BackgroundSleep is an important component of human health and can be measured longitudinally using digital activity trackers. Further, decentralized digital research has the potential to provide a real-world picture of sleep in large populations.ObjectiveThis study examined whether longitudinal sleep patterns from activity trackers could predict risk of obstructive sleep apnea (OSA) and hypertension, as defined by the Berlin questionnaire and self-report, respectively.MethodsWe recruited adults aged ≥18 years nationwide to join our sleep-focused smartphone-based study, called the Research Framework for Exploring Sleep Health. Our sample of 391 adults predominantly comprised women (68%, 247/364) with a mean age of 48 (SD 13.62) years. Participants were asked to fill out health-related surveys, including the Berlin questionnaire and the Horne-Ostberg questionnaire for chronotype. Participants were asked to link their own activity tracker to the app to collect longitudinal sleep data.ResultsWe analyzed sleep data from 391 participants; the cohort was predominantly White (65%, 231/353) followed by multiracial (17%, 61/353) and Hispanic or Latino (6.5%, 23/353) participants. Collinearity testing showed that OSA risk and self-reported hypertension could be considered independently. Holding BMI at a fixed value, the odds of having high OSA risk increased by 159% for every 1-hour increase in weekday sleep variability (odds ratio [OR] 2.592, 95% CI 1.613-4.400; P<.001), and the odds of high OSA risk increased by 93% for each 1-hour increase in weekend sleep variability (OR 1.928, 95% CI 1.197-3.094; P=.01). The odds of having high OSA risk increased by 22% for each unit (kg/m2) increase in BMI, holding both weekday and weekend sleep at separate fixed values (OR 1.217, 95% CI 1.153-1.293; P<.001). Controlling for age, sex, and BMI, the odds of endorsing hypertension increased by 71% for every 1-hour increase in weekday sleep variability (OR 1.712, 95% CI 1.062-2.917; P=.03). Conversely, for weekend sleep, the odds of endorsing hypertension increased by 43% for a 1-hour increase in weekend sleep variability (OR 1.432, 95% CI 1.062-1.928; P=.04). Increased sleep variability predicted a high risk for both OSA and hypertension in this decentralized cohort, when using data from the Berlin questionnaire.ConclusionsOur study demonstrates the utility of decentralized digital health studies in sleep research. It highlights the potential of activity trackers to predict risk for OSA and hypertension without requiring other patient information or assessment. Sleep variability is gaining increasing importance in the context of sleep health. Digital devices have the potential to help individuals assess their risk for certain disorders.

Sleep medicine and research have profound roots and traditions in Europe. Medical and health structure standards, including those related to sleep medicine, are currently regulated by respective national organisations within the European Union. However, despite this fact, the organisation and practice of sleep medicine are far from harmonious across countries, being highly dependent on national medical systems. Moreover, there is a need to involve countries outside the European Union, especially those with underdeveloped sleep medicine, whose individual needs are diverse. The current article reviews the avenues for the movement of the European Sleep Research Society's (ESRS) Assembly of National Sleep Societies (ANSS) towards improved and standardised knowledge in sleep medicine both within and beyond the EU through different activities, including the implementation of the European Examination in Sleep Medicine and the Beyond Boundaries project. Sleep knowledge dissemination, harmonisation in various countries, and advocacy for improvement of the health care system quality are the goals of the current activities.

Abstract Adjusting for covariates in randomized controlled trials can enhance the credibility and efficiency of treatment effect estimation. However, handling numerous covariates and their complex (nonlinear) transformations poses a challenge. Motivated by the case study of the Best Apnea Interventions for Research (BestAIR) trial data from the National Sleep Research Resource (NSRR), where the number of covariates (p=114) is comparable to the sample size (N=196), we propose a principled covariate adjustment with variable selection (COADVISE) framework. COADVISE enables variable selection for covariates most relevant to the outcome while accommodating both linear and nonlinear adjustments. This framework ensures consistent estimates with improved efficiency over unadjusted estimators and provides robust variance estimation, even under outcome model misspecification. We demonstrate efficiency gains through theoretical analysis, extensive simulations, and a re-analysis of the BestAIR trial data to compare alternative variable selection strategies, offering cautionary recommendations. A user-friendly R package, Coadvise, is available to facilitate practical implementation.

Sleep is an important but overlooked component of animal behaviour, especially its social and conservation facets. Here, we use 15 years of data to comprehensively describe the roosting behaviour of cooperatively breeding birds and test hypotheses about its ecological and social determinants. We show that wild Arabian babbler groups in the Arava Desert of Israel preferred roosting in live plants with dense canopies (mostly Acacia tree spp. and reed clusters). Roosting sites were located in the inner areas of territories regardless of territorial conflicts. Groups almost always roosted in intimate huddles but tended to separate into sub-groups that roost in nearby trees as group size increased. Despite the abundance of suitable sites for roosting, each group only used an average of 2.4 main roosting sites within its territory. Social groups thus exhibited strong, non-random fidelity to specific roosting sites that extended over ≥4 group generations and ≥15 years. To the best of our knowledge, this is the longest roosting site fidelity shown for cooperatively breeding birds and mammals. This study stresses the importance of conserving roosting sites of species with strong site fidelity and lays the foundations for advanced sleep research in a highly cooperative species.

