Comparison of Wearable Activity Tracker with Actigraphy for Sleep Evaluation and Circadian Rest-Activity Rhythm Measurement in Healthy Young Adults

Machine learning-based prediction of restless legs syndrome using digital phenotypes from wearables and smartphone data

Restless legs syndrome (RLS) is a relatively common neurosensory disorder that causes an irresistible urge for leg movement. RLS causes sleep disturbances and reduced quality of life, but accurate diagnosis remains challenging owing to the reliance on subjective reporting. This study aimed to propose a predictive machine learning model based on digital phenotypes for RLS diagnosis. Self-reported lifestyle data were integrated via a smartphone application with objective biometric data from wearable devices to obtain 85 features processed based on circadian rhythms. Prediction models used these features to distinguish between the non-RLS (International Restless Legs Study Group Severity Rating Scale [IRLS] score ≤ 10) and RLS symptom groups (10 < IRLS ≤ 20) and between the non-RLS and severe RLS symptom groups (IRLS > 20). The RF model showed the highest performance in predicting the RLS symptom group and XGB model in the severe RLS symptom group. For the RLS symptom group, when using only wearable device data, the AUC, accuracy, precision, recall, and F1 scores were 0.78, 0.70, 0.66, 0.84, and 0.74, respectively, while these scores combining wearable device and application data were 0.86, 0.76, 0.68, 1.00, and 0.81, respectively. For the severe RLS symptom group, when using only wearable device data, XGB achieved AUC, accuracy, precision, recall, and F1 scores of 0.66, 0.84, 0.89, 0.93, and 0.91, respectively, while these scores combining wearable device and application data were 0.70, 0.80, 0.88, 0.90, and 0.89, respectively. Diverse digital phenotypes clinically associated with RLS were processed based on circadian rhythms to demonstrate the potential of digital phenotyping for RLS prediction. Thus, our study establishes early detection and personalized management of RLS.Trial Registration: Clinical Research Information Service (CRIS) KCT0009175 (Registration data: Feb-15-2024) ( https://cris.nih.go.kr/cris/search/detailSearch.do?search_lang=E&focus=reset_12&search_page=M&pageSize=10&page=undefined&seq=26133&status=5&seq_group=26133 ).

Psychophysiological stress response, emotion dysregulation and sleep parameters as predictors of psychopathology in adolescents and young adults

BACKGROUND

Increased emotional reactivity to stress, emotional dysregulation and sleep disturbances are interdependent trans-diagnostic processes that are present in internalising disorders such as depression and anxiety disorders. This study investigated which objective and subjective parameters of stress reactivity, sleep and emotional processing would predict symptoms of anxiety and depression in adolescents and young adults.

METHODS

Participants were adolescents and young adults between the ages of 14 to 21 (N = 106, 25[24 %] male, M age = 17.93). Heart rate, heart rate variability, and subjective stress levels were measured before, during and after a stress induction using the Trier Social Stress Test (TSST). Questionnaires on internalising symptoms, emotion dysregulation, and sleep quality were used. For seven consecutive nights, objective sleep parameters were measured with a wearable device.

RESULTS

Heart rate and heart rate variability after (but not during) the stress induction and emotion dysregulation predicted depressive and anxiety symptoms. Lower subjective sleep quality (but not the objective sleep parameters) was associated with depressive and anxiety symptoms. Emotion dysregulation mediated the relationship between sleep quality and depressive symptoms.

LIMITATIONS

A cross-sectional design, no measurement of daily activity or naps, and only self-report measures of depressive and anxiety symptoms as well as emotion dysregulation.

CONCLUSION

The findings of elevated cardiovascular activation after - but not during - the stress induction and emotion dysregulation underlines problems in regulating and recovering from stress as predictors of youth internalising psychopathology. Differences between subjective and objective measures of sleep and stress reactivity suggests a role of cognitive biases in these domains.

Digital phenotyping in bipolar disorder: Using longitudinal Fitbit data and personalized machine learning to predict mood symptomatology

BACKGROUND

Effective treatment of bipolar disorder (BD) requires prompt response to mood episodes. Preliminary studies suggest that predictions based on passive sensor data from personal digital devices can accurately detect mood episodes (e.g., between routine care appointments), but studies to date do not use methods designed for broad application. This study evaluated whether a novel, personalized machine learning approach, trained entirely on passive Fitbit data, with limited data filtering could accurately detect mood symptomatology in BD patients.

