SPT defined as the amount of time in minutes from sleep onset until lights on at the end of the study, and TST defined as SPT excluding all periods of wake. Other variables calculated included sleep onset latency (SOL) defined as the period from lights off to the first three consecutive epochs of N1 sleep or an epoch of any other stage, sleep efficiency (SE) defined as the time spent asleep as a percentage of the time available for sleep, and the arousal index (ArI) defined as the number of arousals per hour of TST. Respiratory variables included the obstructive apnea hypopnea index (OAHI), which was defined as the total number of obstructive apneas, mixed apneas, obstructive hypopneas, and respiratory event-related arousals per hour of TST, the central apnea hypopnea index (CnAHI) defined as the number of central apneas and hypopneas per hour of TST, SpO2 nadir, the ODI4 defined as the number of times the SpO2 dropped by greater than 4% per hour of TST, ODI90 defined as the number of times the SpO2 dropped below 90% per hour of TST, and the average transcutaneous CO2 during TST. Following scoring of the PSG, the children were grouped according to their OAHI into SDB severity groups: PS, OAHI ≤ 1 event per hour TST; Mild OSA, OAHI > 1–5 events per hour TST; MS OSA, OAHI > 5 events per hour TST. The controls had an OAHI ≤ 1 event per hour TST and no snoring noted on the PSG. For sleep micro-architecture and HRV analysis, PSG data were transferred via European Data Format to data analysis software (LabChart 7.1, ADIinstruments, Sydney, Australia). Sleep micro-architecture was assessed using spectral analysis of the EEG signal [30]. Raw EEG signals were recorded using a band-pass filter ranging from 0.3 to 100 Hz and a sampling frequency of 512 Hz. Spectral analysis was performed on the EEG channels C4-M1 and F4-M1. To remove any low or high frequency artifact from the signal, the entire EEG time series was digitally filtered using a band-pass filter ranging from 0.5 to 30 Hz. Thirty-second epochs containing significant artifact, defined as a 30-s epoch containing >10 s of movement artifact that interrupted the EEG signal, were excluded from analysis. In all studies, the first epoch of the recording was deleted, as well as the last epoch if it did not run for the full 30 s. The following frequency bands were set as delta (0.5–3.9 Hz), theta (4–7.9 Hz), alpha (8–11.9 Hz), sigma (12–13.9 Hz), and β power (14–30 Hz) [10]. Spectral analysis was run using a fast Fourier transform (FFT) size of 1024 over the entire PSG recording with a Hanning window, which allowed edge effects to be avoided [31]. The FFT output provided a total power for 2-s blocks with a frequency resolution of 0.5 Hz. These 0.5-Hz frequency bins were subsequently summed within five frequency bands, producing a single power value for each band. In addition, total power for each 2-s block was determined (0.5–30 Hz). As EEG power is known to be affected by sleep stage in children [10], a mean value for each frequency was calculated for each 30-s epoch and then averaged per sleep stage within each child for values derived from C4 and F4. Heart rate variability analysis Every 2 min epoch from the entire overnight study (including respiratory events) that was free of movement artifact (disruption to the ECG signal caused by gross body movement) on the ECG signal was selected. Two minute epochs were chosen to maximize the overall number of artifact-free epochs available for analysis, while ensuring an adequate length of time to accommodate enough oscillations within the low frequency (LF) range to detect changes. Periods of wakefulness during the sleep period were excluded. Each 2 min epoch was separated from the previous epoch by at least 30s [5]. The intervening 30 s epochs were ECG artifact free. The 2-min epochs were then grouped into wake before sleep onset and all epochs of sleep across the night and a mean value for each state (wake and sleep) calculated for each subject. The power spectral density for the LF (0.04–0.15 Hz) reflecting a mixture of both parasympathetic and sympathetic activity and high frequency (HF) (0.15–0.4 Hz) reflecting parasympathetic activity, bands was determined [32]. Total power (TP) reflects overall automatic activity. The LF/HF ratio was determined as a measure of sympathovagal balance. Statistical analysis Statistical analyses were performed using SPSS (IBM Statistics, version 22). Data were first tested for normality and equal variance. Demographic, sleep macro-architecture, and respiratory data were compared between SDB severity groups using one-way ANOVAs followed by Bonferroni post hoc testing for normally distributed data and a Kruskal–Wallis test followed by Mann–Whitney post hoc testing for data that was not normally distributed. Statistical significance was taken at p <0.05. EEG spectral frequencies were compared between SDB severity groups and sleep stages using linear mixed model analyses to allow for random effects over repeated observations. Group (control, PS, mild OSA, MS OSA) and sleep stage (N1, N2, N3, REM) were entered as fixed effects. Age and BMI z-score were entered as covariates. Subject code was used as the random factor. As no differences between groups were identified, groups were combined for subsequent analyses. The Benjamini and Hochberg False Discovery Rate method was used to correct for multiple testing and statistical significance was taken at p ≤0.01. Significant correlates of the EEG spectral power parameters were identified and used as the independent variables in multiple linear regression analyses, which were performed to determine significant predictors of total, delta, theta, alpha, sigma, and beta power derived from C4 and F4 during N1, N2, N3, and REM sleep. The independent variables were age, OAHI, BMI z-score, TOI, FTOE, and the HRV parameters (HR, TP, LF, HF, LF/HF). Total, delta, theta, alpha, sigma, and beta power were entered as the dependant variables. FTOE and HRV TP were found to be collinear with TOI and the remaining HRV parameters respectively, which was resolved when FTOE and HRV TP were removed from the analysis. Multiple linear regression mediation models were conducted to determine if the effect of heart rate on spectral EEG power was mediated by cerebral (TOI) or peripheral (ODI) oxygenation. Results Demographic, sleep macro-architecture, and