Sleep is a fundamental biological process essential for health and homeostasis. Traditionally investigated through laboratory-based polysomnography (PSG), sleep research has undergone a paradigm shift with the advent of wearable technologies that enable non-invasive, long-term, and real-world monitoring. This review traces the evolution from early analog and actigraphic methods to current multi-sensor and AI-driven wearable systems. We summarize major technological milestones, including the transition from movement-based to physiological and biochemical sensing, and the growing role of edge computing and deep learning in automated sleep staging. Comparative studies with PSG are discussed, alongside the strengths and limitations of emerging devices such as wristbands, rings, headbands, and camera-based systems. The clinical applications of wearable sleep monitors are examined in relation to remote patient management, personalized medicine, and large-scale population research. Finally, we outline future directions toward integrating multimodal biosensing, transparent algorithms, and standardized validation frameworks. By bridging laboratory precision with ecological validity, wearable technologies promise to redefine the gold standard for sleep monitoring, advancing both individualized care and population-level health assessment.

Detecting sleep stages is an essential part of sleep research; inaccuracies in the electroencephalography (EEG) signals will lead to many issues, including inaccurate disease detection, errors in medication description, and incorrect interpretations of a patient's EEG recordings. The proposed study aims to build a novel method for EEG sleep stages classification depending on a Horizontal Visibility Graph (HVG). The majority of current sleep stage classification approaches depend on frequency and time features. This study suggested a method using the HVG and discriminated features for sleep stages identification utilizing a single EEG channel. Firstly, the EEG signals are transformed to HVG, then extract five discriminated features (edgeCt, PathLength, diameter, degrees(1 to 10), and clustering (1 to 10). All features extracted from a single EEG channel (Pz-Oz) are passed to the Least Squares Support Vector Machine (LS-SVM) classifier. Based on the box plot, the convenient features are selected for each two-class pair to obtain the best classification accuracy. The 10-fold approach was used to assess the performance of the suggested model. Cross-validation of the accuracy of 98.10% for awake and Rapid Eye Movement (REM) two-class pairs.

International recommendations for sleep and circadian research in aging and Alzheimer's disease: A Delphi consensus study

Varesco, G, Germain, M, Szocs, S, Martin, A, Toussaint, P-M, and Simonelli, G. Exploring the relationship between sleep hygiene recommendations and outcomes in sleep, fatigue, and cognitive performance among student-athletes. J Strength Cond Res XX(X): 000-000, 2025-Student-athletes often neglect sleep because of academic and training demands, highlighting the value of interventions targeting poor sleep habits. However, sleep hygiene research remains limited to individualized approaches, limiting reproducibility and posing challenges for implementation in large groups, such as collegiate sports teams. In this 4-week study, we assessed the sleep health of an elite collegiate Canadian football team through questionnaires (week 1). Players with poor sleep habits (Pittsburgh Sleep Quality Index > 5; 24 out of 64) participated in a 2-week sleep hygiene intervention. After a familiarization session (week 2), they completed 2 weeks of sleep assessment (actigraphy, sleep diary) and a post-training test, including a fatigue 10-cm visual analog scale, psychomotor vigilance task (PVT), and paced visual serial addition test (PVSAT). Subjects were unaware to be selected based on their poor sleep habits. After week 3, athletes received feedback on poor sleep habits and standardized written, evidence-based sleep hygiene recommendations. They were asked to comply with these recommendations before repeating experimental procedures at the end of week 4. Improvements were observed only in perceived total sleep time (sleep duration; 7:43 ± 1:29 vs. 8:03 ± 1:34 h:mm, p = 0.045) and PVSAT performance (1.44 ± 1.8 vs. 1.36 ± 1.63 seconds, p = 0.004), while PVT speed decreased (3.6 ± 0.4 vs. 3.5 ± 0.4 Hz, p = 0.043). Objective sleep duration (6:53 ± 1:19 vs. 7:04 ± 1:19 h:mm, p = 0.21), fatigue (6 ± 2.1 vs. 5.8 ± 2.1 AU, p = 0.55), and other outcomes remained unchanged (p > 0.37). These results indicate that unsupervised and standardized sleep hygiene recommendations were not associated with changes in sleep habits, with changes in cognitive performance and fatigue level that could not be reconducted to better sleep.

