Prediction of shiftworker alertness, sleep, and circadian phase using a model of arousal dynamics constrained by shift schedules and light exposure

SleepSync: Early Testing of a Personalised Sleep-Wake Management Smartphone Application for Improving Sleep and Cognitive Fitness in Defence Shift Workers

Shift work, long work hours, and operational tasks contribute to sleep and circadian disruption in defence personnel, with profound impacts on cognition. To address this, a digital technology, the SleepSync app, was designed for use in defence. A pre-post design study was undertaken to examine whether four weeks app use improved sleep and cognitive fitness (high performance neurocognition) in a cohort of shift workers from the Royal Australian Air Force. In total, 13 of approximately 20 shift-working personnel from one base volunteered for the study. Sleep outcomes were assessed using the Insomnia Severity Index (ISI), the Patient-Reported Outcomes Measurement Information System (PROMIS), Sleep Disturbance and Sleep-Related Impairment Scales, the Glasgow Sleep Effort Scale, the Sleep Hygiene Index, and mental health was assessed using the Depression, Anxiety, and Stress Scale-21. Sustained attention was measured using the 3-min Psychomotor Vigilance Task (PVT) and controlled response using the NBack. Results showed significant improvements in insomnia (ISI scores 10.31 at baseline and 7.50 after app use), sleep-related impairments (SRI T-scores 53.03 at baseline to 46.75 post-app use), and healthy sleep practices (SHI scores 21.61 at baseline to 18.83 post-app use; all p < 0.001). Trends for improvement were recorded for depression. NBack incorrect responses reduced significantly (9.36 at baseline; reduced by -3.87 at last week of app use, p < 0.001), but no other objective measures improved. These findings suggest that SleepSync may improve sleep and positively enhance cognitive fitness but warrants further investigation in large samples. Randomised control trials with other cohorts of defence personnel are needed to confirm the utility of this intervention in defence settings.

Causal dynamics of sleep, circadian rhythm, and mood symptoms in patients with major depression and bipolar disorder: insights from longitudinal wearable device data

BACKGROUND

Sleep and circadian rhythm disruptions are common in patients with mood disorders. The intricate relationship between these disruptions and mood has been investigated, but their causal dynamics remain unknown.

METHODS

We analysed data from 139 patients (76 female, mean age = 23.5 ± 3.64 years) with mood disorders who participated in a prospective observational study in South Korea. The patients wore wearable devices to monitor sleep and engaged in smartphone-delivered ecological momentary assessment of mood symptoms. Using a mathematical model, we estimated their daily circadian phase based on sleep data. Subsequently, we obtained daily time series for sleep/circadian phase estimates and mood symptoms spanning >40,000 days. We analysed the causal relationship between the time series using transfer entropy, a non-linear causal inference method.

FINDINGS

The transfer entropy analysis suggested causality from circadian phase disturbance to mood symptoms in both patients with MDD (n = 45) and BD type I (n = 35), as 66.7% and 85.7% of the patients with a large dataset (>600 days) showed causality, but not in patients with BD type II (n = 59). Surprisingly, no causal relationship was suggested between sleep phase disturbances and mood symptoms.

INTERPRETATION

Our findings suggest that in patients with mood disorders, circadian phase disturbances directly precede mood symptoms. This underscores the potential of targeting circadian rhythms in digital medicine, such as sleep or light exposure interventions, to restore circadian phase and thereby manage mood disorders effectively.

FUNDING

Institute for Basic Science, the Human Frontiers Science Program Organization, the National Research Foundation of Korea, and the Ministry of Health & Welfare of South Korea.

Imputing missing sleep data from wearables with neural networks in real-world settings

Sleep is a critical component of health and well-being but collecting and analyzing accurate longitudinal sleep data can be challenging, especially outside of laboratory settings. We propose a simple neural network model titled SOMNI (Sleep data restOration using Machine learning and Non-negative matrix factorIzation [NMF]) for imputing missing rest-activity data from actigraphy, which can enable clinicians to better handle missing data and monitor sleep-wake cycles of individuals with highly irregular sleep-wake patterns. The model consists of two hidden layers and uses NMF to capture hidden longitudinal sleep-wake patterns of individuals with disturbed sleep-wake cycles. Based on this, we develop two approaches: the individual approach imputes missing data based on the data from only one participant, while the global approach imputes missing data based on the data across multiple participants. Our models are tested with shift and non-shift workers' data from three independent hospitals. Both approaches can accurately impute missing data up to 24 hours of long dataset (>50 days) even for shift workers with extremely irregular sleep-wake patterns (AUC > 0.86). On the other hand, for short dataset (~15 days), only the global model is accurate (AUC > 0.77). Our approach can be used to help clinicians monitor sleep-wake cycles of patients with sleep disorders outside of laboratory settings without relying on sleep diaries, ultimately improving sleep health outcomes.

Associations between light exposure and sleep timing and sleepiness while awake in a sample of UK adults in everyday life

Experimental and interventional studies show that light can regulate sleep timing and sleepiness while awake by setting the phase of circadian rhythms and supporting alertness. The extent to which differences in light exposure explain variations in sleep and sleepiness within and between individuals in everyday life remains less clear. Here, we establish a method to address this deficit, incorporating an open-source wearable wrist-worn light logger (SpectraWear) and smartphone-based online data collection. We use it to simultaneously record longitudinal light exposure (in melanopic equivalent daylight illuminance), sleep timing, and subjective alertness over seven days in a convenience sample of 59 UK adults without externally imposed circadian challenge (e.g., shift work or jetlag). Participants reliably had strong daily rhythms in light exposure but frequently were exposed to less light during the daytime and more light in pre-bedtime and sleep episodes than recommended [T. M. Brown et al., PLoS Biol. 20, e3001571 (2022)]. Prior light exposure over several hours was associated with lower subjective sleepiness with, in particular, brighter light in the late sleep episode and after wake linked to reduced early morning sleepiness (sleep inertia). Higher pre-bedtime light exposure was associated with longer sleep onset latency. Early sleep timing was correlated with more reproducible and robust daily patterns of light exposure and higher daytime/lower night-time light exposure. Our study establishes a method for collecting longitudinal sleep and health/performance data in everyday life and provides evidence of associations between light exposure and important determinants of sleep health and performance.

