Polysomnographic Data Research Articles

Community prevalence of positional obstructive sleep apnea.

The prevalence of positional obstructive sleep apnea (POSA) in community populations warrants further investigation. Further, more research is needed into the clinical characteristics of its subtypes such as supine predominant OSA (spOSA) and supine isolated OSA (siOSA). A cross-sectional analysis was performed on 1870 Sleep Heart Health Study participants. OSA was defined by an apnea-hypopnea index (AHI) of five events or more per hour of sleep. Participants with OSA were classified as POSA if the supine AHI was ≥2 times the non-supine AHI. Participants with OSA who did not meet this threshold were classified as non-positional OSA participants (non-POSA). Demographics, polysomnographic data, comorbidities, and medications were all considered. The POSA subtypes spOSA and siOSA were also investigated. POSA participants were slightly older, less obese, and had higher systolic blood pressure than non-POSA subjects, in addition to being more prevalent (62% and 38%, respectively). POSA exhibited higher supine and total AHI. The prevalence of comorbidities or prescription drugs did not differ. In the POSA cohort, spOSA was more prevalent than siOSA (56% vs. 44%). When compared to siOSA, spOSA was associated with more fragmented sleep and higher AHI. Furthermore, when compared to the siOSA group, the spOSA group had a higher prevalence of hypertension and diabetes, as well as more frequently prescribed medications for these comorbidities. In the SHHS population the prevalence of POSA is greater than non-POSA. The spOSA subtype is more prevalent and appears to have worse health consequences than siOSA.

The interaction between comorbidities and sleep stages influences oxygen re-saturation characteristics.

Sleep stages exhibit differing patterns of cardiac, respiratory and nervous system activation. Rapid eye movement (REM) sleep is associated with deeper oxygen desaturation events in obstructive sleep apnea. However, no studies have looked to the effect of sleep staging on re-saturation characteristics. Polysomnographic data from the Sleep Heart Health Study were utilised to derive oxygen saturation parameters from events having both desaturation and re-saturation parts. Sleep stages were described as non-REM (NREM) 1 or 2 (combined), 3 and REM. Sleep stage effects on desaturation and re-saturation characteristics were investigated in a healthy subgroup (n = 759) and participants with hypertension (n = 2534), lung disease (n = 715), heart failure (n = 199) and myocardial infarction (n = 713). A total of 3793 participants (48.3% female) were included for analysis (171,976 saturation events; median 20 per participant). Events during REM had the longest duration, deepest nadirs and greatest area for desaturations and re-saturations. Sleep stage effected re-saturation parameters more than desaturation parameters, with a relative difference from NREM 1/2 to REM in duration of 16.7% and 29.8%, in rate of 0.2% and 4.5% and in area of 36.1% and 48.0% for desaturation and re-saturation, respectively, among healthy participants. Similarly, the effect of comorbidities was greater upon re-saturations than desaturations, as participants with heart failure recorded a relative difference to healthy participants of 10.3% and 24.4% for desaturation and re-saturation duration, respectively, 12.8% and 15.0% for rate and 6.4% and 16.4% for area. Sleep stages and comorbidities have significant effects upon nocturnal oxygen re-saturation parameters. Cardiorespiratory comorbidities elicit greater degradations in oxygen re-saturations than they do desaturations.

Open-Source Algorithm for Automated Vigilance State Classification Using Single-Channel Electroencephalogram in Rodents

Accurate identification of sleep stages is essential for understanding sleep physiology and its role in neurological and behavioral research. Manual scoring of polysomnographic data, while reliable, is time-intensive and prone to variability. This study presents a novel Python-based algorithm for automated vigilance state scoring using single-channel electroencephalogram (EEG) recordings from rats and mice. The algorithm employs artifact processing, multi-band frequency analysis, and Gaussian mixture model (GMM)-based clustering to classify wakefulness, non-rapid, and rapid eye movement sleep (NREM and REM sleep, respectively). Combining narrow and broad frequency bands across the delta, theta, and sigma ranges, it uses a majority voting system to enhance accuracy, with tailored preprocessing and voting criteria improving REM detection. Validation on datasets from 10 rats and 10 mice under standard conditions showed sleep–wake state detection accuracies of 92% and 93%, respectively, closely matching manual scoring and comparable to existing methods. REM sleep detection accuracies of 89% (mice) and 91% (rats) align with previously reported (85–90%). Processing a full day of EEG data within several minutes, the algorithm is advantageous for large-scale and longitudinal studies. Its open-source design, flexibility, and scalability make it a robust, efficient tool for automated rodent sleep scoring, advancing research in standard experimental conditions, including aging and sleep deprivation.

