Journal of Sleep Research

Sleep regularity is an important but under-measured dimension of sleep health. Objective indices from actigraphy or wearables are robust but resource-intensive. The Sleep Regularity Questionnaire (SRQ) offers a brief subjective tool, but its validity against objective and diary-based indices in healthy adults is unclear. In Part 1, 31 adults wore a smart ring continuously for 21 nights. Device-derived regularity metrics included the Sleep Regularity Index (SRI), interdaily stability (IS), social jetlag (SJL), composite phase deviation (CPD), and the standard deviation of sleep onset and wake time. In Part 2, 52 adults completed a one-week sleep diary, from which variability in sleep timing, total sleep time (TST), SJL and nightly perceived sleep quality were derived. All participants completed the SRQ and Brief Pittsburgh Sleep Quality Index (B-PSQI). In Part 1, associations between SRQ scores and device-derived SRI, IS, SJL, CPD and timing variability were small (absolute r [≤] 0.36). Higher SRQ Global and Sleep Continuity scores were moderately associated with better B-PSQI global scores (r -0.37 to -0.44). In Part 2, SRQ Global and Circadian Regularity showed small-to-moderate associations with higher diary-rated sleep quality and lower bedtime variability (r {approx} 0.40 and -0.32 to -0.34), while correlations with other diary metrics and B-PSQI were weak (absolute r [≤] 0.25). The SRQ shows modest convergent validity with diary-based timing variability and perceived sleep quality, but only weak correspondence with smart ring-based sleep regularity indices. It is likely to complement, rather than replace, objective monitoring in healthy adults with relatively regular sleep-wake patterns.

PuhrposeTo evaluate the short- and long-term cross-sectional associations between COVID-19 infection and multidimensional sleep health. MethodsData from the COVID-19 Outbreak Public Evaluation (COPE) initiative were used to examine the association between a novel multidimensional sleep health measure (COPE Multidimensional Sleep Health Scale, CMSHS) modeled from the RuSATED instrument and (1) COVID-19 infection and (2) post-acute sequelae of SARS-CoV-2 infection (PASC). ResultsData from 11,326 respondents were used for this study. The cohort was comprised of 51% women, 61% non-Hispanic White, and 17% Hispanic adults. COVID-19 infection was more prevalent among participants who had not received a booster vaccination (55.4% vs. 30.2%, p<0.001); the number of comorbid conditions was higher among those who had been infected (2.2% vs. 1.7%, p<0.001). Participants with COVID-19 infection had significantly lower CMSHS scores indicative of worse sleep health compared with uninfected participants (3.52 {+/-} 1.37 vs. 3.78 {+/-} 1.30; p < 0.001). Participants with PASC had lower CMSHS scores in comparison to those without PASC (2.72 {+/-} 1.30 vs. 3.82 {+/-} 1.28, p<0.001). In adjusted models, a progressive decline in CMSHS scores was observed over 12 months following infection (3.52 {+/-} 0.05 vs. 2.98 {+/-} 0.04; p < 0.001 for <1 month vs. 6-12 months). ConclusionCompared with uninfected individuals, multidimensional sleep health was worse among persons who had a COVID-19 infection. Individuals with PASC had greater and persistent reductions in sleep health for up to 12 months post-infection. Brief summaryO_LISeveral studies have examined the negative effects of COVID-19 on sleep, however the effects of COVID-19 infection on multidimensional sleep health remain poorly understood as do these associations over time. Using a large, population-based cohort, this study evaluates short- and long-term effects of Covid-19 infection on overall sleep health. C_LIO_LIThe study provides evidence that COVID-19 infection is associated with impairments in overall sleep health, with effects persisting up to 12 months post-infection. The findings in this study demonstrate that poor sleep health is an important long-term consequence of COVID-19 infection and emphasizes the need for sleep assessment among patients affected by COVID-19. C_LI