respiratory data are presented in Table 1. There were no differences in the age or the proportion of females between the SDB severity groups. The control children had a significantly lower BMI z-score compared with the children in all SDB groups. There were no significant differences between groups for any of the sleep variables, except the ArI was significantly higher in the children with MS OSA compared with children from the other SDB groups and controls. By study design, the OAHI was significantly higher in the mild OSA and MS OSA groups compared with the PS and control groups. OAHI in the MS OSA group was also significantly higher than in the mild OSA group. CnAHI, SpO2 nadir, ODI90, and ODI4 were significantly higher in the children with MS OSA compared with the other groups, with no differences between control, PS, and Mild OSA groups. Table 1. Demographic, sleep macro-architecture and respiratory data from children (3–12 years) with sleep disordered breathing and nonsnoring controls Control, N = 42 PS, N = 37 Mild OSA, N = 42 MS OSA, N = 37 % female 55 46 43 50 Age years 7.1 (3.2 to 12.1) 7.3 (3.3 to 12.9) 6.0 (3.1 to 12.7) 6.3 (3.5 to 12.9) BMI Z-score 0.32 (−2.02 to 1.94)a 0.84 (−0.79 to 2.86)b 0.96 (−1.15 to 3.6)b 0.77 (−0.26 to 4.48)b TST min 438 (348 to 534) 446 (269 to 496) 434 (294 to 513) 432 (333 to 559) SPT 472 (391 to 551) 477 (404 to 553) 480 (373 to 537) 475 (381 to 566) SE % 86 (74 to 97) 90 (53 to 95) 86 (55 to 96) 87 (69 to 98) SOL min 25 (0 to 106) 24 (5 to 90) 21 (0 to 117) 24 (4 to 118) WASO %SPT 6 (2 to 16) 6 (1 to 44) 8 (1 to 37) 6 (1 to 27) % N1 6 (3 to 12) 6 (1 to 23) 6 (1 to 19) 6 (1 to 25) % N2 48 (35 to 59) 45 (23 to 59) 46 (36 to 58) 42 (30 to 53) % N3 27 (17 to 40) 30 (14 to 48) 28 (16 to 40) 31 (22 to 42) %NREM 80 (71 to 88) 80 (73 to 89) 81 (72 to 93) 79 (72 to 87) %REM 20 (12 to 29) 20 (11 to 27) 19 (7 to 28) 21 (13 to 28) OAHI 0.1 (0.0 to 0.9)a 0.3 (0.0 to 1.0)a 1.9 (1.1 to 4.3)b 10.7 (5.2 to 48.8)c CnAHI 1.2 (0.1 to 5.5)a 1.1 (0.0 to 10.6)a 1.2 (0.1 to 6.3)a 2.2 (0.0 to 12.2)b SpO2 nadir 93 (81 to 97)a 93 (82 to 96)a 93 (76 to 96)a 91 (70 to 96)b ODI90 0.0 (0.0 to 0.4)a 0.0 (0.0 to 0.3)a 0.3 (0.0 to 0.6)a 6.2 (0.0 to 12.1)b ODI4 0.5 (0.0 to 3.6)a 0.3 (0.0 to 5.0)a 0.5 (0.0 to 6.3)a 2.8 (0.0 to 37.7)b Av TCO2 43.9 (5.7 to 61.0) 42.3 (33.3 to 54.0) 42.5 (34.1 to 54.7) 42.2 (28.7 to 51.5) ArI 10.7 (6.3 to 20.7)a 10.1 (5.6 to 21.8)a 11.8 (6.8 to 15.5)a 16.8 (5.6 to 38.3)b Control, N = 42 PS, N = 37 Mild OSA, N = 42 MS OSA, N = 37 % female 55 46 43 50 Age years 7.1 (3.2 to 12.1) 7.3 (3.3 to 12.9) 6.0 (3.1 to 12.7) 6.3 (3.5 to 12.9) BMI Z-score 0.32 (−2.02 to 1.94)a 0.84 (−0.79 to 2.86)b 0.96 (−1.15 to 3.6)b 0.77 (−0.26 to 4.48)b TST min 438 (348 to 534) 446 (269 to 496) 434 (294 to 513) 432 (333 to 559) SPT 472 (391 to 551) 477 (404 to 553) 480 (373 to 537) 475 (381 to 566) SE % 86 (74 to 97) 90 (53 to 95) 86 (55 to 96) 87 (69 to 98) SOL min 25 (0 to 106) 24 (5 to 90) 21 (0 to 117) 24 (4 to 118) WASO %SPT 6 (2 to 16) 6 (1 to 44) 8 (1 to 37) 6 (1 to 27) % N1 6 (3 to 12) 6 (1 to 23) 6 (1 to 19) 6 (1 to 25) % N2 48 (35 to 59) 45 (23 to 59) 46 (36 to 58) 42 (30 to 53) % N3 27 (17 to 40) 30 (14 to 48) 28 (16 to 40) 31 (22 to 42) %NREM 80 (71 to 88) 80 (73 to 89) 81 (72 to 93) 79 (72 to 87) %REM 20 (12 to 29) 20 (11 to 27) 19 (7 to 28) 21 (13 to 28) OAHI 0.1 (0.0 to 0.9)a 0.3 (0.0 to 1.0)a 1.9 (1.1 to 4.3)b 10.7 (5.2 to 48.8)c CnAHI 1.2 (0.1 to 5.5)a 1.1 (0.0 to 10.6)a 1.2 (0.1 to 6.3)a 2.2 (0.0 to 12.2)b SpO2 nadir 93 (81 to 97)a 93 (82 to 96)a 93 (76 to 96)a 91 (70 to 96)b ODI90 0.0 (0.0 to 0.4)a 0.0 (0.0 to 0.3)a 0.3 (0.0 to 0.6)a 6.2 (0.0 to 12.1)b ODI4 0.5 (0.0 to 3.6)a 0.3 (0.0 to 5.0)a 0.5 (0.0 to 6.3)a 2.8 (0.0 to 37.7)b Av TCO2 43.9 (5.7 to 61.0) 42.3 (33.3 to 54.0) 42.5 (34.1 to 54.7) 42.2 (28.7 to 51.5) ArI 10.7 (6.3 to 20.7)a 10.1 (5.6 to 21.8)a 11.8 (6.8 to 15.5)a 16.8 (5.6 to 38.3)b PS = primary snoring; OSA = obstructive sleep apnea; MS = moderate/severe; BMI = body mass index; TST = total sleep time; SPT = sleep period time; SE = sleep efficiency; SOL = sleep onset latency; WASO = wake after sleep onset; NREM = nonrapid eye movement sleep; N1 = NREM stage 1; N2 = NREM stage 2; N3 = NREM stage 3; REM = rapid eye movement sleep; OAHI = obstructive sleep apnea index; CnAHI = central apnea hypopnea index; SpO2 = arterial oxygen saturation; ODI90 = the number of oxygen desaturations below 90% per hour TST; ODI4 = the number of oxygen desaturations greater than 4% per hour sleep; AvTCO2 = average transcutaneous carbon dioxide; ArI = arousal index. Data presented as median (min–max). Columns that do not share a letter in common are significantly different, p < 0.05. Open in new tab Table 1. Demographic, sleep macro-architecture and respiratory data from children (3–12 years) with sleep disordered breathing and nonsnoring controls Control, N = 42 PS, N = 37 Mild OSA, N = 42 MS OSA, N = 37 % female 55 46 43 50 Age years 7.1 (3.2 to 12.1) 7.3 (3.3 to 12.9) 6.0 (3.1 to 12.7) 6.3 (3.5 to 12.9) BMI Z-score 0.32 (−2.02 to 1.94)a 0.84 (−0.79 to 2.86)b 0.96 (−1.15 to 3.6)b 0.77 (−0.26 to 4.48)b TST min 438 (348 to 534) 446 (269 to 496) 434 (294 to 513) 432 (333 to 559) SPT 472 (391 to 551) 477 (404 to 553) 480 (373 to 537) 475 (381 to 566) SE % 86 (74 to 97) 90 (53 to 95) 86 (55 to 96) 87 (69 to 98) SOL min 25 (0 to 106) 24 (5 to 90) 21 (0 to 117) 24 (4 to 118) WASO %SPT 6 (2 to 16) 6 (1 to 44) 8 (1 to 37) 6 (1 to 27) % N1 6 (3 to 12) 6 (1 to 23) 6 (1 to 19) 6 (1 to 25) % N2 48 (35 to 59) 45 (23 to 59) 46 (36 to 58) 42 (30 to 53) % N3 27 (17 to 40) 30 (14 to 48) 28 (16 to 40) 31 (22 to 42) %NREM 80 (71 to 88) 80 (73 to 89) 81 (72 to 93) 79 (72 to 87) %REM 20 (12 to 29) 20 (11 to 27) 19 (7 to 28) 21 (13 to 28) OAHI 0.1 (0.0 to 0.9)a 0.3 (0.0 to 1.0)a 1.9 (1.1 to 4.3)b 10.7 (5.2 to 48.8)c CnAHI 1.2 (0.1 to 5.5)a 1.1 (0.0 to 10.6)a 1.2 (0.1 to 6.3)a 2.2 (0.0 to 12.2)b SpO2 nadir 93 (81 to 97)a 93 (82 to 96)a 93 (76 to 96)a 91 (70 to 96)b ODI90 0.0 (0.0 to 0.4)a 0.0 (0.0 to 0.3)a 0.3 (0.0 to 0.6)a 6.2 (0.0 to 12.1)b ODI4 0.5 (0.0 to 3.6)a 0.3 (0.0 to 5.0)a 0.5 (0.0 to 6.3)a 2.8 (0.0 to 37.7)b Av TCO2 43.9 (5.7 to 61.0) 42.3 (33.3 to 54.0) 42.5 (34.1 to 54.7) 42.2 (28.7 to 51.5) ArI 10.7 (6.3 to 20.7)a 10.1 (5.6 to 21.8)a 11.8 (6.8 to 15.5)a 16.8 (5.6 to 38.3)b Control, N = 42 PS, N = 37 Mild OSA, N = 42 MS OSA, N = 37 % female 55 46 43 50 Age years 7.1 (3.2 to 12.1) 