Maxillomandibular Advancement for Obstructive Sleep Apnea in Patients With Obesity: A Meta-Analysis.

To systematically review polysomnographic and cephalometric data in obstructive sleep apnea (OSA) patients with obesity (body mass index [BMI] ≥30 kg/m2) treated with maxillomandibular advancement (MMA). Scopus, PubMed, CINAHL, and The COCHRANE Library. A search was performed from inception until April 3, 2024, in each database. A total of 14 studies (143 subjects) were included. The mean age was 44.3 years (range: 17-69), 80.2% males (95% CI: 72.5-86.5), mean BMI of 35.3 (95% CI: 33.1-37.5), and mean duration to follow-up post-MMA was 13.7 months (95% CI: 10.1-17.3). All objective outcomes improved significantly; overall, apnea-hypopnea index (AHI) decreased by -57.3 ([95% CI: -71.5 to -43.2], p < 0.0001) lowest oxygen saturation (LSAT) increased by 14.1% ([95% CI: 9.9 to 18.3], p < 0.0001), and Epworth Sleepiness Scale (ESS) decreased by -9.4 ([95% CI: -13.5 to -5.2], p < 0.0001). Surgical cure was 39.2% (95% CI: 20.3-60.0), and surgical success was 85.6% (95% CI: 77.8-91.5). Comparing percent reduction in class 3 obesity (-92.9%) as compared to class 1 (-85.5%) and class 2 (-83.6%) exhibited a significant difference (1 vs 3 p = 0.0012, 2 vs 3 p = 0.015). Our findings suggest that MMA significantly improves subjective and objective outcomes in OSA amongst patients with obesity with results comparable to the overall population. Success rates remained above 80% in studies with the highest mean BMI. In addition, patients with class 3 obesity yielded a significantly increased benefit based on percent reduction in AHI compared with class 1 and 2. 1 Laryngoscope, 135:507-516, 2025.

Obstructive sleep apnea subtyping based on apnea and hypopnea specific hypoxic burden is associated with brain aging and cardiometabolic syndrome.

Conventional metrics such as the apnea-hypopnea index (AHI) may not fully capture the diverse clinical manifestations of obstructive sleep apnea (OSA). This study aims to establish a novel OSA subtype classification based on the patterns of apneic and hypopneic hypoxic burden (HB), a potential biomarker that more accurately reflects the severity and duration of respiratory events. We further examined the associations of these HB-based subtypes with cardiometabolic risk and brain health outcomes. We retrospectively analyzed polysomnography data from 1000 participants including normal, mild, moderate, and severe OSA patients. We performed hierarchical clustering based on apneic and hypopneic HB to identify OSA subtypes. We then compared the prevalence of cardiometabolic syndrome (CMS) and brain health outcomes using the brain age index (BAI) among these subtypes. Five distinct subtypes were identified: 'good sleepers' (subtype 1), 'light hypopneic HB' (subtype 2), 'mild HB' (subtype 3), 'moderate HB' (subtype 4), and 'severe HB with marked apneic HB' (subtype 5). The prevalence of CMS (particularly hypertension) was significantly higher in subtypes 2-5 (p<0.001) compared to subtype 1. BAI was higher in subtypes 4 (3.2 years, p<0.0001) and 5 (11.1 years, p<0001) compared to subtype 1. Greater daytime sleepiness was observed in HB-based subtypes 2 and 5 compared to subtype 1 (p<0.001), whereas no significant differences were found among groups classified by OSA severity using AHI. Our study demonstrates that the HB-based subtypes of OSA are significantly associated with various clinical features and outcomes. These insights could be utilized to improve risk stratification and guide the design of future OSA studies.

Drug Induced Sleep Endoscopy-Directed Tongue Surgery to Treat Persistent Pediatric Obstructive Sleep Apnea: A Systematic Review and Meta-Analysis.