Sleep loss is a major contributor to traffic accidents, yet the differential effects of acute versus cumulative 8-hour sleep loss on vehicle control remain poorly quantified. This study leveraged tightly controlled sleep manipulation together with highly immersive driving simulation to directly compare sleep-loss modes under naturalistic driving conditions. To evaluate subjective sleepiness, driving performance and vigilance, healthy male participants completed a randomized within-subject crossover sleep design, comprising normal sleep, cumulative sleep restriction (four nights with 2 h reduced sleep per night), and acute total sleep deprivation (one night without sleep), each followed by a simulated nighttime-to-early-morning driving protocol segmented into modules. Linear mixed-effects models evaluated condition effects and time-on-task dynamics. Acute total sleep deprivation produced pronounced impairments in vehicle control. Most prominently, lateral control deteriorated rapidly under acute deprivation, with effect magnitudes comparable to those reported at high blood alcohol concentrations. Steering instability and integrated driving performance showed convergent deterioration. In contrast, cumulative sleep restriction resulted in smaller and less consistent changes, while vigilance task performance and speed variability were comparatively preserved across all conditions. Despite increased subjective sleepiness, cumulative sleep loss provoked relatively small changes in driving performance, whereas acute sleep deprivation produced a disproportionate risk to driving safety by selectively and rapidly degrading operational vehicle control. These findings, obtained under ecologically relevant sleep and driving conditions, underscore functional and potentially underlying physiological differences between acute and cumulative sleep loss and, practically, highlight the importance of distinguishing between them when considering sleep-loss-related driving impairment and safety risk. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=109 SRC="FIGDIR/small/26343648v1_ufig1.gif" ALT="Figure 1"> View larger version (49K): org.highwire.dtl.DTLVardef@e136c3org.highwire.dtl.DTLVardef@1d241c9org.highwire.dtl.DTLVardef@dd17f5org.highwire.dtl.DTLVardef@14ff50a_HPS_FORMAT_FIGEXP M_FIG C_FIG Clinical Trials - name, URL and registrationO_ST_ABSClinical Trial nameC_ST_ABSStudy of Identification of Metabolomics-based Sleepiness Markers for Risk Prevention and Traffic Safety - ClinicalTrials.gov Identifier NCT05585515, released on 18.10.2022, https://clinicaltrials.gov/study/NCT05585515; Swiss National Clinical Trial Portal SNCTP000005089, registered on 12.08.2022, https://www.humanforschung-schweiz.ch/de/studiensuche/studien-detail/58930 IRB StatementThis study was approved by the local ethics committee (Kantonale Ethikkommission Zurich, reference number 2022-01273). No identifying images or other personal or clinical details of participants are presented here or will be presented in reports of the trial results. Informed consent materials are available from the corresponding author on request. We confirm all methods were performed in accordance to the Declaration of Helsinki and its later amendments. Statement of significanceSleep loss is common, but not all forms of sleep loss affect driving in the same way. This study shows that a single night without sleep and several nights of restricted sleep have fundamentally different consequences for vehicle control, despite producing significant feelings of sleepiness. By combining controlled sleep manipulation with immersive, ecologically relevant driving assessment, the work reveals that acute sleep deprivation uniquely disrupts the ability to maintain stable control of a vehicle, whereas cumulative sleep loss produces more limited performance changes. These findings challenge assumptions based solely on subjective sleepiness and highlight the need to distinguish between sleep-loss patterns when evaluating driving safety. Future work should address how drivers recognize and respond to these distinct forms of impairment.

BackgroundExcessive daytime sleepiness (EDS) is a major yet under-recognized contributor to road traffic accidents. Traditional diagnostic tools, such as the Maintenance of Wakefulness Test (MWT), assess wakefulness under passive conditions but may not accurately reflect real-world driving risks. To address this gap, we have piloted a Driving Simulation-based MWT (DS-MWT), designed to evaluate sleepiness in an ecologically valid driving scenario. The present study aims to validate the novel DS-MWT against the classical MWT in a functionally relevant cohort of patients with obstructive sleep apnoea (OSA). MethodsThe present monocentric, randomized, controlled, within-subject crossover trial will include 54 participants: 36 patients with OSA undergoing[&ge;] 7-day CPAP withdrawal (W) or continuation (C), and 18 healthy controls. The study employs a well-established CPAP-withdrawal model in patients with prior optimal treatment adherence to transiently induce EDS under controlled conditions. A healthy control group is included to enable between-group comparisons. The primary expected outcome is the difference in mean latencies between MWT and DS-MWT, determined during four standardized test sessions per condition. Secondary exploratory outcomes are defined as the presence, direction, and magnitude of differences or correlations between treatment status (CPAP withdrawal vs. continuation) and driving performance metrics (e.g., lateral position, speed, lane departures, etc.), EEG and eye-tracking features, subjective sleepiness scores, at-home polysomnography (PSG) parameters, and metabolomic biomarkers (saliva, exhaled breath and dried blood spots). Data will be analyzed using linear mixed models, repeated-measures ANOVA, and predictive modeling with cross-validation. DiscussionThis trial addresses a critical limitation in sleep and traffic medicine by introducing a realistic, supposedly more ecologically valid alternative to standard sleepiness assessment tools. The DS-MWT may enhance clinical decision-making regarding fitness to drive (FTD) and provide a framework for identifying physiological and behavioral markers of sleepiness in realistic conditions. Trial registrationClinicalTrials.gov Identifier: NCT06872593, released on 12.03.2025, https://clinicaltrials.gov/study/NCT06872593 Swiss National Clinical Trial Portal SNCTP000006301, released on 19.03.2025, https://www.humanforschung-schweiz.ch/en/trial-search/study-detail/66469

Sleep architecture is essential for metabolic and cardiovascular health, yet the impact of day-to-day dietary variation on objective sleep physiology remains unclear. Using 4.8 thousand person-nights with real-time dietary logs and multi-stage wearable sleep recordings, we examined how prior-day nutrition relates to next-night sleep under free-living conditions. Higher fiber density was associated with increased restorative sleep, including +0.59 pp deep sleep, +0.76 pp REM sleep, -1.35 pp light sleep, and -1.14 bpm lower mean nocturnal heart rate. Greater plant diversity and higher whole-plant food intake were similarly associated with lower nocturnal heart rate (-0.72 to -0.94 bpm). Meal-timing behaviors primarily influenced sleep duration, sleep-onset latency, and autonomic tone: heavier evening meals were associated with +7.7 min longer total sleep time and +0.73 bpm higher nocturnal heart rate. In contrast, short-term variation in macronutrient energy distribution and micronutrient consumption showed no robust associations with sleep outcomes. When analyses were restricted to more extreme dietary contrasts, effect magnitudes increased while remaining directionally consistent. These findings indicate that routine daily dietary choices, particularly plant-forward composition and meal timing, have immediate and measurable effects on objective sleep architecture.

