The association between sleep microarchitecture and cognitive function in middle-aged and older men: a community-based cohort study
2022; American Academy of Sleep Medicine; Volume: 18; Issue: 6 Linguagem: Inglês
10.5664/jcsm.9934
ISSN1550-9397
AutoresJ Parker, Sarah Appleton, Yohannes Adama Melaku, A D’Rozario, Gary Wittert, Sean Martin, Barbara Toson, Peter Catcheside, Bastien Lechat, A Teare, Robert Adams, Andrew Vakulin,
Tópico(s)Obstructive Sleep Apnea Research
ResumoFree AccessScientific InvestigationsThe association between sleep microarchitecture and cognitive function in middle-aged and older men: a community-based cohort study Jesse L. Parker, Hons, Sarah L. Appleton, PhD, Yohannes Adama Melaku, PhD, Angela L. D'Rozario, PhD, Gary A. Wittert, MB Bch, MD, Sean A. Martin, PhD, Barbara Toson, MS, Peter G. Catcheside, PhD, Bastien Lechat, PhD, Alison J. Teare, BA, Robert J. Adams, MBBS, MD, Andrew Vakulin, PhD Jesse L. Parker, Hons Address correspondence to: Jesse Parker, Hons, Flinders Health and Medical Research Institute, Adelaide Institute for Sleep Health, Mark Oliphant Building, Flinders University, 5 Laffer Drive, Bedford Park, SA, 5042, Australia; Tel: 0420431392; Email: E-mail Address: [email protected] Flinders Health and Medical Research Institute, Adelaide Institute for Sleep Health, Flinders University, Adelaide, South Australia, Australia , Sarah L. Appleton, PhD Flinders Health and Medical Research Institute, Adelaide Institute for Sleep Health, Flinders University, Adelaide, South Australia, Australia South Australian Health and Medical Research Institute, Adelaide, South Australia, Australia , Yohannes Adama Melaku, PhD Flinders Health and Medical Research Institute, Adelaide Institute for Sleep Health, Flinders University, Adelaide, South Australia, Australia , Angela L. D'Rozario, PhD CIRUS, Centre for Sleep and Chronobiology, Woolcock Institute of Medical Research, University of Sydney, Sydney, New South Wales, Australia Faculty of Science, School of Psychology, The University of Sydney, Sydney, New South Wales, Australia , Gary A. Wittert, MB Bch, MD South Australian Health and Medical Research Institute, Adelaide, South Australia, Australia Freemasons Centre for Male Health and Wellbeing, Adelaide Medical School, University of Adelaide, Adelaide, South Australia, Australia , Sean A. Martin, PhD South Australian Health and Medical Research Institute, Adelaide, South Australia, Australia Freemasons Centre for Male Health and Wellbeing, Adelaide Medical School, University of Adelaide, Adelaide, South Australia, Australia , Barbara Toson, MS College of Medicine and Public Health, Flinders University, Adelaide, South Australia, Australia , Peter G. Catcheside, PhD Flinders Health and Medical Research Institute, Adelaide Institute for Sleep Health, Flinders University, Adelaide, South Australia, Australia , Bastien Lechat, PhD Flinders Health and Medical Research Institute, Adelaide Institute for Sleep Health, Flinders University, Adelaide, South Australia, Australia , Alison J. Teare, BA Flinders Health and Medical Research Institute, Adelaide Institute for Sleep Health, Flinders University, Adelaide, South Australia, Australia , Robert J. Adams, MBBS, MD Flinders Health and Medical Research Institute, Adelaide Institute for Sleep Health, Flinders University, Adelaide, South Australia, Australia South Australian Health and Medical Research Institute, Adelaide, South Australia, Australia Respiratory and Sleep Services, Southern Adelaide Local Health Network, Bedford Park, Adelaide, South Australia, Australia , Andrew Vakulin, PhD Flinders Health and Medical Research Institute, Adelaide Institute for Sleep Health, Flinders University, Adelaide, South Australia, Australia CIRUS, Centre for Sleep and Chronobiology, Woolcock Institute of Medical Research, University of Sydney, Sydney, New South Wales, Australia Published Online:June 1, 2022https://doi.org/10.5664/jcsm.9934Cited by:1SectionsAbstractEpubPDFSupplemental Material ShareShare onFacebookTwitterLinkedInRedditEmail ToolsAdd to favoritesDownload CitationsTrack Citations AboutABSTRACTStudy Objectives:Sleep microarchitecture parameters determined by quantitative power spectral analysis of electroencephalograms have been proposed as potential brain-specific markers of cognitive