To systematically review the literature for articles evaluating outcomes of drug-induced sleep endoscopy (DISE) directed tongue surgery in children with prior adenotonsillectomy and persistent or recurrent obstructive sleep apnea (OSA), and to perform a meta-analysis on the polysomnographic (PSG) data. Systematic review and metanalysis in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) statement guidelines. Primary, post-operative apnea-hypopnea index (AHI) and lowest oxygen saturation (LSAT); Secondary, surgical response rate. Seven studies (283 patients) met criteria and reported PSG outcome data for the systematic review. The mean ± standard deviation surgical response rate was 70.0% ± 17.0% [95% CI 67.6, 71.6]. Six of the seven studies (270 patients) reported specific pre- and post-operative PSG data included in our meta-analysis. The pre- and post-operative AHI decreased from 9.5 ± 12.1 to 4.2 ± 6.9 events/h (p < 0.04) with a mean difference (MD) of -5.13 [95% CI -7.13, -3.13], Z-score 5.02 (p < 0.00001). LSAT improved from 87.8 ± 5.7 to 90.1% ± 5.1% (p < 0.02) with a MD of 2.71 [95% CI 1.53, 3.89], Z-score 4.51 (p < 0.0001). Existing literature demonstrates DISE-directed tongue surgery in children with persistent OSA can reduce AHI by approximately 50%, improve LSAT by nearly 3%, and have an overall positive response to surgery rate of 70%. There is collective evidence that DISE-directed tongue surgery is effective; limitations include heterogeneity in reported outcomes influenced by confounding factors.

Circadian Misalignment Impacts Cardiac Autonomic Modulation in Adolescence.

Although heart rate variability (HRV), a marker of cardiac autonomic modulation (CAM), is known to predict cardiovascular morbidity, the circadian timing of sleep (CTS) is also involved in autonomic modulation. We examined whether circadian misalignment is associated with blunted HRV in adolescents as a function of entrainment to school or on-breaks. We evaluated 360 subjects from the Penn State Child Cohort (median 16y) who had at least 3-night at-home actigraphy (ACT), in-lab 9-h polysomnography (PSG) and 24-h Holter-monitoring heart rate variability (HRV) data. ACT-measured metrics of circadian misalignment included sleep midpoint (SM), sleep irregularity (SI), and social jetlag (SJL). Five 24-h, daytime and nighttime frequency- and time-domain HRV indices were the primary outcomes. Linear regression models adjusted for sex, race/ethnicity, age, body mass index, apnea/hypopnea index, sleep duration and its variability. These associations were also examined as a function of being in-school or on-break. While on-break, a later SM on weekends was significantly associated with all five nighttime HRV indices. While in-school, greater SI on weekdays was significantly associated with three daytime and three nighttime HRV indices. Greater SJL was not associated with any HRV index. Longitudinal analyses confirmed the association of adolescent SM, SI and SJL with change in nighttime HRV since childhood. An irregular sleep phase during days of entrainment to social demands and a delayed sleep phase during ad-libitum days are associated with blunted HRV in adolescents. Circadian misalignment contributes to increased cardiovascular risk via an altered CAM in youth.

Polysomnographic endotypes of successful multilevel upper airway surgery for obstructive sleep apnea.

Multilevel upper airway surgery is effective for some patients with obstructive sleep apnea (OSA), but prediction the response to surgery remains a challenge. The underlying endotypes of OSA include upper airway collapsibility, muscle compensation, loop gain, and the arousal threshold. This study aimed to explore the effect of surgery on polysomnography (PSG)-derived OSA endotypes and establish a surgical response prediction model. Our study included 54 Chinese patients with OSA who underwent multilevel upper airway surgery. Participants underwent PSG before and after surgery with a median follow-up time of 6.5 months. Using AHIBaseline/AHIpost-surgery ≥ 2 and AHIpost-surgery < 10 events/h as criteria, participants were classified as surgery responders and non-responders. The surgical success rate was 26%. These endotypic traits were derived from a standard PSG data by validated methods. The surgery altered both anatomical and non-anatomical endotypic traits, including increased Vpassive (baseline VS post-surgery: 51.5[18.7-84.2] VS 86.8 [67.4-93.7] %Veupnea, p<0.001), decreased loop gain (baseline VS post-surgery: 0.7 [0.7-0.8] VS 0.6[0.5-0.6]; p<0.001), and a higher arousal threshold (baseline VS post-surgery: 202.9[183.7-222.0] VS 160.7[143.9-177.4] %Veupnea; p<0.001). However, it did not significantly affect muscle compensation. Fully adjusted logistic regression analyses indicated that a favorable response to surgery was independently associated with a lower LG (OR [CI 95%], 0.1[0.0-0.5], p= 0.032). In patients with improved muscle compensation or a more collapsible airway (lower Vpassive), a lower loop gain was more strongly indicative of success. However, when muscle compensation was lower or collapsibility was less severe (higher Vpassive), a lower loop gain was less predictive of success. This study demonstrated that multilevel upper airway surgery altered both anatomical and non-anatomical endotypes in Chinese patients with OSA. An endotype based regression model may meaningfully predict surgical success.