BackgroundSleep regularity plays a significant role in mental health, but its association with anxiety or depression symptoms in midlife and whether movement behaviors moderate these associations is unclear. We examined how variability in sleep timing and duration is associated with these symptoms and tested the moderating effects of physical activity and sedentary time. MethodsWe analyzed data from 3,556 46-year-old participants in the Northern Finland Birth Cohort 1966. Sleep regularity was derived from 7 days of accelerometer-measured sleep timing (midpoint of sleep period) and sleep duration, expressed as standard deviations and categorized as regular, moderately irregular, or highly irregular. Anxiety and depression symptoms were assessed using the Hopkins Symptom Checklist-25. Associations were examined using multivariable linear regression adjusted for gender, marital status, education, smoking, alcohol use, medication, chronotype, work schedule, and physical activity or sedentary time derived from the same accelerometer data, with moderation tested using interaction terms (B coefficients with 95% confidence intervals). ResultsGreater sleep timing irregularity was associated with higher anxiety symptoms (B = 0.025, 95% CI [0.007, 0.043], p = 0.001), and with depressive symptoms only when models included sedentary time (B = 0.035, 95% CI [0.017, 0.053], p < 0.001). Sleep duration irregularity showed no significant associations. Physical activity and sedentary time did not moderate these relationships. ConclusionsSleep timing regularity was more consistently associated with mental health outcomes, especially anxiety, than sleep duration variability. Maintaining regular sleep timing may support anxiety management in midlife, highlighting its potential relevance for preventive intervention.

ObjectivesTo determine associations of common infections with self-reported sleep and explore whether these links differ by race in a diverse cohort. MethodsWe studied 602 participants from the Baltimore Epidemiologic Catchment Area Study (mean age 59.0{+/-}12.8 years; 36.9% male; 35.6% minoritized adults) with data on common infections (IgG antibodies to herpes simplex virus type 1, cytomegalovirus, varicella zoster virus, Epstein-Barr virus, and Toxoplasma gondii) and self-reported symptoms of insomnia (i.e., difficulty falling or staying asleep and early awakening), hypersomnia, and sleep duration. ResultsWe observed differences by race in the association of common infections with sleep disturbances (p-values for interactions<0.05). In fully-adjusted models, among White participants, those who were seropositive for CMV had 51% higher odds of insomnia symptoms (OR=1.51, 95% CI: 0.88, 2.60) and 26% higher odds of early awakening (OR=1.26, 95% CI: 0.67, 2.39). Among minoritized adults, however, CMV seropositivity was associated with 75% lower odds of insomnia symptoms (OR=0.25, 95% CI: 0.09, 0.67) and 73% lower odds of early awakening (OR=0.27, 95% CI: 0.10, 0.74). Seropositivity for a higher number of infections ({beta}=0.24, 95% CI: -0.01, 0.48) or for TOX alone ({beta}=0.64, 95% CI: 0.08, 1.20) was associated with longer sleep among minoritized participants, but shorter sleep among White participants. ConclusionsIn middle-aged and older adults, common infections are differentially associated with sleep disturbances and duration, with infections linked to better sleep profiles and longer sleep among minoritized adults, but poorer sleep among White adults. Research is needed to identify the sources of these differences.