dysfunction. However, data from community samples remain limited. This study examined cross-sectional associations between sleep microarchitecture and cognitive dysfunction in community-dwelling men.Methods:Florey Adelaide Male Ageing Study participants (n = 477) underwent home-based polysomnography (2010–2011). All-night electroencephalogram recordings were processed using quantitative power spectral analysis following artifact exclusion. Cognitive testing (2007–2010) included the inspection time task, Trail-Making Tests A and B, and Fuld object memory evaluation. Complete case cognition, polysomnography, and covariate data were available in 366 men. Multivariable linear regression models controlling for demographic, biomedical, and behavioral confounders determined cross-sectional associations between sleep microarchitecture and cognitive dysfunction overall and by age-stratified subgroups.Results:In the overall sample, worse Trail-Making Test A performance was associated with higher rapid eye movement (REM) theta and alpha and non-REM theta but lower delta power (all P < .05). In men ≥ 65 years, worse Trail-Making Test A performance was associated with lower non-REM delta but higher non-REM and REM theta and alpha power (all P < .05). Furthermore, in men ≥ 65 years, worse Trail-Making Test B performance was associated with lower REM delta but higher theta and alpha power (all P < .05).Conclusions:Sleep microarchitecture parameters may represent important brain-specific markers of cognitive dysfunction, particularly in older community-dwelling men. Therefore, this study extends the emerging community-based cohort literature on a potentially important link between sleep microarchitecture and cognitive dysfunction. The utility of sleep microarchitecture for predicting prospective cognitive dysfunction and decline warrants further investigation.Citation:Parker JL, Appleton SL, Melaku YA, et al. The association between sleep microarchitecture and cognitive function in middle-aged and older men: a community-based cohort study. J Clin Sleep Med. 2022;18(6):1593–1608.BRIEF SUMMARYCurrent Knowledge/Study Rationale: Preliminary evidence predominantly derived from small clinical and case-controlled studies suggests sleep microarchitecture determined by quantitative power spectral analysis of electroencephalograms may represent an important brain-specific marker of cognitive dysfunction. However, previous small studies have not controlled for potential confounders, leaving the nature of this association unclear.Study Impact: This cross-sectional study is one of the first to report that in a community sample worse focused attention and processing speed (Trail-Making Test A performance) and executive function (Trail-Making Test B performance) are independently associated with sleep microarchitecture in older community-dwelling men (≥ 65 years). These data extend the emerging community-based cohort literature and provide further evidence suggesting sleep microarchitecture may represent an important brain-specific marker of cognitive dysfunction.INTRODUCTIONCognitive dysfunction affects a considerable proportion of the general population and is particularly prevalent among older adults.1,2 Insufficient sleep and sleep disorders are similarly associated with cognitive dysfunction.3 Emerging evidence, predominantly derived from small samples, suggests sleep microarchitecture may be associated with daytime cognitive dysfunction.4 However, evidence from community samples controlling for potential confounders remains scarce, leaving the nature of this association unclear.Sleep microarchitecture parameters determined by quantitative power spectral analysis of electroencephalograms (EEGs) may represent important brain-specific markers of cognitive dysfunction.4 However, as reviewed by D'Rozario et al,2 the emerging evidence is inconsistent. Three small case-controlled studies previously examined sleep microarchitecture in patients with mild cognitive impairment (MCI) compared to age-matched controls.5–7 Westerberg et al5 identified that patients with amnestic MCI (aMCI) (n = 18) showed lower rapid eye movement (REM) sleep theta activity and low-frequency non-REM (NREM) sleep delta and theta activity compared to controls (n = 10). Brayet et al7 identified that patients with aMCI (n = 22) showed