Advanced Data Framework for Sleep Medicine Applications: Machine Learning-Based Detection of Sleep Apnea Events

Obstructive Sleep Apnea (OSA) is a prevalent condition that disrupts sleep quality and contributes to significant health risks, necessitating accurate and efficient diagnostic methods. This study introduces a machine learning-based framework aimed at detecting apnea events through analysis of polysomnographic (PSG) and oximetry data. The core component is a Long Short-Term Memory (LSTM) network, which is particularly suited to processing sequential time-series data, capturing complex temporal relationships within physiological signals such as oxygen saturation, heart rate, and airflow. Through extensive feature engineering and preprocessing, the framework optimizes data representation by normalizing, scaling, and encoding input features to enhance computational efficiency and model performance. Key results demonstrate the model’s effectiveness, achieving a balanced accuracy of 79%, precision of 68%, and recall of 76% on the test dataset, with validation set metrics similarly high, supporting the model’s ability to generalize effectively. Comprehensive hyperparameter tuning further contributed to a stable, robust architecture capable of accurately identifying apnea events, providing clinicians with a valuable tool for the early detection and tailored management of OSA. This data-driven framework offers an efficient, reliable solution for OSA diagnostics with the potential to improve clinical decision making and patient outcomes.

Sleep and breathing in children with Joubert syndrome and a review of other rare congenital hindbrain malformations.

Joubert syndrome (JS) is an autosomal recessive disorder with a distinctive mid-hindbrain malformation known as the "molar tooth sign" which involves the breathing control center and its connections with other structures. Literature has reported significant respiratory abnormalities which included hyperpnea interspersed with apneic episodes during wakefulness. Larger-scale studies looking at polysomnographic findings or subjective reports of sleep problems in this population have not yet been published. The primary objectives were (1) compare a large group of children with JS and their unaffected siblings for caregiver-reported sleep difficulties. Secondary objectives were (1) present new polysomnography (PSG) data on our JS cohort; (2) review sleep disordered breathing (SDB) in other rare congenital hindbrain anatomic abnormalities. We conducted a cross-sectional study on a cohort of 109 families affected by JS. Pediatric Sleep Questionnaire (PSQ) and the Children's Sleep Habits Questionnaire (CSHQ) along with general medical health information focused on respiratory and sleep problems were mailed to all patients and families. Caregivers were asked to complete the survey for both children with JS and unaffected siblings, if any. Baseline diagnostic PSG was retrospectively reviewed for those with available studies, and the sleep parameters were compared to a referent cohort. Study participants with JS were older than their unaffected siblings (p = 0.02). Genetic mutations were available for 41 out of 118 individuals, with the most common mutation being MKS3 (31.4%). Patients with JS had higher scores in the PSQ compared to their unaffected siblings (p < 0.001). PSG data showed severe SDB with apnea-hypopnea index (AHI) of 23 ± 15 events/h in patients with JS. Events were primarily obstructive (obstructive AHI 18 ± 15 events/h vs central AHI 4 ± 4 events/h). Abnormal sleep architecture with increased arousal indices, decreased efficiency, and more time awake and in light sleep or wakefulness when compared to the referent data. SDB is common and severe in patients with JS, and the significantly greater obstructive component reported in this cohort makes it necessary to perform complete PSG studies to address or prevent clinical manifestations in this at-risk population. PSQ could represent a viable method to screen for SDB in JS.

Comparative effectiveness of therapies for sleep-disordered breathing in heart failure patients: A comprehensive systematic review and network meta-analysis.