A real-time, personalized sleep intervention using mathematical modeling and wearable devices

The prevalence of artificial light exposure has enabled us to be active any time of the day or night, leading to the need for high alertness outside of traditional daytime hours. To address this need, we developed a personalized sleep intervention framework that analyzes real-world sleep-wake patterns obtained from wearable devices to maximize alertness during specific target periods. Our framework utilizes a mathematical model that tracks the dynamic sleep pressure and circadian rhythm based on the user's sleep history. In this way, the model accurately predicts real-time alertness, even for shift workers with complex sleep and work schedules (N = 71, t = 13~21 days). This allowed us to discover a new sleep-wake pattern called the adaptive circadian split sleep, which incorporates a main sleep period and a late nap to enable high alertness during both work and non-work periods of shift workers. We further developed a mobile application that integrates this framework to recommend practical, personalized sleep schedules for individual users to maximize their alertness during a targeted activity time based on their desired sleep onset and available sleep duration. This can reduce the risk of errors for those who require high alertness during nontraditional activity times and improve the health and quality of life for those leading shift work-like lifestyles.

Pilot feasibility testing of biomathematical model recommendations for personalising sleep timing in shift workers

Sleep disturbances and circadian disruption play a central role in adverse health, safety, and performance outcomes in shift workers. While biomathematical models of sleep and alertness can be used to personalise interventions for shift workers, their practical implementation is undertested. This study tested the feasibility of implementing two biomathematical models-the Phillips-Robinson Model and the Model for Arousal Dynamics-in 28 shift-working nurses, 14 in each group. The study examined the overlap and adherence between model recommendations and sleep behaviours, and changes in sleep following the implementation of recommendations. For both groups combined, the mean (SD) percentage overlap between when a model recommended an individual to sleep and when sleep was obtained was 73.62% (10.24%). Adherence between model recommendations and sleep onset and offset times was significantly higher with the Model of Arousal Dynamics compared to the Phillips-Robinson Model. For the Phillips-Robinson model, 27% of sleep onset and 35% of sleep offset times were within ± 30 min of model recommendations. For the Model of Arousal Dynamics, 49% of sleep onset, and 35% of sleep offset times were within ± 30 min of model recommendations. Compared to pre-study, significant improvements were observed post-study for sleep disturbance (Phillips-Robinson Model), and insomnia severity and sleep-related impairments (Model of Arousal Dynamics). Participants reported that using a digital, automated format for the delivery of sleep recommendations would enable greater uptake. These findings provide a positive proof-of-concept for using biomathematical models to recommend sleep in operational contexts.

Cortical waste clearance in normal and restricted sleep with potential runaway tau buildup in Alzheimer's disease

Accumulation of waste in cortical tissue and glymphatic waste clearance via extracellular voids partly drives the sleep-wake cycle and modeling has reproduced much of its dynamics. Here, new modeling incorporates higher void volume and clearance in sleep, multiple waste compounds, and clearance obstruction by waste. This model reproduces normal sleep-wake cycles, sleep deprivation effects, and performance decreases under chronic sleep restriction (CSR). Once fitted to calibration data, it successfully predicts dynamics in further experiments on sleep deprivation, intermittent CSR, and recovery after restricted sleep. The results imply a central role for waste products with lifetimes similar to tau protein. Strong tau buildup is predicted if pathologically enhanced production or impaired clearance occur, with runaway buildup above a critical threshold. Predicted tau accumulation has timescales consistent with the development of Alzheimer’s disease. The model unifies a wide sweep of phenomena, clarifying the role of glymphatic clearance and targets for interventions against waste buildup.

Working around the Clock: Is a Person’s Endogenous Circadian Timing for Optimal Neurobehavioral Functioning Inherently Task-Dependent?

Neurobehavioral task performance is modulated by the circadian and homeostatic processes of sleep/wake regulation. Biomathematical modeling of the temporal dynamics of these processes and their interaction allows for prospective prediction of performance impairment in shift-workers and provides a basis for fatigue risk management in 24/7 operations. It has been reported, however, that the impact of the circadian rhythm—and in particular its timing—is inherently task-dependent, which would have profound implications for our understanding of the temporal dynamics of neurobehavioral functioning and the accuracy of biomathematical model predictions. We investigated this issue in a laboratory study designed to unambiguously dissociate the influences of the circadian and homeostatic processes on neurobehavioral performance, as measured during a constant routine protocol preceded by three days on either a simulated night shift or a simulated day shift schedule. Neurobehavioral functions were measured every 3 h using three functionally distinct assays: a digit symbol substitution test, a psychomotor vigilance test, and the Karolinska Sleepiness Scale. After dissociating the circadian and homeostatic influences and accounting for inter-individual variability, peak circadian performance occurred in the late biological afternoon (in the “wake maintenance zone”) for all three neurobehavioral assays. Our results are incongruent with the idea of inherent task-dependent differences in the endogenous circadian impact on performance. Rather, our results suggest that neurobehavioral functions are under top-down circadian control, consistent with the way they are accounted for in extant biomathematical models.