greater REM EEG slowing (ratio of slow to fast EEG frequencies) compared to controls (n = 33). Gorgoni et al6 identified that patients with aMCI showed lower NREM fast spindle density (number/minute of ∼13- to 16-Hz EEG bursts, ≥ 0.5 and ≤ 3 seconds) compared to controls (both n = 15). In a population-based case-controlled study, the Study of Osteoporotic Fractures (n = 85 MCI cases and n = 85 age-matched controls), Djonlagic et al8 reported that community-dwelling women ≥ 65 years who had developed MCI 5 years after a baseline sleep study exhibited higher NREM alpha and theta activity and REM alpha and sigma activity compared to controls. In a similarly sized study, Waser et al9 reported that men with cognitive decline from early to late adulthood showed greater NREM EEG slowing compared to men without cognitive decline. Although previous case-controlled studies have investigated differences in sleep microarchitecture parameters between patients with MCI or cognitive decline and controls, these have not thoroughly examined associations between sleep microarchitecture and cognitive dysfunction.Ageing is associated with an increase in sleep disorders such as obstructive sleep apnea (OSA), which is characterized by repeated complete (apnea) or partial (hypopnea) pharyngeal upper airway collapse.10 These nocturnal events lead to intermittent hypoxemia and hypercapnia, augmented breathing, sleep fragmentation, and blood pressure surges associated with frequent arousals.11 Sleep microarchitecture has also been impaired in patients with OSA relative to matched controls.4,12–14 Reported abnormalities include lower NREM delta activity,14–16 higher fast-frequency beta activity,15,16 decreased spindle frequency and occurrence,17 reduced K-complex density (number/minute of < 1-Hz EEG bursts),18 and greater REM EEG slowing.12,13A community-based cohort study (n = 664) reported that increased intermittent hypoxemia was independently associated with greater REM EEG slowing and higher NREM fast-frequency beta activity.19 Another community-based cohort study (n = 3,819) that recruited late middle-aged and older participants from 2 independent community-based cohorts, the Multi-Ethnic Study of Atherosclerosis and Osteoporotic Fractures in Men Study, found lower NREM delta activity was associated with worse executive function while accounting for OSA and other potential confounders.20 However, no additional community-based cohort studies have investigated potential links between sleep microarchitecture and cognitive dysfunction while accounting for OSA and other potential confounders. Furthermore, no community-based cohort studies have determined whether sleep microarchitecture parameters are differentially associated with cognitive dysfunction among early to middle-aged vs older community-dwelling participants.The primary aim of the present study was to extend the emerging community-based cohort literature by examining cross-sectional associations between sleep microarchitecture and cognitive dysfunction in community-dwelling men. A secondary aim was to examine cross-sectional associations between sleep microarchitecture and cognitive dysfunction in early to middle-aged (< 65 years) and older (≥ 65 years) men to determine whether sleep microarchitecture is differentially associated with cognitive dysfunction among early to middle-aged vs older community-dwelling men. It was hypothesized that lower NREM delta power, higher power in faster-frequency EEG bands during NREM sleep, and greater REM EEG slowing would be independently associated with worse cognitive function.METHODSStudy participantsThe Men Androgen Inflammation Lifestyle Environment and Stress (MAILES) study comprises 2,569 unselected urban community-dwelling men harmonized from 2 independent prospective population-based cohorts; all participants of the Florey Adelaide Male Ageing Study (FAMAS) and male participants of the North West Adelaide Health Study.21 The present study includes FAMAS participants (n = 1,195) aged 35–80 years at baseline (2002) and living in the northern and western regions of Adelaide, South Australia.22,23During a computer-assisted telephone interview follow-up in 2010 (n = 858), FAMAS participants who reported no previous OSA diagnosis (n = 767) were invited to undergo unattended home-based 