The prevalence of sleep-disordered breathing (SDB) in patients with heart failure (HF) is a significant concern, leading to adverse outcomes. This network meta-analysis (NMA) is dedicated to evaluate the relative effectiveness of diverse therapeutic approaches for SDB treatments in the context of HF. An extensive search up to May 19, 2023, was implemented in PubMed, Cochrane, Embase, and Web of Science to identify randomized controlled trials (RCTs). These trials compared adaptive servo ventilation (ASV), automatic positive airway pressure (APAP), Bi-level positive pressure ventilation (Bi-level PPV), continuous positive airway pressure (CPAP), and oxygen therapy against placebo or standard treatment. Key outcomes included cardiac function indicators (left ventricular ejection fraction [LVEF], plasma B-type natriuretic peptide [BNP] levels), polysomnography data (apnea-hypopnea index [AHI], Mean and Minimum Oxygen Saturation), and patient-reported quality of life measures (Epworth sleepiness scale [ESS]). A random-effects Bayesian NMA was utilized to evaluate and compare the efficacy of these interventions. The study revealed significant therapeutic effects of different interventions in sleep apnea patients. For patients with obstructive sleep apnea (OSA), CPAP demonstrated superior efficacy in enhancing average oxygen saturation (mean difference (MD)=0.98; 95% credible interval (CI) [0.07,2.09]), improving LVEF (MD=5.66; 95% CI [0.71, 10.08]), and reducing AHI (MD=-20.61; 95% CI [-33.84, -9.9]). In the case of central sleep apnea (CSA), the therapeutic approaches showed varying degrees of effectiveness. CPAP proved most effective for enhancing average oxygen saturation(MD=1.1; 95% CI [0.26,1.98]). ASV was the most effective in lowering AHI (MD=-32.11; 95% CI [-47.5, -17.16]), increasing lowest oxygen saturation (MD=8.14; 95% CI [1.72,14.92]), and enhancing LVEF (MD=6.58; 95% CI [0.06,12.49]). No single SDB treatment uniformly benefits all clinical indicators in SDB and HF patients.

Effects of Metabolic Factors on Left Ventricular Diastolic Function in Patients with Obstructive Sleep Apnea.

The effect of metabolic factors on cardiovascular risk in obstructive sleep apnea (OSA) is unclear. This study aimed to investigate the effect of metabolic factors on the left ventricular diastolic function in patients with OSA. This cross-sectional study included a total of 478 patients with OSA from September 2018 to September 2023. After propensity score matching, wherein 193 patients with OSA with metabolic syndrome (MS) were 1:1 matched to patients with OSA without MS by sex and age, data from 386 patients were ultimately analyzed. Furthermore, all patients were divided into mild, moderate, and severe OSA groups according to their sleep apnea-hypopnea index (AHI). Measurements included nocturnal polysomnography, biochemical testing, and transthoracic echocardiography data. The AHI in the MS group was higher (30.24±21.69 vs 23.19±17.65, p<0.001) and the lowest oxygen saturation at night was lower (77.67±9.23 vs 80.59±9.26, p<0.001) than those in the non-MS group. Additionally, the left ventricular end-diastolic diameter (LVEDD), left ventricular end-systolic diameter (LVESD), end-diastolic ventricular septal thickness (IVST), left ventricular end-diastolic posterior wall thickness (LVPWT), left atrial internal diameter (LAD), and E peak to A peak velocity ratio (E/A) in the MS group were higher than those in the non-MS group (P<0.05). The E peak to e' peak velocity ratio (E/e') in the MS group was higher than that in the non-MS group (12.02±3.68 vs 11.13±3.12, P=0.011) and was positively correlated with the diagnosis of MS and metabolic factors (r=0.115, p=0.024; r=0.131, p=0.010, respectively). Patients with five metabolic factors had a significantly higher risk of E/e' elevation than patients in the non-MS group (odds ratio=4.238, p=0.007). MS may be related to OSA severity and left ventricular diastolic dysfunction. An increase in metabolic factors may increase the risk of diastolic dysfunction. Among metabolic factors, blood pressure may be the most important.

Sleep-disordered breathing increases mortality in patients with diabetes.