7.3 (3.3 to 12.9) 6.0 (3.1 to 12.7) 6.3 (3.5 to 12.9) BMI Z-score 0.32 (−2.02 to 1.94)a 0.84 (−0.79 to 2.86)b 0.96 (−1.15 to 3.6)b 0.77 (−0.26 to 4.48)b TST min 438 (348 to 534) 446 (269 to 496) 434 (294 to 513) 432 (333 to 559) SPT 472 (391 to 551) 477 (404 to 553) 480 (373 to 537) 475 (381 to 566) SE % 86 (74 to 97) 90 (53 to 95) 86 (55 to 96) 87 (69 to 98) SOL min 25 (0 to 106) 24 (5 to 90) 21 (0 to 117) 24 (4 to 118) WASO %SPT 6 (2 to 16) 6 (1 to 44) 8 (1 to 37) 6 (1 to 27) % N1 6 (3 to 12) 6 (1 to 23) 6 (1 to 19) 6 (1 to 25) % N2 48 (35 to 59) 45 (23 to 59) 46 (36 to 58) 42 (30 to 53) % N3 27 (17 to 40) 30 (14 to 48) 28 (16 to 40) 31 (22 to 42) %NREM 80 (71 to 88) 80 (73 to 89) 81 (72 to 93) 79 (72 to 87) %REM 20 (12 to 29) 20 (11 to 27) 19 (7 to 28) 21 (13 to 28) OAHI 0.1 (0.0 to 0.9)a 0.3 (0.0 to 1.0)a 1.9 (1.1 to 4.3)b 10.7 (5.2 to 48.8)c CnAHI 1.2 (0.1 to 5.5)a 1.1 (0.0 to 10.6)a 1.2 (0.1 to 6.3)a 2.2 (0.0 to 12.2)b SpO2 nadir 93 (81 to 97)a 93 (82 to 96)a 93 (76 to 96)a 91 (70 to 96)b ODI90 0.0 (0.0 to 0.4)a 0.0 (0.0 to 0.3)a 0.3 (0.0 to 0.6)a 6.2 (0.0 to 12.1)b ODI4 0.5 (0.0 to 3.6)a 0.3 (0.0 to 5.0)a 0.5 (0.0 to 6.3)a 2.8 (0.0 to 37.7)b Av TCO2 43.9 (5.7 to 61.0) 42.3 (33.3 to 54.0) 42.5 (34.1 to 54.7) 42.2 (28.7 to 51.5) ArI 10.7 (6.3 to 20.7)a 10.1 (5.6 to 21.8)a 11.8 (6.8 to 15.5)a 16.8 (5.6 to 38.3)b PS = primary snoring; OSA = obstructive sleep apnea; MS = moderate/severe; BMI = body mass index; TST = total sleep time; SPT = sleep period time; SE = sleep efficiency; SOL = sleep onset latency; WASO = wake after sleep onset; NREM = nonrapid eye movement sleep; N1 = NREM stage 1; N2 = NREM stage 2; N3 = NREM stage 3; REM = rapid eye movement sleep; OAHI = obstructive sleep apnea index; CnAHI = central apnea hypopnea index; SpO2 = arterial oxygen saturation; ODI90 = the number of oxygen desaturations below 90% per hour TST; ODI4 = the number of oxygen desaturations greater than 4% per hour sleep; AvTCO2 = average transcutaneous carbon dioxide; ArI = arousal index. Data presented as median (min–max). Columns that do not share a letter in common are significantly different, p < 0.05. Open in new tab Figure 1 compares the sleep micro-architecture between SDB severity groups during N1, N2, N3, and REM sleep derived from C4-M1. Sleep micro-architecture was compared between SDB severity groups during N1, N2, N3, and REM sleep. The mixed model analysis revealed that there was no main effect of SDB severity group or significant interaction between group and sleep stage for any EEG frequency. Figure 1. Open in new tabDownload slide C4 derived EEG power spectral analysis for (a) total power, (b) delta power, (c) theta power, (d) alpha power, (e) sigma power, and (f) beta power, during N1, N2, N3, and REM sleep in nonsnoring controls and children with SDB. White columns represent the control group; light gray the PS group; dark gray the mild OSA group; and black the MS OSA group. Data presented as mean (SD). Figure 1. Open in new tabDownload slide C4 derived EEG power spectral analysis for (a) total power, (b) delta power, (c) theta power, (d) alpha power, (e) sigma power, and (f) beta power, during N1, N2, N3, and REM sleep in nonsnoring controls and children with SDB. White columns represent the control group; light gray the PS group; dark gray the mild OSA group; and black the MS OSA group. Data presented as mean (SD). Figure 2 compares the sleep micro-architecture between SDB severity groups during N1, N2, N3, and REM sleep derived from F4-M1. Sleep micro-architecture was compared between SDB severity groups during N1, N2, N3, and REM sleep. The mixed model analysis revealed that there was no main effect of SDB severity group or significant interaction between group and sleep stage. Figure 2. Open in new tabDownload slide F4 derived EEG power spectral analysis for (a) total power, (b) delta power, (c) theta power, (d) alpha power, (e) sigma power, and (f) beta power, during N1, N2, N3, and REM sleep in nonsnoring controls and children with SDB. White columns represent the control group; light gray the PS group; dark gray the mild OSA group; and black the MS OSA group. Data presented as mean (SD). Figure 2. Open in new tabDownload slide F4 derived EEG power spectral analysis for (a) total power, (b) delta power, (c) theta power, (d) alpha power, (e) sigma power, and (f) beta power, during N1, N2, N3, and REM sleep in nonsnoring controls and children with SDB. White columns represent the control group; light gray the PS group; dark gray the mild OSA group; and black the MS OSA group. Data presented as mean (SD). Multiple stepwise linear regression revealed differences in the significant determinants of the EEG spectral power parameters that were both sleep stage and brain region specific. Significant determinants of total, delta, theta, alpha, sigma, and beta power during the sleep stages when derived from C4-M1, are presented in Table 2. OAHI, TOI, BMI Z-score, and HRV LF/HF ratio were not significant determinants of EEG spectral power for any frequency when derived from C4. Age was the predominant determinant of EEG spectral power. Increasing age significantly predicted decreasing total EEG power during N1, N2, and REM; decreasing delta power during N1 and REM; decreasing theta power during N2, N3, and REM; and increasing sigma power during N2 and N3. Increasing heart rate significantly determined decreasing total EEG power during N2, N3, and REM; decreasing delta power during N2, N3, and REM; and decreasing theta power during REM. Increasing HRV LF power significantly predicted decreasing total EEG power during N1, and decreasing theta power during N3. HRV HF power was a significant determinant of total EEG power during N1. Table 2. Significant determinants of total, delta, theta, alpha, sigma, and beta power derived from C4 during N1, N2, N3, and REM sleep in the combined group of children with and without SDB children R2 B β 95% CI of B P Total power  N1 0.33   Age −54.57 −0.41 −80.78, −28.33 <0.001   LF power −0.04 −0.30 −0.07, −0.01 0.008   HF power 0.02 0.26 0.001, 0.04 0.01  N2 0.14   Age −27.05 −0.21 −47.23, −6.87 0.009   HR −11.20 −0.34 −17.58, −4.81 0.001  N3 0.11   HR −22.97 −0.20 −42.62, −3.33 0.01  REM 0.20   Age −33.78 −0.28 −53.00, −14.58 0.001   HR −8.32 −0.28 −14.04, −2.61 0.005 Delta power  N1 0.28   Age −43.55 −0.37 −67.89, −19.21 0.001   LF −0.04 −0.32 −0.07, −0.01 