8-channel polysomnography (PSG) (2010–2011) as part of a substudy of the MAILES study.21,24 Approximately 75% of eligible participants (n = 575) agreed to undergo PSG. However, 98 PSGs were not completed due to time and budget constraints leading to a final PSG sample of 477 (Figure 1). As previously described, there was minor healthy volunteer responder bias with participants who underwent PSG.25 Participants who underwent PSG were on average younger, less obese, and less commonly reported poor general health compared to participants who did not undergo PSG.25 FAMAS was conducted in accordance with the Declaration of Helsinki and approved by the Royal Adelaide Hospital Human Research Ethics Committee (approval number: 020305). All participants provided written informed consent.Figure 1: Clinical and sleep study assessments and cognitive function testing.CATI = computer-assisted telephone interview, EEG = electroencephalography, FAMAS = Florey Adelaide Male Ageing Study, OSA = obstructive sleep apnea, PSG = polysomnography, qEEG = quantitative electroencephalography.Download FigureSleep study assessmentParticipants underwent home-based 8-channel ambulatory PSG (Embletta X100; Embla Systems, Thornton, CO), which recorded electrical brain activity (EEG, F4-M1) and left electrooculography (EOG) with a 12-bit signal resolution, a sampling rate of 200 Hz, and band-pass filters (0.3–35 Hz) along with submental electromyography (EMG), nasal cannula pressure, thoracic and abdominal motion, finger pulse oximetry, and body position. Before PSG set-up, trained staff obtained anthropometric measurements (height, weight, and body mass index [BMI, kg/m2]).A single experienced sleep technician manually scored all PSG measures according to 2007 American Academy of Sleep Medicine (AASM) alternative scoring criteria,26 which was recommended by an expert panel of the Australasian Sleep Association for use in prospective epidemiological studies.27 OSA was identified by an apnea-hypopnea index (AHI) ≥10 events/h and further categorized as mild (10–19 events/h), moderate (20–29 events/h), or severe (AHI ≥ 30 events/h). Ruehland et al have shown that an AHI of 5 events/h used to define sleep-disordered breathing by the AASM 2007 recommended criteria is approximately equivalent to 10 events/h using the alternative criteria and 15 events/h using the older 1999 Chicago criteria.26 Therefore, an AHI cut-off of 10 events/h was chosen to maintain comparability with previous work. Apnea was defined as complete or near-complete airflow cessation (≥ 90%) measured using nasal cannula pressure excursions with breathing lasting ≥ 10 seconds. Hypopnea was defined as a ≥ 50% decrease in nasal cannula pressure excursions along with an associated ≥ 3% oxygen desaturation or EEG arousal.26 Sleep hypoxemia was assessed from the percentage of total sleep time with oxygen saturation < 90%.26 Sleep studies were considered acceptable with ≥ 3.5 hours of sleep and ≥ 5.5 hours of total-recorded study time with technically acceptable respiratory and EEG signal quality for the majority of the recording.EEG data processingA detailed description of quantitative EEG (qEEG) analysis used in this study has been previously described.28,29 Synchronized European Data Format and sleep stage files were generated using Embla REMLogic PSG Software (Natus Medical, Inc., Pleasanton, California). Of the 477 men who underwent sleep studies, PSG data were of adequate quality for qEEG analysis in 415. An algorithm identified artifactual EEG data over consecutive nonoverlapping 5-second epochs based on previously validated artifact detection amplitude threshold parameters.28 Contaminated 5-second epochs, including arousals where EEG traces went outside the amplitude boundaries, were subsequently excluded from qEEG analysis.Manual verification of automated artifact scoring accuracyAutomated artifact scoring accuracy was verified by manual review in 10% of randomly selected PSGs (n = 36). Four agreement measures were calculated, including accuracy, sensitivity, specificity, and Cohen's kappa (k). Consistent with the original artifact detection validation study,28 our algorithm displayed excellent accuracy (mean ± standard deviation) (96.6% ± 4.4%) and specificity (99.9% ± 28.1%) and good to moderate sensitivity (59.1% ± 0.1%) and