Sleep-disordered breathing and diabetes mellitus (DM) are often concomitant; however, data on the impact of sleep-disordered breathing on mortality in the population with diabetes remain scarce. The population from the Sleep Heart Health Study, a multicenter prospective observational study representing 5,780 patients with polysomnography and mortality data, including 453 patients with DM, was analyzed to assess the impact of sleep-disordered breathing variables and the presence of DM on all-cause, cardiovascular disease, and noncardiovascular disease associated mortality. Survival analysis and proportional hazard regression models were used to calculate the adjusted hazard ratios (aHRs) for mortality. Patients with DM and the average oxygen saturation > 91.4% had significantly lower all-cause (aHR 0.52, confidence interval [CI] 0.34-0.80) and cardiovascular disease mortality risk (aHR 0.44, CI 0.22-0.87) as compared with patients with oxygen saturation below this value. Apnea-hypopnea index > 31 (aHR 1.58, CI 1.10-2.28) and oxygen desaturation index > 13.3 (aHR 1.58, CI 1.10-2.25) were associated with increased all-cause mortality in participants with DM on treatment. Sleep efficiency and proportion of rapid eye movement sleep did not have any impact on mortality in patients with DM and thus differed significantly from individuals without DM, where increased all-cause mortality was observed in those with sleep efficiency < 81.4% (aHR 0.77, CI 0.68-0.87) or rapid eye movement sleep < 14.9% (aHR 0.78, CI 0.68-0.89). Patients with diabetes on treatment and moderate to severe sleep-disordered breathing experience increased all-cause mortality. Reduced average oxygen saturation predicted both all-cause and cardiovascular death in the population with diabetes. Vichova T, Petras M, Waldauf P, Westlake K, Vimmerova-Lattova Z, Polak J. Sleep-disordered breathing increases mortality in patients with diabetes. J Clin Sleep Med. 2025;21(1):89-99.

The brain, rapid eye movement sleep, and major depressive disorder: A multimodal neuroimaging study

Evidence has established the prominent involvement of rapid eye movement (REM) sleep disturbance in major depressive disorder (MDD). However, the neural correlates of REM sleep in MDD and their clinical significance are less clear. Cross-sectional and longitudinal polysomnography and resting-state functional MRI data were collected from 131 MDD patients and 71 healthy controls to measure REM sleep and voxel-mirrored homotopic connectivity (VMHC). Correlation and mediation analyses were performed to examine the associations between REM sleep, VMHC, and clinical variables. Moreover, we conducted spatial correlations between the neural correlates of REM sleep and a multimodal collection of reference brain maps to facilitate genetic, structural and functional annotations. MDD patients exhibited REM sleep abnormalities manifesting as higher REM sleep latency and lower REM sleep duration, which were correlated with decreased VMHC of the precentral gyrus and inferior parietal lobe and mediated their associations with more severe anxiety symptoms. Longitudinal data showed that VMHC increase of the inferior parietal lobe was related to improvement of depression symptoms in MDD patients. Spatial correlation analyses revealed that the neural correlates of REM sleep in MDD were linked to gene categories primarily involving cellular metabolic process, signal pathway, and ion channel activity as well as linked to cortical microstructure, metabolism, electrophysiology, and cannabinoid receptor. These findings may add important context to the growing literature on the complex interplay between sleep and MDD, and more broadly may inform future treatment for depression via regulating sleep.

Developing a Sleep Algxorithm to Support a Digital Medicine System: Noninterventional, Observational Sleep Study.