BackgroundNocturnal hypoglycaemia (NH) is a common and challenging complication in Type 1 Diabetes (T1D), disrupting blood glucose control and sleep physiology. Its real-world impact on sleep architecture remains poorly characterised. Consumer wearables offer a way to examine these associations under free-living conditions, providing detailed insight into behavioural and physiological responses to nocturnal blood glucose fluctuations. This study aims to assess how wearable-derived sleep metrics and physiological features could be used as indicators of NH, including the effects of how low blood glucose levels fall during hypoglycaemic events and the associated pre-event changes. MethodsWe conducted a comparative observational analysis of paired continuous glucose monitoring (CGM) and Garmin smartwatch data collected over 12 weeks from 17 adults with T1D. Nights were categorised as normoglycaemia, hyperglycaemia, or hypoglycaemia Level 1 ([&ge;]3.1 and <3.9 mmol/L), and hypoglycaemia Level 2 (<3.0 mmol/L). Thirteen sleep metrics, including total sleep time, wake after sleep onset (WASO), sleep-stage proportions, fragmentation indices, and physiological features such as heart rate, were compared using non-parametric tests. Pre-hypoglycaemic event analyses examined 60-minute and 15-minute windows preceding hypoglycaemia to identify early deviations in sleep and physiological metrics. ResultsAcross 573 nights, 17.5% involved Level 1 and 7.3% Level 2 hypoglycaemia. Level 2 hypoglycaemia was associated with 31 minutes less wakefulness, 17-25 minutes more REM, and up to 74% more deep sleep compared with normo-glycaemic nights. Sleep efficiency increased during hypoglycaemic events despite greater fragmentation. Pre-hypoglycaemic episode analyses revealed shorter awake and light-sleep bouts, as well as a 9.8% higher heart rate, preceding Level 2 episodes. ConclusionsWearable-derived sleep and physiological signals reveal clear intraindividual changes both before and during NH. Our findings indicate that Level 2 episodes are associated with deeper sleep and reduced behavioural arousal, suggesting that CGM alarms may be less effective at waking individuals during level2 NH. By characterising pre-hypoglycaemic changes that differ based on hypoglycaemia level, this work provides preliminary evidence for personalised, wearable-based early-warning systems. Such approaches could help distinguish nocturnal hypoglycaemic events and support more effective alerting, particularly in settings with limited or no access to CGM. Author SummaryO_ST_ABSWhy was this study done?C_ST_ABSPeople with Type 1 Diabetes (T1D) frequently experience nocturnal hypoglycaemia (low blood glucose at night), a dangerous event that often goes unnoticed because individuals are less able to recognise symptoms or wake up during sleep. These events also disrupt sleep in ways that are not well characterised under real-world conditions. Limited access to continuous glucose monitoring (CGM), especially in low- and middle-income countries, highlights the need for affordable alternatives to ensure nighttime safety. What did we do and find?Using more than 500 nights of paired smartwatch and CGM data, we investigated how sleep features change when blood glucose levels fall overnight. We found that hypoglycaemic nights show distinct alterations in sleep architecture, including increased REM and deep sleep, and greater micro-fragmentation. A key finding was that Level 2 hypoglycaemia was associated with deeper sleep and reduced wakefulness. This pattern indicates that individuals may be less likely to awaken during more severe events, even when alarms are present. Pre-hypoglycaemic episode analysis revealed additional early-warning signals, such as shorter awake and light-sleep bouts and elevated heart rate, before level 2 hypoglycaemia occurred. What do these findings mean?Smartwatches can capture sleep-based changes that appear before and during nocturnal hypoglycaemia. Because deeper sleep during Level 2 episodes may reduce responsiveness to CGM alerts, these results suggest that current alarm approaches could be improved by incorporating sleep features alongside glucose data. Such sleep-informed detection may enhance the reliability of hypoglycaemia alerts, reduce missed events during deep sleep, and provide a foundation for low-cost early-warning systems in settings where CGM is unavailable or unaffordable. Further research is needed in larger and more diverse populations, but this work provides early evidence that wearable-derived sleep features can meaningfully strengthen nocturnal hypoglycaemia detection.

BackgroundWhile accurate sleep measurement is vital for older adults, the validity of actigraphy (AG) in free-living environments remains controversial, particularly given the flexible sleep-wake schedules common in this demographic. To address this uncertainty, we assessed the accuracy of wrist AG against in-home portable electroencephalography (EEG) among community-dwelling older adults. MethodsCommunity-dwelling older adults underwent concurrent sleep monitoring using a portable EEG device and a wrist-worn AG for five consecutive nights whenever possible, with monitoring extended to up to seven nights when feasible. Key sleep parameters, including total sleep time (TST), sleep onset latency (SOL), wake after sleep onset (WASO), and sleep efficiency, were derived from both devices. Measurement agreement was assessed using Bland-Altman plots and multilevel modeling, while reliability and accuracy were quantified via intraclass correlation coefficients (ICCs) and mean absolute percentage error (MAPE). ResultsForty-nine adults contributed 217 nights of recordings. On average, AG slightly overestimated TST and sleep efficiency and underestimated SOL and WASO compared with EEG. Single-measure ICCs were 0.73 for TST and 0.38 for WASO (0.84 and 0.55 for averages across nights), and the MAPE was 11% for TST but exceeded 50% for SOL and WASO, indicating poor accuracy for these indices. ConclusionIn community-dwelling older adults, wrist AG yielded acceptably accurate estimates of average TST, supporting its use in epidemiological monitoring of sleep duration. However, large errors for SOL and WASO indicate that portable EEG- or polysomnography-based assessment remains indispensable when precise evaluation of sleep initiation and nocturnal wakefulness is required.

We previously documented and released a benchmark dataset for machine learning research on sleep stage classification [1]. Subsequently, it was pointed out in a preprint [2] that some recordings in the National Sleep Research Resource [3] include only binary wake-sleep annotations, instead of full sleep stage scoring using the Rechtschaffen and Kales (R&K) [4] or American Academy of Sleep Medicine (AASM) [5] standards. Because wake-sleep labels are an ontological mismatch and not just label noise, they do not belong in a dataset designed for full sleep stage classification. Therefore, we have updated our benchmark dataset (henceforth known as benchmark v1.0) to replace 16 recordings with suitable recordings from age- and sex-matched subjects, while all other dataset selection criteria and distributions have been preserved. Additionally, the total number of recordings and the composition of the training, validation, and testing sets remain unchanged. While this update is a minor revision, we want to distinguish its use from v1.0, and therefore have titled this update as benchmark v1.1. The file listings are provided on the GitHub repository (https://github.com/adammj/ecg-sleep-staging/).