agreement (k = 0.68 ± 0.26).EEG power spectral analysisAfter rejecting artifactual epochs, power spectra were obtained using a standard fast Fourier transform algorithm with a rectangular weighting window for each nonoverlapping 5-second epoch of EEG. Absolute spectral power (µV2) was calculated in the delta, theta, alpha, sigma, and beta frequency bands and was defined as EEG activity of 0.5–4.5, 4.5–8, 8–12, 12–15, and 15–32 Hz, respectively, during NREM and REM sleep. The EEG power for each sleep-staged 30-second epoch was calculated by averaging data from 6 artifact-free 5-second epochs that made up each 30-second recording segment. The weighted average spectral power or spectral variance over the frequency interval within the defined frequency bands was computed for NREM (N2 and N3) and REM sleep. Weighted average spectral power is a weighted average, based on sleep stage or type, that was calculated by averaging the absolute power of 30-second epochs of the EEG. Relative spectral power for each frequency band during NREM and REM sleep (eg, delta/delta + theta + alpha + sigma + beta) was calculated. A global measure of NREM and REM EEG slowing (ie, a ratio of slow to fast EEG frequencies ([delta + theta]/[alpha + sigma + beta]) was also calculated.Cognitive assessmentsParticipants completed 4 standardized, validated, and well-established cognitive tests outlined below during the 2007–2010 follow-up as previously described in greater detail.30 The average time lag between cognitive and PSG testing was 26 (range, 3–51) months. Health Study participants completed PSG testing, only FAMAS participants completed both the cognitive and PSG assessments, thus comprising the sample included in all analyses.Inspection time taskThis inspection time task measured visual processing speed determined as the average duration in milliseconds that a stimulus was presented to participants before they correctly identified which of 2 vertical lines displayed on a screen was longer on ≥ 75% of trials.31 As a measure of the early stage of information processing, inspection time was associated with cognitive ageing.32 Impairments in inspection time have also been associated with the severity of Alzheimer's disease.33Trail-Making TestThe Trail-Making Test (TMT) assesses visual search, scanning, processing speed, mental flexibility, focused attention, and executive function.34,35 The test consisted of 2 parts requiring participants to map out a sequential path. TMT-A is a focused-attention measure requiring participants to connect encircled numbers (1–25) in sequence.36 TMT-B was an executive function measure requiring participants to connect circles containing numbers with the corresponding letters in the appropriate sequence (1–A, 2–B, 3–C, etc).36 The time, in seconds, needed to complete each path was scored.Fuld object memory evaluationThe Fuld object memory evaluation (FOME) test utilizes multiple sensory pathways (tactile, visual, and verbal) to assess working memory performance.37 This multisensory method evaluates encoding, storage, and retrieval of unrelated objects. The maximum possible score is 10, with higher scores representing intact memory and lower scores impaired memory. The FOME test helps identify memory decline, and low scores may indicate dementia.38Covariate assessmentsSelf-completed questionnaires determined demographic (age, financial stress [spends > earns vs saves a little/lot], highest educational attainment [≥diploma, certificate, trade, bachelor's degree or higher vs ≤ high school], and marital status [married/partner vs other]) and other health-related (current smoking status, alcohol intake, and physical activity [low/moderate/vigorous vs sedentary behavior]) risk factors and quality of life (the 36-item short-form survey instrument [SF-36]). BMI was categorized according to international criteria from the World Health Organization (< 25 [underweight/normal], 25 to < 30 [overweight], and ≥ 30 kg/m2 [obese]).39 Relative social disadvantage, based on participants' residential postcode, was determined with the Australian Bureau of Statistics' Socio-Economic Indexes for Areas (SEIFA) Index of Relative Socio-Economic Disadvantage.40 Clinic assessment (2007–2010) included anthropometry (BMI and waist circumference), seated sphygmomanometer blood pressure, and a