Accuracy of large language models in data extraction from randomized controlled trials in sleep medicine: A proof-of-concept study.

This study proposes a Personalized Deep Learning-Based Sleep Recommendation System Using Lifelog Data (PDSRS-LD). Traditional sleep research primarily relies on bio signals such as EEG and ECG recorded during sleep but often fails to sufficiently reflect the influence of daily activities on sleep quality. To address this limitation, we collect lifelog data such as stress levels, fatigue, and sleep satisfaction via wearable devices and use them to construct individual user profiles. Subsequently, real sleep data obtained from an AI-powered motion bed are incorporated for secondary training to enhance recommendation performance. PDSRS-LD considers comprehensive user data, including gender, age, and physical activity, to analyze the relationships among sleep quality, stress, and fatigue. Based on this analysis, the system provides personalized sleep improvement strategies. Experimental results demonstrate that the proposed system outperforms existing models in terms of F1 score and Average Precision (mAP). These results suggest that PDSRS-LD is effective for real-time, user-centric sleep management and holds significant potential for integration into future smart healthcare systems.

Respiratory rate (RR) is a valuable, yet underexploited lung functional parameter in preclinical lung research, aiding drug toxicity studies, lung disease assessments, stress, pain, and sleep research. It may also enhance translatability between animal and human studies. Longitudinal micro-computed tomography (microCT) lung data acquisitions not only contain spatial information on lung disease, volumes and patterns, but also temporal information covering many breathing cycles. This enables reliable and non-invasive extraction of lung morphological and functional biomarkers, including RR, with a single measurement from free-breathing animals, crucial for accurate measurements. Here, we aimed to develop a non-invasive pipeline, for longitudinal RR monitoring as a biomarker for lung function and pathology based on the X-ray projections of lung microCT acquisitions. First, we mechanically ventilated a mouse and scanned it using microCT at different breathing rates, 60 to 185 breaths per minute (bpm), serving as ground-truth data for our RR measurements. Next, we obtained raw intensity curves from these ground-truth X-ray projections, which contained noise and signals from multiple sources such as respiratory and cardiac cycles. To find the optimal algorithm and isolate the respiratory signals, we post-processed these raw intensity curves with different signal processing techniques. Adept at handling non-uniformly sampled signals in time domain, the Lomb-Scargle (LS) algorithm outperformed the other signal processing techniques, exhibiting robust prediction of RR with an error margin of 3%. Next, we applied this pipeline to benchmark the longitudinal RR data as a biomarker of lung damage and repair in a mouse model of lung epithelial injury. Our RR monitoring pipeline detected a transient loss of lung function in diseased mice, marked by a temporary RR decrease and a simultaneous increase in total lung and aerated lung volumes. Adopting this X-ray-based pipeline would allow lung researchers to non-invasively collect both morphological and functional data in a single measurement, improving insights into lung disease progression and host response thereto by providing relevant biomarkers. This approach contributes to facilitating translation of preclinical study results toward clinical trials.

Sleep is essential for optimal functioning and health. Interconnected to multiple biological, psychological, and socio-environmental factors (i.e., biopsychosocial factors), the multidimensional nature of sleep is rarely capitalized on in research. Here, we deployed a data-driven approach to identify sleep-biopsychosocial profiles that linked self-reported sleep patterns to inter-individual variability in health, cognition, and lifestyle factors in 770 healthy young adults. We uncovered five profiles, including two profiles reflecting general psychopathology associated with either reports of general poor sleep or an absence of sleep complaints (i.e., sleep resilience), respectively. The three other profiles were driven by the use of sleep aids and social satisfaction, sleep duration, and cognitive performance, and sleep disturbance linked to cognition and mental health. Furthermore, identified sleep-biopsychosocial profiles displayed unique patterns of brain network organization. In particular, somatomotor network connectivity alterations were involved in the relationships between sleep and biopsychosocial factors. These profiles can potentially untangle the interplay between individuals' variability in sleep, health, cognition, and lifestyle-equipping research and clinical settings to better support individual's well-being.

An overview of longitudinal data analyses in sleep research.

Perceptions of sleep and sleep research among African American adults.