Data-driven sleep research relies on polysomnography data from various public repositories and vendor systems. Yet the lack of standardized access methods creates substantial barriers to multi-dataset research, method reuse, and reproducibility. We present SleePyPhases, an open-source Python framework providing unified access to multiple sleep data repositories. It offers integrated data harmonization, configuration-based preprocessing, and the development of machine learning pipelines. The framework unifies channel naming, annotation semantics, and data formats across several public repositories (including SHHS, MESA, MrOS, PhysioNet, and SleepEDF) and commercial vendor formats (Philips Alice and Somnomedics Domino). We validated the framework by reproducing five published sleep analysis studies covering diverse datasets, sleep scoring tasks (sleep staging, arousal, leg movement, respiratory event detection), preprocessing methods (signal preprocessing and spectrograms), machine learning methods (supervised and unsupervised learning), and model architectures (convolutional, recurrent, and transformer networks). Four reproductions achieved near-identical results, confirming data fidelity and methodological flexibility. SleePyPhases is open-source and provides a foundation for reproducible sleep research, enabling researchers to focus on scientific questions rather than data infrastructure.

Recent methodological advances have enabled minimally invasive, convenient self-assessment of central circadian phase in home-based settings [1]. In this report, we describe salivary melatonin onset secretion profiles in 261 participants (36% male; age range, 9-83 years) who were recruited from 18 clinics in North America or by self-referral. All participants received a standardized, at-home central circadian phase assessment kit by mail or directly through their provider, along with written and video sample collection instructions. The standard protocol consisted of 7 or 9 saliva samples collected at 1-hour intervals in dim light on a single occasion, starting up to 7 hours prior to an individuals habitual bedtime and proceeding until 1-5 hours past bedtime. The number of samples and the start and end times were highly dependent on the nature and timing of the individuals primary concern and were dictated by the referring sleep-health provider. Samples were frozen immediately after collection and then returned to a central laboratory via 2-day shipping after an additional 48 hours of home freezing. Samples were subsequently kept frozen until the day of melatonin assay. Of the 261 participants, 91 (34.9%) exhibited DLMOs within the Predicted Onset Window (POW). The remaining 170 participants (65.2%) showed profiles that did not meet criteria for a predicted, aligned phase onset, including 162 (62.1% of the total sample) that could be classified into atypical phenotypes, distributed as follows: 28 showed a Delayed Melatonin Onset (10.7%), 28 showed an Advanced Melatonin Onset (10.7%), 27 exhibited Hypermelatoninemia (10.3%), 36 exhibited Hypomelatoninemia (13.8%), and 43 presented with Irregular/Multipeak profiles (16.5%). The remaining 8 participants (3.1%) exhibited profiles that could not be reliably classified. These results highlight the unexpectedly high proportion of non-predicted melatonin patterns and demonstrate that over 80% of profiles can be reliably assigned to clinically meaningful circadian phenotypes. The Discussion explores how real-world melatonin profiling can identify relevant circadian contributors to sleep disruption that symptom reports alone may fail to detect. We also describe and discuss these phenotypes in detail, considering their biological contexts and potential clinical relevance.

Sleep is essential for physical and mental health, yet large-scale assessment of sleep stages and sleep apnea is limited by the cost and burden of clinical polysomnography. Wrist accelerometry provides a scalable lower-fidelity alternative, but its usefulness is highly reliant on modeling choices. Generative self-supervised learning has remained unexplored for this purpose. To address this, we developed and pretrained physiology-aware masked autoencoders to capture pulse-and respiration-related motion using [~]700,000 days of wrist accelerometry from 108,904 recordings in the UK Biobank cohort. This effort culminated in AcceleRest, a transformer model pretrained using a respiratory amplification objective. Performance was validated against polysomnography across 478 recordings from six cohorts and devices including independent external test cohorts. AcceleRest feature vectors enabled linear wake-NREM-REM sleep staging with a macro F1 score of 0.69 and respiratory event detection with a macro F1 of 0.56. The combined model outputs enabled sleep apnea severity evaluation with a 67% sensitivity and 96% specificity for severe apnea. Overall agreement between polysomnography and AcceleRest showed a bias of 0.8 min for total sleep duration, with 95% limits of agreement (LoA) of -101.6 to 103.2 min, and 32.5 min for REM sleep duration and 95% LoA of -68.7 to 133.6 min. These findings demonstrate that physiology-aware pretraining can enable robust and clinically meaningful sleep phenotyping from wrist accelerometers, supporting scalable screening and longitudinal monitoring of sleep health. To the best of our knowledge, AcceleRest represents the first wrist accelerometry model for joint sleep stage and apnea evaluation. All code and models will be made available upon final peer-reviewed publication.