Sleep-wake patterns are important behavioral biomarkers for patients with serious mental illness (SMI), providing insight into their well-being. The gold standard for monitoring sleep is polysomnography (PSG), which requires a sleep lab facility; however, advances in wearable sensor technology allow for real-world sleep-wake monitoring. The goal of this study was to develop a PSG-validated sleep algorithm using accelerometer (ACC) and electrocardiogram (ECG) data from a wearable patch to accurately quantify sleep in a real-world setting. In this noninterventional, nonsignificant-risk, abbreviated investigational device exemption, single-site study, participants wore the reusable wearable sensor version 2 (RW2) patch. The RW2 patch is part of a digital medicine system (aripiprazole with sensor) designed to provide objective records of medication ingestion for patients with schizophrenia, bipolar I disorder, and major depressive disorder. This study developed a sleep algorithm from patch data and did not contain any study-related or digitized medication. Patch-acquired ACC and ECG data were compared against PSG data to build machine learning classification models to distinguish periods of wake from sleep. The PSG data provided sleep stage classifications at 30-second intervals, which were combined into 5-minute windows and labeled as sleep or wake based on the majority of sleep stages within the window. ACC and ECG features were derived for each 5-minute window. The algorithm that most accurately predicted sleep parameters against PSG data was compared to commercially available wearable devices to further benchmark model performance. Of 80 participants enrolled, 60 had at least 1 night of analyzable ACC and ECG data (25 healthy volunteers and 35 participants with diagnosed SMI). Overall, 10,574 valid 5-minute windows were identified (5854 from participants with SMI), and 84% (n=8830) were classified as greater than half sleep. Of the 3 models tested, the conditional random field algorithm provided the most robust sleep-wake classification. Performance was comparable to the middle 50% of commercial devices evaluated in a recent publication, providing a sleep detection performance of 0.93 (sensitivity) and wake detection performance of 0.60 (specificity) at a prediction probability threshold of 0.75. The conditional random field algorithm retained this performance for individual sleep parameters, including total sleep time, sleep efficiency, and wake after sleep onset (within the middle 50% to top 25% of the assessed devices). The only parameter where the model performance was lower was sleep onset latency (within the bottom 25% of all comparator devices). Using industry-best practices, we developed a sleep algorithm for use with the RW2 patch that can accurately detect sleep and wake windows compared to PSG-labeled sleep data. This algorithm may be used for a more complete understanding of well-being for patients with SMI in a real-world setting, without the need for PSG and a sleep lab.

Sex Differences in Sleep Profiles and the Effect of Elexacaftor/Tezacaftor/Ivacaftor on Sleep Quality in Adult People with Cystic Fibrosis: A Prospective Observational Study.

Recent studies indicate that sleep and sleep disorders differ between men and women, but corresponding data in people with chronic lung diseases are lacking. This study aims to answer the question of what the sex-specific differences in sleep profiles and responses to elexacaftor/tezacaftor/ivacaftor (ETI) therapy in people with cystic fibrosis (pwCF) are. Adult pwCF and a matched control group (adults with suspected sleep-disordered breathing undergoing in-laboratory polysomnography (PSG)) were included. PSG data at baseline and after 6 months' ETI therapy were compared between men (mwCF) and women (wwCF) with cystic fibrosis. PSG data at baseline and 6-month follow-up for mwCF/wwCF were compared with baseline PSG data for men/women in the control group. Daytime sleepiness was evaluated using the Epworth Sleepiness Scale (ESS). Correlations between change in percentage predicted forced expiratory volume in 1 s from baseline to 6 months were correlated with corresponding changes in key sleep parameters. Changes in transferrin during ETI therapy were also documented. Twenty-eight pwCF (12 wwCF, 16 mwCF) and 28 matched controls were included. Both mwCF (4 ± 5 vs. 9 ± 20 events/h, p = 0.028) and wwCF (3 ± 3 vs. 8 ± 9 events/h, p = 0.004) had fewer respiratory events during sleep versus male and female controls, but worse sleep efficiency (75 ± 11% vs. 84 ± 11%; p = 0.004 and 76 ± 10% vs. 83 ± 11%; p = 0.011. The baseline ESS score was significantly higher in wwCF versus female controls (8 ± 4 vs. 14 ± 8; p = 0.040). Although some sleep parameters normalized during ETI therapy in pwCF, sleep quality remained poor. The transferrin levels at baseline (2.7 ± 0.4 vs. 2.2 ± 0.5; p = 0.049) and 6 months (3.8 ± 0.4 vs. 2.6 ± 0.5; p < 0.001) were significantly higher in the wwCF versus the mwCF, and the change from baseline during ETI therapy was significantly greater in women versus men (1.1 ± 0.6 vs. 0.4 ± 0.4; p < 0.001). These data suggest that wwCF and mwCF should be managed differently with respect to their sleep.

Detection of OSA Through the Application of Deep Learning on Polysomnography Data

Detection of OSA Through the Application of Deep Learning on Polysomnography Data

LPSGM: A Unified Flexible Large PSG Model for Sleep Staging and Mental Disorder Diagnosis.

