Explore the vast range of titles available.

Metabolic rate reduction has been considered the mechanism by which sleep conserves energy, similar to torpor or hibernation. This mechanism of energy savings is in conflict with the known upregulation (compared to wake) of diverse functions during sleep and neglects a potential role in energy conservation for partitioning of biological operations by behavioral state. Indeed, energy savings as derived from state-dependent resource allocations have yet to be examined. A mathematical model is presented based on relative rates of energy deployment for biological processes upregulated during either wake or sleep. Using this model, energy savings from sleep-wake cycling over constant wakefulness is computed by comparing stable limit cycles for systems of differential equations. A primary objective is to compare potential energy savings derived from state-dependent metabolic partitioning versus metabolic rate reduction. Additionally, energy conservation from sleep quota and the circadian system are also quantified in relation to a continuous wake condition. As a function of metabolic partitioning, our calculations show that coupling of metabolic operations with behavioral state may provide comparatively greater energy savings than the measured decrease in metabolic rate, suggesting that actual energy savings derived from sleep may be more than 4-fold greater than previous estimates. A combination of state-dependent metabolic partitioning and modest metabolic rate reduction during sleep may enhance energy savings beyond what is achievable through metabolic partitioning alone; however, the relative contribution from metabolic partitioning diminishes as metabolic rate is decreased during the rest phase. Sleep quota and the circadian system further augment energy savings in the model. Finally, we propose that state-dependent resource allocation underpins both sleep homeostasis and the optimization of daily energy conservation across species. This new paradigm identifies an evolutionary selective advantage for the upregulation of central and peripheral biological processes during sleep, presenting a unifying construct to understand sleep function.

AASM guidelines are the result of decades of efforts aiming at standardizing sleep scoring procedure, with the final goal of sharing a worldwide common methodology. The guidelines cover several aspects from the technical/digital specifications, e.g., recommended EEG derivations, to detailed sleep scoring rules accordingly to age. Automated sleep scoring systems have always largely exploited the standards as fundamental guidelines. In this context, deep learning has demonstrated better performance compared to classical machine learning. Our present work shows that a deep learning-based sleep scoring algorithm may not need to fully exploit the clinical knowledge or to strictly adhere to the AASM guidelines. Specifically, we demonstrate that U-Sleep, a state-of-the-art sleep scoring algorithm, can be strong enough to solve the scoring task even using clinically non-recommended or non-conventional derivations, and with no need to exploit information about the chronological age of the subjects. We finally strengthen a well-known finding that using data from multiple data centers always results in a better performing model compared with training on a single cohort. Indeed, we show that this latter statement is still valid even by increasing the size and the heterogeneity of the single data cohort. In all our experiments we used 28528 polysomnography studies from 13 different clinical studies.

Automatic sleep staging with deep learning has advanced considerably, yet clinical adoption remains hindered by limited generalization, model bias, and inconsistent evaluation practices. We present SLEEPYLAND, an open-source framework comprising ~ 220,000 h of in-domain and ~ 84,000 h of out-of-domain polysomnographic recordings, spanning diverse ages, disorders, and hardware configurations. We release pre-trained state-of-the-art models, evaluating them across single- and multi-channel EEG/EOG setups. We introduce SOMNUS, an ensemble that integrates models via soft-voting, achieving robust performance across 24 datasets (macro-F1, 68.7–87.2%), outperforming individual models in 94.9% of cases and exceeding prior state-of-the-art. Exploiting the Bern-Sleep-Wake-Registry ( N  = 6633), we show that while SOMNUS improves generalization, no model architecture consistently minimizes model demographic/clinical bias. On multi-annotated datasets, SOMNUS surpasses the best human scorer (macro-F1, 85.2% vs 80.8% on DOD-H, and 80.2% vs 75.9% on DOD-O), more closely reproducing consensus. Finally, ensemble disagreement metrics predict scorer ambiguity (ROC-AUC 82.8%), providing reliable proxies for human uncertainty.

Sleep apnoea, one of the most common chronic diseases, is a risk factor for ischaemic stroke, stroke recurrence, and poor functional recovery after stroke. More than half of stroke survivors present with sleep apnoea during the acute phase after stroke, with obstructive sleep apnoea being the most common subtype. Following a stroke, sleep apnoea frequency and severity might decrease over time, but moderate to severe sleep apnoea is nevertheless present in up to a third of patients in the chronic phase after an ischaemic stroke. Over the past few decades evidence suggests that treatment for sleep apnoea is feasible during the acute phase of stroke and might favourably affect recovery and long-term outcomes. Nevertheless, sleep apnoea still remains underdiagnosed and untreated in many cases, due to challenges in the detection and prediction of post-stroke sleep apnoea, uncertainty as to the optimal timing for its diagnosis, and a scarcity of clear treatment guidelines (ie, uncertainty on when to treat and the optimal treatment strategy). Moreover, the pathophysiology of sleep apnoea associated with stroke, the proportion of stroke survivors with obstructive and central sleep apnoea, and the temporal evolution of sleep apnoea subtypes following stroke remain to be clarified. To address these shortcomings, the management of sleep apnoea associated with stroke should be integrated into a multidisciplinary diagnostic, treatment, and follow-up strategy.