Background/ObjectivesElectromyography (EMG), video-polysomnography (vPSG), and wrist actigraphy are each used to develop diagnostic algorithms for Rapid eye movement sleep behavior disorder (RBD). However, the extent to which they capture overlapping versus distinct motor phenomena remains unknown. We evaluated the respective contributions of actigraphy, EMG and vPSG to the measurement of REM-sleep motor activity. MethodsSeventeen adults with RBD (Mount Sinai n = 9; Stanford n = 8) and eight control participants from an open Newcastle dataset underwent vPSG and concomitant wrist actigraphy. Flexor digitorum superficialis EMG activity and video-detected movements were manually scored in 3-second mini epochs. Actigraphy was quantified using an acceleration-magnitude-based activity count model. Statistical and agreement analyses were performed to assess the motor events captured by all three, any two, or by each modality independently during REM sleep. ResultsIn participants with RBD, actigraphy-derived movement load was significantly higher during REM sleep than during non-REM stages, a pattern not observed in control participants. Across 12,941 3-second mini epochs, EMG, actigraphy, and video detected 1,703, 1,613, and 811 motor events, of which 413 were detected concurrently by all three modalities. Pairwise agreement was moderate and increased from EMG-actigraphy ({kappa} = 0.27 {+/-} 0.10) to actigraphy-video ({kappa} = 0.41 {+/-} 0.12) and EMG-video ({kappa} = 0.45 {+/-} 0.15). Of EMG-detected events, 49.0% were also detected by actigraphy; of actigraphy-detected events, 37.2% were detected by EMG and 34.9% by video. Actigraphy activity counts were highest for events detected by all three modalities and lowest for actigraphy-only events. ConclusionActigraphy-measured REM-related motor activity was elevated in RBD but not in controls. EMG, actigraphy, and video captured partially overlapping motor events in RBD patient, with actigraphy showing the highest sensitivity and manually scored video the lowest.

RationaleObstructive sleep apnea (OSA) is linked to cardiovascular, metabolic, and cognitive morbidity. Although COVID-19 has been associated with long-term respiratory and neurological sequelae, its role in precipitating new-onset OSA remains unclear. ObjectivesTo evaluate whether SARS-CoV-2 infection increases risk of developing OSA up to 4.5 years post-infection and how risk varies by hospitalization status, demographics, comorbidities, and vaccination status. MethodsThis retrospective cohort study used electronic health records from the Montefiore Health System in the Bronx. Adults tested for SARS-CoV-2 between March 1, 2020, and August 17, 2024, were classified as hospitalized COVID+, non-hospitalized COVID+, or COVID-. Patients with prior OSA or inadequate follow-up were excluded. Inverse probability weighting adjusted for demographic, clinical, socioeconomic, and vaccination covariates. New-onset OSA was assessed using weighted Cox proportional hazards models. Secondary outcomes including hypertension, myocardial infarction, heart failure, stroke, arrhythmia, pulmonary hypertension, type 2 diabetes, and obesity were evaluated with Poisson regression. Sensitivity analysis used a pre-pandemic control cohort. ResultsAmong 910,393 eligible patients, hospitalized [HR 1.41 (95% CI 1.14-1.73)] and non-hospitalized [HR 1.33 (95% CI 1.22-1.46)] COVID+ patients had higher adjusted risk of new-onset OSA versus COVID- controls. Similar findings were observed using historical controls (n=621046). After OSA onset, hospitalized COVID+ patients had higher risks of heart failure and pulmonary hypertension, while non-hospitalized COVID+ patients had higher risk of obesity vs COVID- patients. ConclusionsSARS-CoV-2 infection is independently associated with increased risk of new-onset OSA. These findings support targeted screening in post-COVID populations.

BackgroundObstructive sleep apnea (OSA), as measured by the Apnea Hypopnea Index (AHI), is associated with adverse outcomes. Measures that characterize the temporal variability in events may provide information over and beyond a simple summary of event frequency as measured by the AHI. Research QuestionTo assess whether temporal variability in the occurrence of obstructive apnea/hypopneas during the night is associated with all-cause mortality or incident cardiovascular disease (CVD). Study Design and MethodsData from the Sleep Heart Health Study (SHHS), a prospective multi-site community-based cohort were analyzed. For each person, the intervals between apnea/hypopnea events (inter-event interval; IEI) were used to calculate a coefficient of variation for their IEIs (IEI_CV). Risk for mortality (n=5,701) and incident CVD (n=4,373) were estimated by adjusted Cox proportional hazard models. Sensitivity analyses were conducted to test potential explanatory variables such as hypoxic burden and duration of uninterrupted sleep. ResultsIn 11.8 years of follow-up (median, IQR 10.6-12.2), 1,287 deaths occurred. After adjusting for potential confounders, including OSA severity, participants in the lowest quartile of IEI_CV (Q1) had a 40% higher risk of all-cause mortality compared with those in the highest quartile (Q4) (hazard ratio [HR] = 1.40; 95% confidence interval [CI], 1.20-1.64). In 11.5 years of follow-up (IQR 7.9-12.7), 867 CVD events occurred. The adjusted hazard rate for CVD was 29% higher (HR=1.29 [1.06-1.56]) for those with less variable IEI. Minimal reductions in effects sizes were observed after additional adjustment for hypoxic burden and additional novel and traditional covariates. In sensitivity analyses, adjusting for the longest bout of uninterrupted sleep without respiratory events attenuated the association for CVD incidence (HR=1.15 [0.89-1.50]). InterpretationThe temporal distribution of respiratory events - specifically, less variability in inter-event intervals (more regular event occurrences) - is associated with higher mortality and incident CVD.