0.008  N2 0.15   HR −9.75 −0.37 −14.77, −4.72 <0.001  N3 0.10   HR −20.58 −0.19 −38.97, −2.19 0.01  REM 0.15   Age −23.95 −0.21 −42.07, −5.83 0.01   HR −7.13 −0.26 −12.52, −1.73 0.01 Theta power  N2 0.22   Age −12.67 −0.43 −17.01, −8.33 <0.001  N3 0.17   Age −17.92 −0.26 −28.39, −7.45 0.001   LF power 0.03 0.27 0.003, 0.061 0.01  REM 0.17   Age −6.40 −0.28 −10.02, −2.79 0.001 R2 B β 95% CI of B P Total power  N1 0.33   Age −54.57 −0.41 −80.78, −28.33 <0.001   LF power −0.04 −0.30 −0.07, −0.01 0.008   HF power 0.02 0.26 0.001, 0.04 0.01  N2 0.14   Age −27.05 −0.21 −47.23, −6.87 0.009   HR −11.20 −0.34 −17.58, −4.81 0.001  N3 0.11   HR −22.97 −0.20 −42.62, −3.33 0.01  REM 0.20   Age −33.78 −0.28 −53.00, −14.58 0.001   HR −8.32 −0.28 −14.04, −2.61 0.005 Delta power  N1 0.28   Age −43.55 −0.37 −67.89, −19.21 0.001   LF −0.04 −0.32 −0.07, −0.01 0.008  N2 0.15   HR −9.75 −0.37 −14.77, −4.72 <0.001  N3 0.10   HR −20.58 −0.19 −38.97, −2.19 0.01  REM 0.15   Age −23.95 −0.21 −42.07, −5.83 0.01   HR −7.13 −0.26 −12.52, −1.73 0.01 Theta power  N2 0.22   Age −12.67 −0.43 −17.01, −8.33 <0.001  N3 0.17   Age −17.92 −0.26 −28.39, −7.45 0.001   LF power 0.03 0.27 0.003, 0.061 0.01  REM 0.17   Age −6.40 −0.28 −10.02, −2.79 0.001 LF = low frequency; HF = high frequency; HR = heart rate. Open in new tab Table 2. Significant determinants of total, delta, theta, alpha, sigma, and beta power derived from C4 during N1, N2, N3, and REM sleep in the combined group of children with and without SDB children R2 B β 95% CI of B P Total power  N1 0.33   Age −54.57 −0.41 −80.78, −28.33 <0.001   LF power −0.04 −0.30 −0.07, −0.01 0.008   HF power 0.02 0.26 0.001, 0.04 0.01  N2 0.14   Age −27.05 −0.21 −47.23, −6.87 0.009   HR −11.20 −0.34 −17.58, −4.81 0.001  N3 0.11   HR −22.97 −0.20 −42.62, −3.33 0.01  REM 0.20   Age −33.78 −0.28 −53.00, −14.58 0.001   HR −8.32 −0.28 −14.04, −2.61 0.005 Delta power  N1 0.28   Age −43.55 −0.37 −67.89, −19.21 0.001   LF −0.04 −0.32 −0.07, −0.01 0.008  N2 0.15   HR −9.75 −0.37 −14.77, −4.72 <0.001  N3 0.10   HR −20.58 −0.19 −38.97, −2.19 0.01  REM 0.15   Age −23.95 −0.21 −42.07, −5.83 0.01   HR −7.13 −0.26 −12.52, −1.73 0.01 Theta power  N2 0.22   Age −12.67 −0.43 −17.01, −8.33 <0.001  N3 0.17   Age −17.92 −0.26 −28.39, −7.45 0.001   LF power 0.03 0.27 0.003, 0.061 0.01  REM 0.17   Age −6.40 −0.28 −10.02, −2.79 0.001 R2 B β 95% CI of B P Total power  N1 0.33   Age −54.57 −0.41 −80.78, −28.33 <0.001   LF power −0.04 −0.30 −0.07, −0.01 0.008   HF power 0.02 0.26 0.001, 0.04 0.01  N2 0.14   Age −27.05 −0.21 −47.23, −6.87 0.009   HR −11.20 −0.34 −17.58, −4.81 0.001  N3 0.11   HR −22.97 −0.20 −42.62, −3.33 0.01  REM 0.20   Age −33.78 −0.28 −53.00, −14.58 0.001   HR −8.32 −0.28 −14.04, −2.61 0.005 Delta power  N1 0.28   Age −43.55 −0.37 −67.89, −19.21 0.001   LF −0.04 −0.32 −0.07, −0.01 0.008  N2 0.15   HR −9.75 −0.37 −14.77, −4.72 <0.001  N3 0.10   HR −20.58 −0.19 −38.97, −2.19 0.01  REM 0.15   Age −23.95 −0.21 −42.07, −5.83 0.01   HR −7.13 −0.26 −12.52, −1.73 0.01 Theta power  N2 0.22   Age −12.67 −0.43 −17.01, −8.33 <0.001  N3 0.17   Age −17.92 −0.26 −28.39, −7.45 0.001   LF power 0.03 0.27 0.003, 0.061 0.01  REM 0.17   Age −6.40 −0.28 −10.02, −2.79 0.001 LF = low frequency; HF = high frequency; HR = heart rate. Open in new tab Significant determinants of total, delta, theta, alpha, sigma, and beta power during the sleep stages when derived from F4-M1 are presented in Table 3. As identified for C4, OAHI and BMI Z-score were not significant determinants of EEG spectral power for any frequency when derived from F4. HRV HF power and LF/HF ratio were also not significant determinants of EEG spectral power for any frequency when derived from F4. As identified for C4, age was the predominant determinant of EEG spectral power for F4. Increasing age significantly predicted decreasing total EEG power during N1 and REM; decreasing delta power during N1 and REM; decreasing theta power during N1, N3, and REM; and decreasing sigma power during N2 and N3. Heart rate was a significant determinant of decreasing alpha power during N1. HRV LF power significantly predicted decreasing total EEG and delta power during N1, and increasing total EEG and delta power during N2; and increasing beta power during N2. Table 3. Significant determinants of total, delta, theta, alpha, sigma, and beta power derived from F4 during N1, N2, N3, and REM sleep in the combined group of children with and without SDB R2 B β 95% CI of B P Total power  N1 0.33   Age −61.37 −0.42 −90.64, −32.10 <0.001   LF power −0.04 −0.29 −0.08, −0.01 0.01  N2 0.10   LF power 0.14 0.28 0.03, 0.26 0.01  REM 0.16   Age −40.02 −0.27 −63.71, −16.33 0.001 Delta power  N1 0.30   Age −49.50 −0.40 −75.07, −23.94 <0.001   LF power −0.04 −0.29 −0.07, −0.01 0.01  N2 0.10   LF power 0.09 0.26 0.01, 0.18 0.01  REM 0.14   Age −32.94 −0.25 −54.34, −11.55 0.003 Theta power  N1 0.31   Age −10.96 −0.40 −16.48, −5.43 <0.001  N3 0.17   Age −20.10 −0.28 −31.01, −9.19 <0.001  REM 0.27   Age −7.44 −0.41 −10.19, −4.68 <0.001 Alpha power  N1 0.16   HR −0.28 −0.28 −0.52, −0.04 0.01 Sigma power  N2 0.22   Age 1.74 0.44 1.16, 2.32 <0.001  N3 0.10   Age 1.24 0.30 0.57, 1.92 <0.001 Beta power  N2 0.06   LF 0.001 0.24 0.00, 0.002 0.047 R2 B β 95% CI of B P Total power  N1 0.33   Age −61.37 −0.42 −90.64, −32.10 <0.001   LF power −0.04 −0.29 −0.08, −0.01 0.01  N2 0.10   LF power 0.14 0.28 0.03, 0.26 0.01  REM 0.16   Age −40.02 −0.27 −63.71, −16.33 0.001 Delta power  N1 0.30   Age −49.50 −0.40 −75.07, −23.94 <0.001   LF power −0.04 −0.29 −0.07, −0.01 0.01  N2 0.10   LF power 0.09 0.26 0.01, 0.18 0.01  REM 0.14   Age −32.94 −0.25 −54.34, −11.55 0.003 Theta power  N1 0.31   Age −10.96 −0.40 −16.48, −5.43 <0.001  N3 0.17   Age −20.10 −0.28 −31.01, −9.19 <0.001  REM 0.27   Age −7.44 −0.41 −10.19, −4.68 <0.001 Alpha power  N1 0.16   HR −0.28 −0.28 −0.52, −0.04 0.01 Sigma power  N2 0.22   Age 1.74 0.44 1.16, 2.32 <0.001  N3 0.10   Age 1.24 0.30 0.57, 1.92 <0.001 Beta power  N2 0.06   LF 0.001 0.24 0.00, 0.002 0.047 SDB = sleep disordered breathing; HR = heart rate; LF = HRV low frequency; TOI = tissue oxygenation index. Open in new tab Table 3. Significant determinants of total, delta, theta, alpha, sigma, and beta power derived from F4 during N1, N2, N3, and REM sleep in the combined group of children with and without SDB R2 B β 95% CI of B P Total power  N1 0.33   Age −61.37 −0.42 −90.64, −32.10 <0.001   LF power −0.04 −0.29 −0.08, −0.01 0.01  N2 0.10   LF power 0.14 0.28 0.03, 0.26 0.01  REM 0.16   Age −40.02 −0.27 −63.71, −16.33 0.001 Delta power  N1 0.30   Age −49.50 −0.40 −75.07, −23.94 <0.001   LF power −0.04 −0.29 −0.07, −0.01 0.01  N2 0.10   LF power 0.09 0.26 0.01, 0.18 0.01  REM 0.14   Age −32.94 −0.25 −54.34, −11.55 0.003 Theta power  N1 0.31   Age −10.96 −0.40 −16.48, −5.43 <0.001  N3 0.17   Age −20.10 −0.28 −31.01, −9.19 <0.001  REM 0.27   Age −7.44 −0.41 −10.19, −4.68 <0.001 Alpha power  N1 0.16   HR −0.28 −0.28 −0.52, −0.04 0.01 Sigma power  N2 0.22   Age 1.74 0.44 1.16, 2.32 <0.001  N3 0.10   Age 1.24 0.30 0.57, 1.92 <0.001 Beta power  N2 0.06   LF 0.001 0.24 0.00, 0.002 0.047 R2 B β 95% CI of B P Total power  N1 0.33   Age −61.37 −0.42 −90.64, −32.10 <0.001   LF power −0.04 −0.29 −0.08, −0.01 0.01  N2 0.10   LF power 0.14 0.28 0.03, 0.26 0.01  REM 0.16   Age −40.02 −0.27 −63.71, −16.33 0.001 Delta power  N1 0.30   Age −49.50 −0.40 −75.07, −23.94 <0.001   LF power −0.04 −0.29 −0.07, −0.01 0.01  N2 0.10   LF power 0.09 0.26 0.01, 0.18 0.01  REM 0.14   Age −32.94 −0.25 −54.34, −11.55 0.003 Theta power  N1 0.31   Age −10.96 −0.40 −16.48, −5.43 <0.001  N3 0.17   Age −20.10 −0.28 −31.01, −9.19 <0.001  REM 0.27   Age −7.44 −0.41 −10.19, −4.68 <0.001 Alpha power  N1 0.16   HR −0.28 −0.28 −0.52, −0.04 0.01 Sigma power  N2 0.22   Age 1.74 0.44 1.16, 2.32 <0.001  N3 0.10   Age 1.24 0.30 0.57, 1.92 <0.001 Beta power  N2 0.06   LF 0.001 0.24 0.00, 0.002 0.047 SDB = sleep disordered breathing; HR = heart rate; LF = HRV low frequency; TOI = tissue oxygenation index. Open in new tab TOI did not significantly predict EEG spectral power derived from either brain region. The effect of HR on EEG spectral power was not significantly altered when accounting for the indirect effect of either cerebral (TOI) or peripheral (ODI) oxygenation for any EEG frequency, or sleep stage, when determined at either C4 or F4. This indicates that the effect of HR on EEG spectral power was not mediated by cerebral or peripheral oxygenation. Discussion In this study, we have demonstrated that cerebral oxygenation and SDB severity are not significant determinants of EEG spectral power, which was predominately predicted by age, HR, and HRV. To our knowledge, this is the first study to analyze the association between autonomic control, cerebral oxygenation, and EEG spectral power in children with and without SDB. In contrast to studies in infants and adults, there was not a relationship between measures of cerebral oxygenation and EEG spectral power. Furthermore, also in contrast to studies in adults, SDB in children did not affect the relationship between autonomic activity and EEG spectral power. In contrast to studies in adults which have identified that SDB alters the tight coupling between the CNS and the ANS during sleep, we did not find any effect of SDB severity on EEG spectral power in the children studied. There have been a limited number of studies assessing the association between autonomic control and EEG spectral power in adults with SDB [18, 19, 33–36]. Coherence analysis is a method that characterizes the linear relationship between two signals in the frequency domain. A coherence value of one represents a very strong linear link and the gain value is the ratio between the amplitude of both signals. This method was used to analyze normalized HRV HF power and delta power in men with OSA and controls, and demonstrated decreased coherence and gain from controls to severe OSA [19]. The authors concluded that apneas and hypopneas affect the link between the autonomic control of HR and delta power. This difference in findings of a relationship between OSA severity and EEG spectral power in adult studies and our study which found no relationship, may reflect the difference in OSA severity between adults and children, whereby OSA is more severe in adults compared with children. Previous infant and adult studies led us to hypothesize that cerebral oxygenation would be predictive of EEG spectral power [13–16, 37]. Unexpectedly, in our cohort of children, cerebral oxygenation did not play a role in EEG spectral power. A number of studies concerning cerebral oxygenation and EEG spectral power have been conducted in neonates [13–15, 37]. Overall these studies present a picture of increasing maturation of EEG activity being associated with decreased variability in cerebral oxygen extraction in infants. Therefore, our findings would suggest that the brain had matured enough in our cohort of children, either with or without SDB, so that cerebral oxygenation no longer had a significant relationship with EEG activity. We have identified that the significant determinants of EEG spectral power vary between brain regions for specific sleep stages. However overall, in both brain regions, age, HR, and HRV parameters were the predominant determinants of EEG spectral power, with age generally being the strongest determinant of decreasing total, delta, theta, and alpha power, and of increasing sigma and beta power, depending on the sleep stage and brain region. A steady age-related reduction in EEG spectral power during NREM sleep has been demonstrated for delta, theta, alpha, sigma, and beta frequencies in healthy subjects across childhood, adolescence, and adulthood [38]. Early studies attribute these changes across childhood to either ontogenetic alterations in cortical synaptic density, which peak at around 10 years of age and then undergo reorganization during adolescence [39], or to a higher level of synchronization of cortical neurons [40]. In our study, in addition to age, HR and HRV LF power were also significant determinants of either decreasing or increasing EEG spectral power, which varied according to sleep stage, brain region, and the EEG frequency examined. Mediation analysis identified that neither cerebral nor peripheral oxygenation mediated the effect of HR and HRV on EEG spectral power. This is supported by our recent research which identified that SDB ameliorates the relationship between autonomic cardiovascular control and cerebral oxygenation found in nonsnoring children [22]. However, it must be noted that our model did not explain all of the variance in the EEG spectral power and other physiological parameters that were not included also have an impact on EEG spectral power. Cardiac activity