Abstract Three papers currently published in SLEEP using two different mouse models of narcolepsy, including either Hcrt-tTa;TetO diptheria toxin-A (DTA) or Hypocretin knock-out (Hcrt-KO) mice, suggest important gender differences in narcolepsy expression. Specifically, these recent data corroborate previous findings in mice demonstrating that females show more cataplexy events and more total cataplexy expression than males. Moreover, in the neurotoxic DTA mouse model, females show earlier onset of cataplexy expression than males during active Hcrt cell loss. Finally, females show a doubling of cataplexy during estrous compared to other phases of the estrous cycle. These findings are reviewed in the broader context of prior published literature, including reported gender differences in Hcrt expression and hormonal influences on sleep and wakefulness. Although similar findings have not been reported in humans, a systematic evaluation of gender differences in human narcolepsy has yet to be performed. Taken together, these animal data suggest that more research exploring gender differences in human narcolepsy is warranted.

Orexins (also called hypocretins) are hypothalamic neuropeptides that carry out essential functions in the central nervous system; however, little is known about their release and range of action in vivo owing to the limited resolution of current detection technologies. Here we developed a genetically encoded orexin sensor (OxLight1) based on the engineering of circularly permutated green fluorescent protein into the human type-2 orexin receptor. In mice OxLight1 detects optogenetically evoked release of endogenous orexins in vivo with high sensitivity. Photometry recordings of OxLight1 in mice show rapid orexin release associated with spontaneous running behavior, acute stress and sleep-to-wake transitions in different brain areas. Moreover, two-photon imaging of OxLight1 reveals orexin release in layer 2/3 of the mouse somatosensory cortex during emergence from anesthesia. Thus, OxLight1 enables sensitive and direct optical detection of orexin neuropeptides with high spatiotemporal resolution in living animals. OxLight1 is a genetically encoded sensor for the orexin neuropeptides. It has been applied in fiber photometry recordings and two-photon imaging in mice during a variety of behaviors.

Abstract The Swiss Primary Hypersomnolence and Narcolepsy Cohort Study (SPHYNCS) is a multicenter research initiative to identify new biomarkers in central disorders of hypersomnolence (CDH). Whereas narcolepsy type 1 (NT1) is well characterized, other CDH disorders lack precise biomarkers. In SPHYNCS, we utilized Fitbit smartwatches to monitor physical activity, heart rate, and sleep parameters over 1 year. We examined the feasibility of long-term ambulatory monitoring using the wearable device. We then explored digital biomarkers differentiating patients with NT1 from healthy controls (HC). A total of 115 participants received a Fitbit smartwatch. Using a adherence metric to evaluate the usability of the wearable device, we found an overall adherence rate of 80% over 1 year. We calculated daily physical activity, heart rate, and sleep parameters from 2 weeks of greatest adherence to compare NT1 (n = 20) and HC (n = 9) participants. Compared to controls, NT1 patients demonstrated findings consistent with increased sleep fragmentation, including significantly greater wake-after-sleep onset (p = .007) and awakening index (p = .025), as well as standard deviation of time in bed (p = .044). Moreover, NT1 patients exhibited a significantly shorter REM latency (p = .019), and sleep latency (p = .001), as well as a lower peak heart rate (p = .008), heart rate standard deviation (p = .039) and high-intensity activity (p = .009) compared to HC. This ongoing study demonstrates the feasibility of long-term monitoring with wearable technology in patients with CDH and potentially identifies a digital biomarker profile for NT1. While further validation is needed in larger datasets, these data suggest that long-term wearable technology may play a future role in diagnosing and managing narcolepsy. Graphical Abstract Graphical Abstract

This study aims to enhance the clinical use of automated sleep-scoring algorithms by incorporating an uncertainty estimation approach to efficiently assist clinicians in the manual review of predicted hypnograms, a necessity due to the notable inter-scorer variability inherent in polysomnography (PSG) databases. Our efforts target the extent of review required to achieve predefined agreement levels, examining both in-domain (ID) and out-of-domain (OOD) data, and considering subjects' diagnoses. A total of 19,578 PSGs from 13 open-access databases were used to train U-Sleep, a state-of-the-art sleep-scoring algorithm. We leveraged a comprehensive clinical database of an additional 8832 PSGs, covering a full spectrum of ages (0-91 years) and sleep-disorders, to refine the U-Sleep, and to evaluate different uncertainty-quantification approaches, including our novel confidence network. The ID data consisted of PSGs scored by over 50 physicians, and the two OOD sets comprised recordings each scored by a unique senior physician. U-Sleep demonstrated robust performance, with Cohen's kappa (K) at 76.2% on ID and 73.8-78.8% on OOD data. The confidence network excelled at identifying uncertain predictions, achieving AUROC scores of 85.7% on ID and 82.5-85.6% on OOD data. Independently of sleep-disorder status, statistical evaluations revealed significant differences in confidence scores between aligning vs discording predictions, and significant correlations of confidence scores with classification performance metrics. To achieve κ ≥ 90% with physician intervention, examining less than 29.0% of uncertain epochs was required, substantially reducing physicians' workload, and facilitating near-perfect agreement. Inter-scorer variability limits the accuracy of the scoring algorithms to ~80%. By integrating an uncertainty estimation with U-Sleep, we enhance the review of predicted hypnograms, to align with the scoring taste of a responsible physician. Validated across ID and OOD data and various sleep-disorders, our approach offers a strategy to boost automated scoring tools' usability in clinical settings.