METHODS

We analyzed data from 54 adults with BD, who wore Fitbits and completed bi-weekly self-report measures for 9 months. We applied machine learning (ML) models to Fitbit data aggregated over two-week observation windows to detect occurrences of depressive and (hypo)manic symptomatology, which were defined as two-week windows with scores above established clinical cutoffs for the Patient Health Questionnaire-8 (PHQ-8) and Altman Self-Rating Mania Scale (ASRM) respectively.

RESULTS

As hypothesized, among several ML algorithms, Binary Mixed Model (BiMM) forest achieved the highest area under the receiver operating curve (ROC-AUC) in the validation process. In the testing set, the ROC-AUC was 86.0% for depression and 85.2% for (hypo)mania. Using optimized thresholds calculated with Youden's J statistic, predictive accuracy was 80.1% for depression (sensitivity of 71.2% and specificity of 85.6%) and 89.1% for (hypo)mania (sensitivity of 80.0% and specificity of 90.1%).

CONCLUSION

We achieved sound performance in detecting mood symptomatology in BD patients using methods designed for broad application. Findings expand upon evidence that Fitbit data can produce accurate mood symptomatology predictions. Additionally, to the best of our knowledge, this represents the first application of BiMM forest for mood symptomatology prediction. Overall, results move the field a step toward personalized algorithms suitable for the full population of patients, rather than only those with high compliance, access to specialized devices, or willingness to share invasive data.

The effects of COVID-19 lockdown and reopening on rest-activity rhythms in Singaporean working adults: A longitudinal age group comparison study

STUDY OBJECTIVES

COVID-19 mobility restrictions disrupted daily rhythms worldwide, but how this rhythm disruption differs across age groups is unclear. We examined the course of age-related differences in trajectories of rest-activity rhythm during the COVID-19 pandemic lockdown and reopening in Singapore. We also evaluated the association of these patterns with mental well-being.

METHODS

24-hour step count data (Fitbit) were obtained from 617 younger (age range: 21-40) and 602 older adults (age range: 55-70) from January 2020 (baseline) through lockdown (April 2020) and reopening periods until August 2021. Nonparametric rest-activity rhythm metrics: interdaily stability, intradaily variability and most active 10-hour period (M10) were computed. Longitudinal changes in rest-activity rhythm, age-related differences in changes, and the associations between mental well-being and these changes were assessed using nonlinear latent-growth models.

RESULTS

In younger adults, mobility restrictions during lockdown caused significant decline in interdaily stability and M10, alongside significant increase in intradaily variability. However, in older adults, changes were confined to increased intradaily variability and decreased M10. Older adults also showed less change in intradaily variability and M10 compared to younger adults. Gradual recovery of rest-activity rhythm metrics during reopening was observed, with interdaily stability and M10 remaining lower after 15months post-lockdown. In younger but not older adults, a larger decline in interdaily stability was associated with poorer mental well-being 15months post-lockdown.

CONCLUSION

Younger adults appear more vulnerable than older adults to mobility restrictions as reflected in their rest-activity rhythm metrics. A significant disruption of daily routine may have long-lasting effects on younger adults' mental well-being.

STATEMENT OF SIGNIFICANCE

Although stringent mobility restrictions imposed to curb the spread of COVID-19 were imposed primarily to protect older adults, we found that younger adults were more vulnerable to rhythm disruption arising from mobility restrictions. Disrupted rhythm stability was associated with poorer mental well-being 15months after the lockdown ended in younger but not older adults. These asymmetric long-term effects on mental health on younger relative to older adults should be kept in mind when planning for large-scale catastrophes linked to mobility restrictions.

Exploring the Psychological and Physiological Insights Through Digital Phenotyping by Analyzing the Discrepancies Between Subjective Insomnia Severity and Activity-Based Objective Sleep Measures: Observational Cohort Study

BACKGROUND

Insomnia is a prevalent sleep disorder affecting millions worldwide, with significant impacts on daily functioning and quality of life. While traditionally assessed through subjective measures such as the Insomnia Severity Index (ISI), the advent of wearable technology has enabled continuous, objective sleep monitoring in natural environments. However, the relationship between subjective insomnia severity and objective sleep parameters remains unclear.

OBJECTIVE

This study aims to (1) explore the relationship between subjective insomnia severity, as measured by ISI scores, and activity-based objective sleep parameters obtained through wearable devices; (2) determine whether subjective perceptions of insomnia align with objective measures of sleep; and (3) identify key psychological and physiological factors contributing to the severity of subjective insomnia complaints.