BackgroundAs sleep data from wearable devices are increasingly available in health research, there are new opportunities to understand sleep regulation behaviors as modifiable risk factors for disease. At such a large scale (tens of thousands of people over millions of day-level observations), prioritizing and interpreting sleep behaviors is challenging while maintaining biological relevance and modifiability. In this work, we aim to address this challenge by proposing a framework to interpret Fitbit data through a well-known neurobiological framing of sleep regulation, the two-process model. MethodsWe use data from the All of Us Research Program, a national biobank with passively collected Fitbit data for 32,292 people across 15,754,893 total days. We map Fitbit behaviors (b) to either circadian (C) or homeostatic (S) processes. Using iterative exploratory factor analysis to obtain weights, the Fitbit Cb and Sb are then weighted at the level of each day to create Cb and Sb scores. FindingsCb and Sb scores were found to align with expected real-world relationships with age, seasonality, shift work, and napping. Cb and Sb scores were interpreted with relation to depression, where it was found that Sb scores are highly associated with likelihood of diagnosis (OR = 1.5, p < 2e-16) while Cb and Sb scores are equally associated with severity (Sb score {beta} = 0.2, Cb score {beta} = 0.21, p < 2e-16). InterpretationCb and Sb scores support longitudinal interpretation (e.g., changes in Sb around treatment), aggregation (e.g., differences in Cb between two groups), and actionable modification (e.g., reduce naps to improve poor Sb). Overall, our behavior scores allow for interpretation of wearables sleep data and can be utilized across many disease contexts to better understand how sleep influences health. FundingThis work was supported by NIH training grant T32GM145734 and NIH R21HL172038.

Background and ObjectivesSleep has been associated with cognition and risk of dementia. However, sleep is a highly complex and multi-dimensional state, and there is uncertainty about which aspects of sleep are most relevant to cognitive performance and dementia risk. We applied a data-driven approach to identify clusters of sleep variables that reflect meaningful sleep composites and examined their association with cognitive performance and dementia risk. MethodsData from the Sleep and Dementia Consortium, consisting of 5 US population-based cohorts were utilized. Participants had methodologically consistent, home-based polysomnography, self-report habitual sleep, neuropsychological assessments, and dementia risk surveillance. The pooled cognitive analysis included 5,958 participants aged [&ge;]45 years, and the incident dementia analysis included 5,471 participants aged [&ge;]60 years. A cluster around latent variables analysis was used to derive 9 latent sleep composites from 44 sleep metrics. Global cognitive composite z-scores were derived from principal component analysis. Linear regression models were used to assess associations between sleep composites and cognitive performance. Cox proportional hazard models assessed associations between sleep composites and incident dementia. ResultsMean (SD) age was 70 {+/-} 11 and 74 {+/-} 12 years for the cognitive and dementia analysis, respectively. There were 1,134 incident dementia cases (median follow-up time of 5-19 years). 9 sleep composites were identified, together explaining 49% of the total variance in the original 44 sleep metrics: Sleep quantity and efficiency, sleep fragmentation, light NREM predominance, N3 predominance, spindle number and duration, REM sleep bouts, respiratory disturbances, slow oscillation-spindle coupling and spindle amplitude. Of these, composites reflecting greater sleep quantity and efficiency (i.e., longer and more consolidated sleep; pooled {beta} per one-unit change in composite, 0.03; 95% CI: 0.004 - 0.06; p=0.033) and stronger slow oscillation-spindle coupling (pooled {beta}, 0.04; 95% CI: 0.003 - 0.07; p=0.039) were associated with better global cognition. However, no significant associations were identified between the 9 sleep composites and dementia risk. DiscussionOur data-driven approach identified longer, more consolidated sleep and stronger slow oscillation-spindle coupling as the composites of sleep most strongly related to cognitive performance. These composites may be useful in guiding further investigations of sleep-brain health relationships.