BackgroundSleep plays a critical role in overall health and well-being. While most sleep research focuses on high-income countries, there is limited knowledge about sleep quality in Sub-Saharan Africa (SSA), especially among women living in urban informal settlements. Many factors, including physical, psychological, cultural, and environmental influences, can affect sleep quality. This study, which uses Bronfenbrenner’s ecological model, aims to explore the prevalence of sleep disturbances and self-reported factors associated with poor sleep quality among a representative sample of 800 women living in two informal settlements in Nairobi, Kenya.MethodsThe data, collected in September 2022, are from the baseline assessment of an 18-month longitudinal cohort study examining mental health and climate change among women living in two informal settlements in Nairobi–Mathare and Kibera. Items from the Brief Pittsburgh Sleep Quality Index (B-PSQI) were collected to examine women’s sleep habits and quality. Quality of sleep scores were calculated. We used t-tests, bivariate regressions, and ANOVAs to assess the bivariate associations between key predictors of poor sleep with the Brief Pittsburgh Sleep Quality Index (B-PSQI) score. We also conducted a cross-sectional multivariable regression analysis to explore the factors influencing sleep disturbances. Open-ended questions were asked about factors contributing to sleep disturbance, and a thematic analysis was conducted to summarize the findings.Findings29% of women (N = 229) met the criteria for poor-quality sleep. Open-ended findings identify stress as the main factor affecting sleep. Childcare, financial instability, physical health, climate, grief, and loss also impacted women’s sleep. Significant quantitative predictors of poor sleep quality among women included the severity of disability, depression, and food insecurity. Anxiety also showed a trend toward significance, underscoring the complex interplay of physical, mental, and socioeconomic factors on sleep.InterpretationThis study underscores the need for further research on sleep quality among women in SSA’s informal settlements. By enhancing understanding and awareness of sleep’s health impacts, policymakers and interventionists can develop more effective interventions tailored to the unique challenges faced by this population. Our findings contribute to the knowledge base, supporting the creation of targeted policies and practices that address and improve sleep quality for women in these communities.

This study proposes a method to automatically identify dream experience (DE) and no experience (NE) during the sleep stage N2 of nonrapid eye movement (NREM). We investigated the use of machine learning (ML) to automatically identify when a subject is having a dream during NREM from electroencephalography (EEG) signals. We use permutation-based channel selection to identify the most informative EEG channels for the classification of DE and NE and to select a set of channels that allow focus on the most important areas of the brain at the scalp level. The results show that when the ML models are trained on a balanced dataset containing both DE and NE reports, along with high-density EEG, they can achieve a classification performance of up to 0.94 in accuracy, F1 score, precision and recall, an Area Under the Receiver Operating Characteristic of 0.97, and a kappa of 0.88. Performance decreases while we reduce the number of channels, but it remains like using up to 30–40 EEG channels. We show that ML models trained on high-density EEG to classify NE and DE can identify whether a subject was dreaming, achieving an accuracy of 0.7 on a separate set of dream reports where subjects reported a dream experience without recall, and channel selection methods have shown that performance could increase by 0.02 when EEG channels are removed from the occipital area. Our results show a high classification performance for automatic dream detection and the need to reduce the number of EEG channels needed to create the ML models, thus obtaining low-cost portable devices that can be used in real-life scenarios.Statement of SignificanceDreaming during non-REM sleep is difficult to detect objectively. Here, we show that machine learning models leveraging high-density EEG can accurately classify the presence vs absence of dream experiences during N2 sleep. A key finding is that this high performance is maintained with a substantially reduced channel set, establishing the feasibility of portable, low-density EEG systems for real-time dream detection. Our work provides a critical objective correlate for subjective dreaming and opens new avenues for practical applications in sleep research and clinical diagnostics.

RimeSleepNet: A hybrid deep learning network for s-EEG sleep stage classification.

Clarifying the role of metformin and statistical interpretation in sleep quality research among patients with diabetes

Immigrant Status, Socioeconomic Status, and Sleep Disparities in Early Adolescence: Findings From the National Adolescent Brain Cognitive Development Study.

Sleep is a fundamental biological process essential to health, particularly during early life when sleep patterns are developing and sleep disorders are common. Yet pediatric sleep research is hindered by a lack of large-scale, high-quality polysomnography (PSG) datasets. To address this need, we introduce the Boston Children's Hospital (BCH) Sleep Corpus-the largest pediatric PSG dataset available-comprising 15 695 overnight recordings from 12 640 unique patients (median age ~ 6years). The dataset includes 16.7 million annotated sleep stages, 2.25 million respiratory, arousal, and limb movement events, and over 11 000 patient diagnoses linked through de-identified electronic health records. Each PSG has a median duration of 8.9hours, totaling 139 208hours of EEG data. Sleep staging follows American Academy of Sleep Medicine guidelines and reveals age-related trends: REM sleep decreases from 33.5% in neonates to 16.3% in teenagers, while N2 sleep increases from 21.7% to 35.4%. Central apneas decline with age, while obstructive hypopneas and respiratory effort related arousals events rise. Limb movements are not scored in <1yr but remain at around 30 per PSG across older age groups. We also present age- and region-specific EEG spectral norms and respiratory event trends across the pediatric age range. The dataset is organized in Brain Imaging Data Structure (BIDS) format and publicly available via the Brain Data Science Platform. The dataset provides a valuable resource for improving our scientific understanding of pediatric sleep and developing automated PSG analysis with artificial intelligence tools.