fasting blood sample to assess blood glucose.22 Composite cardiovascular disease (self-reported, doctor-diagnosed myocardial infarction, angina, transient ischemic attack, or stroke) and diabetes mellitus (self-reported, doctor-diagnosed, fasting plasma glucose ≥ 7.0 mmol/L [126 mg/dL], hemoglobin A1C [≥ 6.5%], or reported antidiabetic medication use [oral hypoglycemic agents and/or insulin]) were also determined. Men were classified as having insomnia symptoms if they reported difficulty initiating or maintaining sleep occurring at least 3 nights per week (Pittsburgh Sleep Quality Index dimensions) and significant daytime fatigue defined as an SF-36 Vitality Scale score 1 standard deviation below the mean.41 Hypertension was defined as systolic blood pressure ≥ 140 mm Hg and/or diastolic blood pressure ≥ 90 mm Hg or reported antihypertensive medication use.42Statistical analysis methodologyComplete case cognition, PSG, and covariate data were available in 366 men. Data were analyzed using IBM SPSS version 25.0 (IBM Corporation, Armonk, New York). Descriptive statistics for NREM and REM relative spectral powers, cognitive test scores, and continuous covariates are reported as mean standard deviation. NREM and REM EEG slowing ratio is reported as median due to nonnormality. For descriptive analyses, dichotomous and categorical risk factor covariates are reported as percentages (proportion).For analysis of cognitive function in relation to demographic, biomedical, and behavioral risk factors, 1-way analyses of variance and independent samples t tests were performed. Mann-Whitney U tests were used to test for differences in NREM and REM EEG slowing ratio between middle-aged (< 65 years) and older (≥ 65 years) men. Moreover, independent samples t tests were used to test for differences in NREM and REM relative spectral powers between middle-aged and older men. Pearson's chi-squared tests were used to examine differences in demographic, biomedical, and behavioral risk factors between middle-aged and older men.Univariable and multivariable linear regression models determined cross-sectional associations of cognitive dysfunction with NREM and REM relative spectral powers and logarithmically (10-base) transformed EEG slowing ratio. Unstandardized beta (B) coefficients (95% confidence interval [CI]) and adjusted R2 values are reported. For each sleep microarchitecture metric, 3 covariate-adjusted regression models were constructed. Model 1 was adjusted for age and OSA; model 2 was additionally adjusted for demographic (financial stress, education, socioeconomic disadvantage, and marital status) risk factors; and model 3 was additionally adjusted for total sleep time and biomedical and behavioral (BMI, alcohol risk, current smoking status, cardiovascular disease, diabetes mellitus, blood glucose, insomnia, and hypertension) risk factors. Age, BMI, blood glucose, and total sleep time were treated continuously, with all other covariates treated dichotomously or categorically.Moderator analysis was performed using age × qEEG as an interaction term to determine if age significantly moderated observed fully adjusted associations between sleep microarchitecture and cognitive dysfunction. After identifying significant moderation, age-stratified (< 65 vs ≥ 65 years) linear regression analyses were performed to determine if sleep microarchitecture parameters were differentially associated with cognitive dysfunction among early to middle-aged vs older community-dwelling men.For age-stratified multivariable linear regression analyses, the purposeful-selection-of-covariates procedure proposed by Hosmer and Lemeshow43 was applied to construct a robust multivariable model. Accordingly, unadjusted analyses were first performed to examine crude associations between covariates and cognitive outcomes, with covariates returning P-values < 0.25 selected as potential candidates for adjustment. Nonsignificant covariates were gradually removed until only significant covariates remained. Covariates not initially selected as potential candidates for adjustment were then gradually re-added with the significant covariates retained earlier to identify which of these covariates were important in the presence of initially selected covariates. This purposeful covariate selection procedure reduced the final