has been demonstrated in humans to precede changes in EEG signals during sleep [41–43], and a raft of rodent and human studies have demonstrated an association between autonomic control of HR and EEG spectral power [20, 21, 43–47]. Studies in healthy adults [45, 48] and rats [44, 46] have reported that delta power was negatively correlated with the HRV LF/HF ratio, which reflects sympathovagal balance during NREM sleep. Delta power was also negatively correlated with HRV LF during NREM in humans [48]. A small pilot study reported that EEG delta, sigma, and beta bands exhibited a strong correlation with cardiac HRV parameters in different sleep stages in OSA patients [33]. This was confirmed in a later publication from the same authors, reporting data from 8 healthy adults and 11 OSA patients [18]. Similar to our study, there were differences in the correlations between the different EEG spectral frequencies and the HRV parameters for the different sleep stages. This was apparent not only in the strength of the correlations but also in the direction of the correlations. A recent review of publications reporting on CNS and ANS coupling during sleep in adults concluded that overall the data suggest that HRV measures mainly reflecting parasympathetic modulation fluctuate together with EEG delta power, with the cardiac changes temporally preceding cortical changes [17]. These different findings, may suggest fundamental differences between adults and children with regard to the relationship between autonomic control of the HR and CNS control of the EEG waveforms. The relationship between cardiac and cortical changes in the other EEG spectral frequencies remains to be elucidated in adults and our findings need confirmation with further research. EEG oscillations are the result of the interaction of large populations of neurons giving rise to rhythmic electrical events in the brain [49]. The frequency of these oscillations are determined not only by the overall activity of the brain, but also by the intrinsic properties of the neurons, and importantly, the function of the specific brain regions [50]. Therefore, to investigate regional differences in EEG spectral power, our data were analyzed for both the central and frontal brain regions. Our research group [51] and others [52–54] have reported changes to the brain in children with SDB, using a range of MRI technologies. The current study suggests that these changes to the brain are not reflected in changes to EEG spectral power. Our findings of no significant differences in sleep micro-architecture, between nonsnoring controls and the SDB severity groups during any sleep stage, for any EEG spectral power are similar to previous studies by ourselves and others. A study that included 40 young adults, mean age 24.3 ± 4.9 years, with OSA (apnea hypopnea index [AHI] >5 events/h), reported a higher AHI was associated with a lower slow- (11–13 Hz)-to-fast (13–17 Hz) sigma ratio, however, AHI was not correlated with delta, theta, alpha, beta slow sigma, or fast sigma power [55]. We found that SDB severity, as indicated by the OAHI, was not a significant determinant of EEG spectral power when derived from either brain region, confirming that SDB severity did not have a significant role in determining sleep micro-architecture in these children. This concurs with a previous study by our research group in 7- to 12-year-old children with SDB that also reported no significant differences in EEG spectral power between nonsnoring control children and children from all SDB severity groups [10]. However, there were significant differences in the distribution of EEG spectral power in delta, theta, and beta power as the night progressed between children with moderate/severe OSA and children with mild OSA, PS, or nonsnoring controls. The current study differs from the Yang et al. [10] study in that we included preschool-aged as well as 7- to 12-year-old children in the analysis. Other studies that have investigated EEG spectral power in children with SDB, focused their analyses on changes to spectral power surrounding respiratory events and arousals [7, 56], or respiratory cycle-related EEG changes [8, 57]. We acknowledge that there are limitations inherent to measuring EEG spectral power and HRV spectral power, including noise in the EEG and ECG signals. EEG and ECG signal noise is much more difficult to recognize in the power spectrum than in the raw signals. The source of noise in the signals can be either physiological in nature, or can be caused by electrode contact movement. However, our raw EEG and ECG signals underwent a rigorous manual data reduction process whereby all epochs containing artifact were removed from further analysis. Age was a common significant determinant of EEG spectral power, and it is a limitation of the study that pubertal status of the children was not recorded. Future research should include pubertal status and analyze the data in developmental groups. Conclusion Neither SDB severity nor cerebral oxygenation had a significant effect on EEG spectral power in the children in our study. However age, HR, and HRV, were differentially predictive of EEG spectral power according to the region of the brain, sleep stage and the EEG frequency being analyzed. While age is a nonmodifiable factor, HR and autonomic control can be modified. Therefore, this study highlights the importance of the early detection of SDB in children before the adverse cardiovascular outcomes associated with SDB impact on brain activity reflected in alterations to EEG spectral power. Further research is needed to elucidate how the physiology that underlies our findings impacts on the cardiovascular and neurocognitive sequelae in children with SDB. Funding This study was funded by The National Health and Medical Research Council of Australia APP1063500 and the Victorian Government’s Operational Infrastructure Support Program. Conflict of interest statement. None declared. Acknowledgment We would like to thank all the children and their parents who participated in this study and the staff of the Melbourne Children’s Sleep Centre where the studies were carried out. References 1. Lumeng JC , et al. Epidemiology of pediatric obstructive sleep apnea . Proc Am Thorac Soc. 2008 ; 5 ( 2 ): 242 – 252 . Google Scholar Crossref Search ADS PubMed WorldCat 2. Bourke R , et al. Cognitive and academic functions are impaired in children with all severities of sleep-disordered breathing . Sleep Med. 2011 ; 12 ( 5 ): 489 – 496 . Google Scholar Crossref Search ADS PubMed WorldCat 3. Bourke RS , et al. Neurobehavioral function is impaired in children with all severities of sleep disordered breathing . Sleep Med. 2011 ; 12 ( 3 ): 222 – 229 . Google Scholar Crossref Search ADS PubMed WorldCat 4. Horne RS , et al. Elevated blood pressure during sleep and wake in children with sleep-disordered breathing . Pediatrics. 