We present the Large PSG Model (LPSGM), a unified and flexible framework for sleep staging and disease diagnosis using polysomnography (PSG) data. LPSGM is designed to address the challenges of cross-center generalization in sleep staging and to enable fine-tuning for downstream disease diagnosis tasks. LPSGM introduces a unified training framework for heterogeneous datasets and allows flexible channel input adjustments during inference. The model is first trained on 220,500 hours whole-night PSG from 16 public datasets, achieving robust sleep staging performance. It is then fine-tuned on target center data for various disease classification tasks, including narcolepsy diagnosis, anxiety and depression detection, and the classification of healthy versus depressed individuals. LPSGM outperforms baseline models on both sleep staging and disease diagnosis tasks. Our results demonstrate that LPSGM not only enhances sleep staging accuracy but also improves the diagnosis of sleep-related and psychiatric disorders, showing promise for clinical applications in sleep medicine and psychiatry.

Waiting for in-Lab Polysomnography May Unnecessarily Prolong Treatment Start in Patients with Moderate or Severe OSA at Home Sleep Apnea Testing.

The gold standard in obstructive sleep apnea (OSA) diagnostics is nocturnal full-night polysomnography (PSG). Due to high costs and high time effort portable respiratory polygraphy (PG or home sleep apnea testing-HSAT) has been developed. In contrast to PG the PSG gains relevant further information concerning sleep stages, arousals and leg movements. However, the role of PG in the diagnostic of OSA remains largely undefined. The aim of this study was to investigate the difference of PG- and PSG- related metrics in OSA, to understand if there is a difference in PG and PSG-based treatment decision and show up the time between performed PG and PSG. 99 consecutive patients with existing outpatient performed PG and followed PSG in our tertiary care otorhinolaryngology department between February 2020 and December 2023 were retrospectively assessed. All patients were treatment-naive at the time of consultation. The time between performed outpatient PG and PSG was calculated. Furthermore, clinical baseline parameter and PG as well as PSG data were evaluated. All data were then blinded presented with relevant comorbid diseases to two experts in sleep medicine in our tertiary care centre to decide whether PAP therapy was indicated or not. Mean AHI was significantly higher in PSG (32.32 ± 22.78/h) compared to PG (22.60 ± 15.12/h) (p<0.001). Mean duration between performed PG and PSG was 194.99 ± 131.96 days (range between 37 and 842 days). Only in two patients PAP-therapy was indicated with PG results but not with PSG results. Only in one case PAP-therapy was not indicated with PG results but with PSG results. These data suggest initiating OSA therapy based on PG results for patients with at least moderate OSA on PG, followed by a confirming PSG and a control PSG under treatment to avoid unnecessary prolongation of treatment start.

Distribution patterns of obstructive and central apnea in adults with obstructive sleep apnea-hypopnea syndrome

Objective: This study aimed to investigate the distribution patterns of obstructive and central apnea events in adults with obstructive sleep apnea-hypopnea syndrome (OSAHS). Methods: Utilizing polysomnography (PSG) data, we selected patients with a habitual snoring condition and categorized them into OSAHS and non-OSAHS groups. The OSAHS group was further stratified into mild, moderate, and severe subgroups based on disease severity. We compared the distribution of obstructive sleep apnea (OSA) and central sleep apnea (CSA) events between the groups, analyzing their correlation with the apnea-hypopnea index (AHI), age, gender, and other factors. Results: A total of 460 patients were enrolled for the study, 374 patients were in the OSAHS group and 86 patients in the non-OSAHS group. (1) There were statistical differences in AHI and obstructive apnea index (OAI) in all groups (P &lt; 0.05), with the OSAHS group having a higher AHI and OAI than the non-OSAHS group. (2) In the OSAHS group, OAI and AHI had the highest correlation (r = 0.884, P &lt; 0.01), and central apnea index (CAI) and AHI had the lowest correlation (r = 0.237, P &lt; 0.01); in the non-OSAHS group, OAI and AHI had the highest correlation (r = 0.520, P &lt; 0.01), and CAI and AHI had the lowest correlation (r = 0.312, P &lt; 0.01). (3) In the OSAHS group, OAI was significantly correlated with the severity of patient disease (r = 0.828, P &lt; 0.01), whereas CAI was marginally correlated with the severity of patient disease (r = 0.235, P &lt; 0.01), but it increased as the condition became more severe. Conclusion: Patients with OSAHS exhibited both CSA and OSA, with OAI having a significantly greater impact. OAI can serve as a valuable index for evaluating OSAHS severity, while CSA appears to be primarily influenced by the overall disease severity. Notably, patients with OSAHS had a higher CAI compared to those without OSAHS.