ObjectiveTo explore how younger and older adults in the UK experience, manage, and seek treatment for long COVID-related cognitive impairments, and to identify their treatment preferences and expectations. DesignQualitative study using Charmazs (2014) constructivist grounded theory methodology. SettingOnline, semi-structured interviews conducted with UK adults reporting long COVID-related cognitive symptoms, recruited via patient support groups and research panels. ParticipantsTwenty-one adults with long COVID cognitive impairments: 10 younger participants aged 25-38 years and 11 older participants aged 60-72 years. Main outcome measuresPatients lived experiences, coping strategies, healthcare interactions, and treatment preferences and expectations for non-pharmacological and pharmacological interventions, analysed through iterative coding, constant comparison, and theory generation. ResultsIn the absence of effective treatments, both age groups relied heavily on self-management strategies (e.g., memory aids, structured routines, pacing). Interactions with healthcare were characterised by validation from some clinicians but widespread frustration at the lack of treatment options. Younger adults were more proactive in seeking experimental therapies and clinical trials, while older adults emphasised pragmatic adaptation, independence, and cautious optimism. Across groups, participants preferred non-pharmacological interventions (e.g., cognitive rehabilitation) but also expressed hope for biomedical treatments. The central process identified was "striving for agency" in the face of ongoing cognitive difficulties. ConclusionsThis study highlights the urgent unmet need for evidence-based interventions to address long COVID-related cognitive impairment. Health services should provide practical cognitive rehabilitation and support, and clinicians should acknowledge and validate patients cognitive struggles rather than dismissing them as normal ageing or purely psychogenic in origin. Research into therapeutics (e.g. cognitive training programs, pharmacotherapies) is urgently desired by the participants in this study. Effective solutions will need to be holistic and individualised - addressing not only memory and concentration deficits but also the psychological and social challenges associated with long COVID cognitive impairment in different age groups. WHAT IS ALREADY KNOWN ON THIS TOPICO_LIA large proportion of people with long COVID experience persistent cognitive difficulties, often described as "brain fog," with substantial impact on daily functioning and employment. C_LIO_LIMost studies have focused on symptom prevalence, biological mechanisms, or broad psychosocial consequences. C_LIO_LILittle is known about patients treatment preferences and expectations, and almost no research has explored generational differences between younger and older adults. C_LI WHAT THIS STUDY ADDSO_LITo our knowledge, this is the first qualitative study in the UK to compare younger and older adults perspectives on managing long COVID-related cognitive impairment. C_LIO_LIBoth groups described "striving for agency" through self-management, validation-seeking in healthcare, and balancing preferences for pharmacological and non-pharmacological treatments. C_LIO_LIYounger adults were more proactive in seeking experimental therapies and trial participation, while older adults emphasised pragmatic adaptation and maintaining independence. C_LI HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE, OR POLICYO_LIHealthcare services should expand access to cognitive rehabilitation and psychosocial support, while clinicians should validate patients experiences rather than attributing them to ageing or stress. C_LIO_LIFuture research must test both pharmacological and non-pharmacological interventions, with study designs informed by patient priorities and generational needs. C_LIO_LIPolicymakers and service providers should tailor care pathways to life stage: supporting younger adults with workplace accommodations and research opportunities, while providing older adults with reassurance, independence-focused care, and monitoring for possible accelerated cognitive ageing. C_LI

ImportanceCognitive impairment (CI) is often under detected in primary care due to time and resource constraints. Passive analysis of clinical dialogue may offer an accessible approach for screening. ObjectiveTo assess whether audio recordings of patient-physician dialogue during routine primary care visits can be used to identify CI using acoustic speech features and machine learning (ML). DesignThis observational study conducted among older primary care patients involved audio recording primary care visits using a microphone and portable device. An external validation cohort was recruited in a separate city to assess reproducibility of findings. SettingThe study was conducted in primary care practices in New York City, with additional participants recruited from primary care practices in Chicago, Illinois, for validation. ParticipantsThe study included 787 English-speaking patients aged 55 years and older, without documented history of dementia or mild CI. Eligible patients were recruited from primary care practices during routine visits. For validation, 179 patients meeting the same eligibility criteria were recruited from primary care practices in Chicago. ExposuresMultiple thirty-second speech segments were extracted from recordings. Acoustic features were derived using foundation models (Whisper, HuBERT, Wav2Vec 2.0) and expert-defined methods (eGeMAPS, prosody). Main Outcomes and MeasuresCI was defined as Montreal Cognitive Assessment score [&ge;]1.0 standard deviations below age and education-adjusted norms. ML classifiers were trained to predict CI status from audio recordings. We calculated area under the receiver operating characteristic curve (AUC-ROC) and maximum F1 score (Fmax) for identifying CI participants. ResultsThe mean age was 66.8 years and 21% had CI. Models using Whisper-derived acoustic features performed best (AUC-ROC=0.733, 95% confidence interval [95%CI]=0.714-0.752; Fmax(CI)=0.504, 95%CI=0.474-0.534). Results generalized to the external site with similar performance (AUC-ROC=0.727, 95%CI=0.714-0.740; Fmax(CI)=0.459, 95%CI=0.442-0.476). Model interpretation identified pitch, timing, and variability features as key predictors. When used for screening, the algorithm achieved positive predictive value of 30.4% (95%CI=28.7%-32.1%), sensitivity of 68.2% (95%CI=61.8%-74.6%), and specificity of 63.6% (95%CI=59.8%-67.4%) on the holdout cohort. Conclusions and RelevanceML models trained on acoustic features from brief clinical conversations identified CI with high accuracy. These findings support the feasibility of passive, speech-based screening during routine primary care. Key Points QuestionCan acoustic features extracted from audio recordings of patient-physician conversations during routine primary care visits be used to screen for cognitive impairment? FindingsIn this study including 787 older adults without diagnosis of cognitive problems, machine learning models trained on acoustic features from audio segments of recordings of primary care visits achieved area under the receiver operating characteristic curve values of 0.72 for predicting cognitive impairment. The algorithm achieved a sensitivity of 83%, specificity of 44%, and positive predictive value of 28%, identifying a subset of primary care patients at higher risk for cognitive impairment. Models performed similarly on an external validation dataset of 179 participants. Interpretability analyses highlighted patient pause duration and energy-related features as salient indicators of cognition status. MeaningThese findings suggest that short segments of naturalistic clinical dialogue may contain useful acoustic signals for passively screening patients for cognitive impairment.