2011 ; 128 ( 1 ): e85 – e92 . Google Scholar Crossref Search ADS PubMed WorldCat 5. Walter LM , et al. Autonomic dysfunction in children with sleep disordered breathing . Sleep Breath. 2013 ; 17 ( 2 ): 605 – 613 . Google Scholar Crossref Search ADS PubMed WorldCat 6. Walter LM , et al. Impaired blood pressure control in children with obstructive sleep apnea . Sleep Med. 2013 ; 14 ( 9 ): 858 – 866 . Google Scholar Crossref Search ADS PubMed WorldCat 7. Bandla HP , et al. Dynamic changes in EEG spectra during obstructive apnea in children . Pediatr Pulmonol. 2000 ; 29 ( 5 ): 359 – 365 . Google Scholar Crossref Search ADS PubMed WorldCat 8. Chervin RD , et al. Correlates of respiratory cycle-related EEG changes in children with sleep-disordered breathing . Sleep. 2004 ; 27 ( 1 ): 116 – 121 . Google Scholar Crossref Search ADS PubMed WorldCat 9. Jenni OG , et al. Regional differences of the sleep electroencephalogram in adolescents . J Sleep Res. 2005 ; 14 ( 2 ): 141 – 147 . Google Scholar Crossref Search ADS PubMed WorldCat 10. Yang JS , et al. Determining sleep quality in children with sleep disordered breathing: EEG spectral analysis compared with conventional polysomnography . Sleep. 2010 ; 33 ( 9 ): 1165 – 1172 . Google Scholar Crossref Search ADS PubMed WorldCat 11. Finelli LA , et al. Functional topography of the human nonREM sleep electroencephalogram . Eur J Neurosci. 2001 ; 13 ( 12 ): 2282 – 2290 . Google Scholar Crossref Search ADS PubMed WorldCat 12. Jussila M , et al. Sleep apnea reduces the amount of computational deep sleep in the right frontopolar area in school-aged children . Clin Neurophysiol. 2016 ; 127 ( 5 ): 2167 – 2174 . Google Scholar Crossref Search ADS PubMed WorldCat 13. El-Dib M , et al. EEG maturation and stability of cerebral oxygen extraction in very low birth weight infants . J Perinatol. 2016 ; 36 ( 4 ): 311 – 316 . Google Scholar Crossref Search ADS PubMed WorldCat 14. Roche-Labarbe N , et al. Coupled oxygenation oscillation measured by NIRS and intermittent cerebral activation on EEG in premature infants . Neuroimage. 2007 ; 36 ( 3 ): 718 – 727 . Google Scholar Crossref Search ADS PubMed WorldCat 15. ter Horst HJ , et al. The relationship between electrocerebral activity and cerebral fractional tissue oxygen extraction in preterm infants . Pediatr Res. 2011 ; 70 ( 4 ): 384 – 388 . Google Scholar Crossref Search ADS PubMed WorldCat 16. Tüshaus L , et al. In human non-REM sleep, more slow-wave activity leads to less blood flow in the prefrontal cortex . Sci Rep. 2017 ; 7 ( 1 ): 14993 . Google Scholar Crossref Search ADS PubMed WorldCat 17. de Zambotti M , et al. Dynamic coupling between the central and autonomic nervous systems during sleep: a review . Neurosci Biobehav Rev. 2018 ; 90 : 84 – 103 . Google Scholar Crossref Search ADS PubMed WorldCat 18. Abdullah H , et al. Cross-correlation of EEG frequency bands and heart rate variability for sleep apnoea classification . Med Biol Eng Comput. 2010 ; 48 ( 12 ): 1261 – 1269 . Google Scholar Crossref Search ADS PubMed WorldCat 19. Jurysta F , et al. The link between cardiac autonomic activity and sleep delta power is altered in men with sleep apnea-hypopnea syndrome . Am J Physiol Regul Integr Comp Physiol. 2006 ; 291 ( 4 ): R1165 – R1171 . Google Scholar Crossref Search ADS PubMed WorldCat 20. Jurysta F , et al. A study of the dynamic interactions between sleep EEG and heart rate variability in healthy young men . Clin Neurophysiol. 2003 ; 114 ( 11 ): 2146 – 2155 . Google Scholar Crossref Search ADS PubMed WorldCat 21. Pedemonte M , et al. Electroencephalographic frequencies associated with heart changes in RR interval variability during paradoxical sleep . Auton Neurosci. 2005 ; 123 ( 1–2 ): 82 – 86 . Google Scholar Crossref Search ADS PubMed WorldCat 22. Walter LM , et al. Sleep disordered breathing in children disrupts the maturation of autonomic control of heart rate and its association with cerebral oxygenation . J Physiol. 2019 ; 597 ( 3 ): 819 – 830 . Google Scholar Crossref Search ADS PubMed WorldCat 23. Ogden CL , et al. Centers for disease control and prevention 2000 growth charts for the United States: improvements to the 1977 National Center for Health Statistics version . Pediatrics. 2002 ; 109 ( 1 ): 45 – 60 . Google Scholar Crossref Search ADS PubMed WorldCat 24. Iber C , et al. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specification . 1st ed . Westchester, IL : American Academy of Sleep Medicine ; 2007 . Google Preview WorldCat COPAC 25. Naulaers G , et al. Use of tissue oxygenation index and fractional tissue oxygen extraction as non-invasive parameters for cerebral oxygenation. A validation study in piglets . Neonatology. 2007 ; 92 ( 2 ): 120 – 126 . Google Scholar Crossref Search ADS PubMed WorldCat 26. Pichler G , et al. Reference ranges for regional cerebral tissue oxygen saturation and fractional oxygen extraction in neonates during immediate transition after birth . J Pediatr. 2013 ; 163 ( 6 ): 1558 – 1563 . Google Scholar Crossref Search ADS PubMed WorldCat 27. Tamanyan K , et al. Age effects on cerebral oxygenation and behavior in children with sleep-disordered breathing . Am J Respir Crit Care Med. 2018 ; 197 ( 11 ): 1468 – 1477 . Google Scholar Crossref Search ADS PubMed WorldCat 28. van Bel F , et al. Monitoring neonatal regional cerebral oxygen saturation in clinical practice: value and pitfalls . Neonatology. 2008 ; 94 ( 4 ): 237 – 244 . Google Scholar Crossref Search ADS PubMed WorldCat 29. Nixon GM , et al. The impact of recent changes to the respiratory scoring rules in pediatrics . J Clin Sleep Med. 2014 ; 10 ( 11 ): 1217 – 1221 . Google Scholar PubMed WorldCat 30. Kellaway P. An Orderley Approach to Visual Analysis: Characteristics of the Normal EEG of Adults and Children . New York : Raven Press ; 1990 . Google Preview WorldCat COPAC 31. Jenni OG , et al. Spectral analysis of the sleep electroencephalogram during adolescence . Sleep. 