multivariable model to 8 covariates, including age, AHI, financial stress, socioeconomic disadvantage, marital status, education, total sleep time, and cardio-metabolic conditions (including 1 or more of diabetes, hypertension, or cardiovascular disease).Assumptions of linear regression modeling, including linearity, independence, homoscedasticity, and normality, were met. Furthermore, all variance inflation factor values were near 1, indicating absence of multicollinearity. For all analyses, a 2-sided P < .05 was considered statistically significant. No multiple comparison adjustments were performed.44,45RESULTSOf the 477 men who underwent PSG, 397 had adequate quality sleep microarchitecture and cognition data available for analysis. In total, 366 men were included in the analysis after excluding 31 (7.8%) who reported regularly using psychoactive medication(s), including opiates, antipsychotics, antiepileptics, antidepressants, or benzodiazepines, which may potentially disrupt sleep microarchitecture.Participant characteristicsParticipant characteristics, overall and stratified by age, are reported in Table 1. Of the men included in the analysis, 52.5% had at least mild OSA (AHI ≥ 10 events/h), and 12.9% had severe OSA (AHI ≥ 30 events/h). Approximately one-third (31.7%) were obese (BMI ≥ 30 kg/m2). Higher age, lower education, diabetes, and presence of 1 or more cardio-metabolic conditions were associated with worse cognitive function (Table S1 in the supplemental material).Table 1 Participant characteristics (relative spectral powers, EEG slowing ratio, OSA parameters, and demographic and other risk factors).Participant CharacteristicsOverall Sample (n = 366)< 65 Years (n = 257)≥ 65 Years (n = 109)Demographic risk factors Age (years), mean (SD)59.0 (10.4)53.5 (6.4)71.9 (5.3)* Financial stress, % (n) Spends > earns15.3 (56)14.4 (37)17.4 (19) Saves a little/lot84.7 (310)85.6 (220)82.6 (90) Highest educational attainment, % (n) Diploma, certificate, trade, bachelor's degree or higher71.3 (261)73.2 (188)67.0 (73) SEIFA IRSD, % (n) Quintile 1 (highest disadvantage)21.6 (79)24.9 (64)13.8 (15)† Quintile 210.7 (39)11.7 (30)8.3 (9) Quintile 329.0 (106)32.3 (83)21.1 (23)† Quintile 426.0 (95)22.6 (58)33.9 (37)† Quintile 5 (lowest disadvantage)12.8 (47)8.6 (22)22.9 (25)† Married/partner, % (n)85.0 (311)85.6 (220)83.5 (91)Relative spectral powers, mean (SD) NREM Delta power (0.5–4.5 Hz), %81.3 (6.9)81.5 (6.9)80.8 (6.6) Theta power (4–8 Hz), %8.1 (2.7)7.9 (2.7)8.5 (2.7)* Alpha power (8–12 Hz), %5.3 (2.4)5.3 (2.5)5.3 (2.4) Sigma power (12–15 Hz), %2.0 (1.0)2.1 (1.0)1.9 (1.0) Beta power (15–32 Hz), %3.3 (2.0)3.3 (2.1)3.5 (1.9) REM Delta power (0.5–4.5 Hz), %70.1 (9.6)70 (9.5)70.3 (10) Theta power (4.5–8 Hz), %11.8 (4.1)11.9 (4.0)11.6 (4.4) Alpha power (8–12 Hz), %6.8 (2.5)6.8 (2.5)6.9 (2.7) Sigma power (12–15 Hz), %2.9 (1.1)2.9 (1.1)2.9 (1.1) Beta power (15–32 Hz), %8.4 (3.9)8.4 (3.9)8.3 (4.0)EEG slowing ratios, median (IQR) NREM slowing ratio9.0 (6.8, 12.4)9.0 (6.7, 12.5)8.8 (6.8, 12.2) REM slowing ratio4.5 (3.4, 6.4)4.5 (3.4, 6.2)4.5 (3.2, 6.6) OSA severity categories (AHI), % (n) < 10 events/h47.5 (162)50.6 (130)43.1 (47)† 10–19 events/h26.7 (91)28 (72)22.0 (24)† 20–29 events/h12.9 (44)11.3 (29)18.3 (20)† ≥ 30 events/h12.9 (44)10.1 (26)16.5 (18)†Other risk factors Medium–very high alcohol risk, % (n)5.2 (19)6.6 (17)1.8 (2)† Low/moderate/vigorous physical activity, % (n)77.3 (283)75.1 (193)82.6 (90)† BMI (kg/m2), % (n) < 25 (underweight/normal)19.9 (73)20.6 (53)18.3 (20) 25 to < 30 (overweight)48.4 (177)48.2 (124)48.6 (53) ≥ 30 (obese)31.7 (116)31.1 (80)33.0 (36) Current smokers, % (n)16.9 (62)20.2 (52)9.2 (10)† Cardiovascular disease, % (n)6.8 (25)3.1 (8)15.6 (17)† Insomnia, % (n)12.8 (47)13.2 (34)11.9 (13) Diabetes mellitus, % (n)17.2 (63)13.6 (35)25.7 (28)† Hypertension, % (n)47.0 (172)40.5 (104)62.4 (68)† Cardiometabolic conditions, % (n)56.6 (204)48.2 (124)73.4 (80)† Total sleep time < 360 minutes, % (n)36.9 (135)34.6 (89)42.2 (46)†All PSG measures were scored according to AASM 2007 alternative scoring criteria in which an AHI of 10 events/h is approximately equivalent to an AHI of 5 events/h used to define sleep-disordered breathing by the AASM 2007 recommended scoring criteria. SEIFA IRSD: Socio-Economic Indexes for Areas Index of Relative Socio-Ec
Referência(s)