2004 ; 27 ( 4 ): 774 – 783 . Google Scholar PubMed WorldCat 32. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology . Heart rate variability, standards of measurement, physiological interpretation, and clinical use . Eur Heart J . 1996 ; 17 : 354 – 381 . Crossref Search ADS PubMed WorldCat 33. Abdullah H , et al. Correlation of sleep EEG frequency bands and heart rate variability . Conf Proc IEEE Eng Med Biol Soc. 2009 ; 2009 : 5014 – 5017 . Google Scholar PubMed WorldCat 34. Jurysta F , et al. Long-term CPAP treatment partially improves the link between cardiac vagal influence and delta sleep . BMC Pulm Med. 2013 ; 13 : 29 . Google Scholar Crossref Search ADS PubMed WorldCat 35. Rothenberger SD , et al. Time-varying correlations between delta EEG power and heart rate variability in midlife women: the SWAN sleep study . Psychophysiology. 2015 ; 52 ( 4 ): 572 – 584 . Google Scholar Crossref Search ADS PubMed WorldCat 36. Thomas RJ , et al. Relationship between delta power and the electrocardiogram-derived cardiopulmonary spectrogram: possible implications for assessing the effectiveness of sleep . Sleep Med. 2014 ; 15 ( 1 ): 125 – 131 . Google Scholar Crossref Search ADS PubMed WorldCat 37. Victor S , et al. Relationship between blood pressure, cerebral electrical activity, cerebral fractional oxygen extraction, and peripheral blood flow in very low birth weight newborn infants . Pediatr Res. 2006 ; 59 ( 2 ): 314 – 319 . Google Scholar Crossref Search ADS PubMed WorldCat 38. Gaudreau H , et al. Age-related modifications of NREM sleep EEG: from childhood to middle age . J Sleep Res. 2001 ; 10 ( 3 ): 165 – 172 . Google Scholar Crossref Search ADS PubMed WorldCat 39. Feinberg I , et al. Gamma distribution model describes maturational curves for delta wave amplitude, cortical metabolic rate and synaptic density . J Theor Biol. 1990 ; 142 ( 2 ): 149 – 161 . Google Scholar Crossref Search ADS PubMed WorldCat 40. Aström C , et al. Relationship of age to power spectrum analysis of EEG during sleep . J Clin Neurophysiol. 1992 ; 9 ( 3 ): 424 – 430 . Google Scholar Crossref Search ADS PubMed WorldCat 41. Bonnet MH , et al. Heart rate variability: sleep stage, time of night, and arousal influences . Electroencephalogr Clin Neurophysiol. 1997 ; 102 ( 5 ): 390 – 396 . Google Scholar Crossref Search ADS PubMed WorldCat 42. Otzenberger H , et al. Dynamic heart rate variability: a tool for exploring sympathovagal balance continuously during sleep in men . Am J Physiol. 1998 ; 275 ( 3 ): H946 – H950 . Google Scholar PubMed WorldCat 43. Brandenberger G , et al. Inverse coupling between ultradian oscillations in delta wave activity and heart rate variability during sleep . Clin Neurophysiol. 2001 ; 112 ( 6 ): 992 – 996 . Google Scholar Crossref Search ADS PubMed WorldCat 44. Kuo TB , et al. Scatterplot analysis of EEG slow-wave magnitude and heart rate variability: an integrative exploration of cerebral cortical and autonomic functions . Sleep. 2004 ; 27 ( 4 ): 648 – 656 . Google Scholar Crossref Search ADS PubMed WorldCat 45. Yang CC , et al. Relationship between electroencephalogram slow-wave magnitude and heart rate variability during sleep in humans . Neurosci Lett. 2002 ; 329 ( 2 ): 213 – 216 . Google Scholar Crossref Search ADS PubMed WorldCat 46. Yang CC , et al. Relationship between electroencephalogram slow-wave magnitude and heart rate variability during sleep in rats . Neurosci Lett. 2003 ; 336 ( 1 ): 21 – 24 . Google Scholar Crossref Search ADS PubMed WorldCat 47. Brandenberger G , et al. Sleep stage 2: an electroencephalographic, autonomic, and hormonal duality . Sleep. 2005 ; 28 ( 12 ): 1535 – 1540 . Google Scholar Crossref Search ADS PubMed WorldCat 48. Ako M , et al. Correlation between electroencephalography and heart rate variability during sleep . Psychiatry Clin Neurosci. 2003 ; 57 ( 1 ): 59 – 65 . Google Scholar Crossref Search ADS PubMed WorldCat 49. Miller LM , et al. Stimulus-based state control in the thalamocortical system . J Neurosci. 2000 ; 20 ( 18 ): 7011 – 7016 . Google Scholar Crossref Search ADS PubMed WorldCat 50. Fingelkurts AA , et al. Short-term EEG spectral pattern as a single event in EEG phenomenology . Open Neuroimag J. 2010 ; 4 : 130 – 156 . Google Scholar PubMed WorldCat 51. Horne RSC , et al. The impact of sleep disordered breathing on cardiovascular health in overweight children . Sleep Med. 2018 ; 41 : 58 – 68 . Google Scholar Crossref Search ADS PubMed WorldCat 52. Chan KC , et al. Neurocognitive dysfunction and grey matter density deficit in children with obstructive sleep apnoea . Sleep Med. 2014 ; 15 ( 9 ): 1055 – 1061 . Google Scholar Crossref Search ADS PubMed WorldCat 53. Kheirandish-Gozal L , et al. Preliminary functional MRI neural correlates of executive functioning and empathy in children with obstructive sleep apnea . Sleep. 2014 ; 37 ( 3 ): 587 – 592 . Google Scholar Crossref Search ADS PubMed WorldCat 54. Philby MF , et al. Reduced regional grey matter volumes in pediatric obstructive sleep apnea . Sci Rep. 2017 ; 7 : 44566 . Google Scholar Crossref Search ADS PubMed WorldCat 55. Lee YJ , et al. Sleep EEG characteristics in young and elderly patients with obstructive sleep apnea syndrome . Psychiatry Investig. 2016 ; 13 ( 2 ): 217 – 221 . Google Scholar Crossref Search ADS PubMed WorldCat 56. Yang JS , et al. EEG spectral analysis of apnoeic events confirms visual scoring in childhood sleep disordered breathing . Sleep Breath. 2012 ; 16 ( 2 ): 491 – 497 . Google Scholar Crossref Search ADS PubMed WorldCat 57. Chervin RD , et al. Method for detection of respiratory cycle-related EEG changes in sleep-disordered breathing . Sleep. 2004 ; 27 ( 1 ): 110 – 115 . Google Scholar Crossref Search ADS PubMed WorldCat © Sleep Research Society 2019. Published by Oxford University Press on behalf of the Sleep Research Society. All rights reserved. For permissions, please e-mail journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Age and autonomic control, but not cerebral oxygenation, are significant determinants of EEG spectral power in children
JF - SLEEP
DO - 10.1093/sleep/zsz118
DA - 2019-09-06
UR - https://www.deepdyve.com/lp/oxford-university-press/age-and-autonomic-control-but-not-cerebral-oxygenation-are-significant-C8W0eDpuY3
VL - 42
IS - 9
DP - DeepDyve
ER -