METHODS

A total of 250 participants, including both individuals with and without insomnia aged 19-70 years, were recruited from March 2023 to November 2023. Participants were grouped based on ISI scores: no insomnia, mild, moderate, and severe insomnia. Data collection involved subjective assessments through self-reported questionnaires and objective measurements using wearable devices (Fitbit Inspire 3) that monitored sleep parameters, physical activity, and heart rate. The participants also used a smartphone app for ecological momentary assessment, recording daily alcohol consumption, caffeine intake, exercise, and stress. Statistical analyses were used to compare groups on subjective and objective measures.

RESULTS

Results indicated no significant differences in general sleep structure (eg, total sleep time, rapid eye movement sleep time, and light sleep time) among the insomnia groups (mild, moderate, and severe) as classified by ISI scores (all P>.05). Interestingly, the no insomnia group had longer total awake times and lower sleep quality compared with the insomnia groups. Among the insomnia groups, no significant differences were observed regarding sleep structure (all P>.05), suggesting similar sleep patterns regardless of subjective insomnia severity. There were significant differences among the insomnia groups in stress levels, dysfunctional beliefs about sleep, and symptoms of restless leg syndrome (all P≤.001), with higher severity associated with higher scores in these factors. Contrary to expectations, no significant differences were observed in caffeine intake (P=.42) and alcohol consumption (P=.07) between the groups.

CONCLUSIONS

The findings demonstrate a discrepancy between subjective perceptions of insomnia severity and activity-based objective sleep parameters, suggesting that factors beyond sleep duration and quality may contribute to subjective sleep complaints. Psychological factors, such as stress, dysfunctional sleep beliefs, and symptoms of restless legs syndrome, appear to play significant roles in the perception of insomnia severity. These results highlight the importance of considering both subjective and objective assessments in the evaluation and treatment of insomnia and suggest potential avenues for personalized treatment strategies that address both psychological and physiological aspects of sleep disturbances.

TRIAL REGISTRATION

Clinical Research Information Service KCT0009175; https://cris.nih.go.kr/cris/search/detailSearch.do?seq=26133.

Utility of a Wearable Tracker to Assess Sleep Quality in Nurses and Their Spouses: A Prospective Cohort Study

INTRODUCTION

Sleep disturbances among nurses engaged in night duty and their spouses need to be improved to ensure their ability to provide care and perform daily tasks. Therefore, an objective investigation is needed to establish a sleep improvement strategy.

OBJECTIVE

To investigate the utility of a sleep tracker to assess sleep quality in nurses and spouses.

METHOD

Nurses (n = 30) and spouses (n = 30) wore a sleep tracker for 14 days to investigate sleep scores. Sleep quality and number of steps were evaluated by Fitbit. They responded to the Richards-Campbell Sleep Questionnaire and Pittsburgh Sleep Quality Index. A multiple regression analysis was performed to identify the factors affecting sleep quality.

RESULTS

Factors affecting sleep scores in nurses were hypnotic medication, night duty, and steps, while those in spouses were mental instability, hypnotic medication, alcohol, night duty, and steps. Factors affecting the Richards-Campbell Sleep Questionnaire in nurses were household chores, night duty, and steps, while those in spouses were hypnotic medication and steps.

CONCLUSION

The sleep quality of nurses was affected by household chores, hypnotic medication, night duty, and steps. Besides the factors of nurses, spouses were affected by mental instability and alcohol. Night duty affected negativity in both nurses and spouses. Steps exerted positive effects in both the sleep tracker and the Richards-Campbell Sleep Questionnaire. The sleep tracker may be useful for identifying factors that improve sleep quality.

Biometrically measured sleep in medical students as a predictor of psychological health and academic experiences in the preclinical years

BACKGROUND

Student wellness is of increasing concern in medical education. Increased rates of burnout, sleep disturbances, and psychological concerns in medical students are well documented. These concerns lead to impacts on current educational goals and may set students on a path for long-term health consequences.

METHODS

Undergraduate medical students were recruited to participate in a novel longitudinal wellness tracking project. This project utilized validated wellness surveys to assess emotional health, sleep health, and burnout at multiple timepoints. Biometric information was collected from participant Fitbit devices that tracked longitudinal sleep patterns.

RESULTS

Eighty-one students from three cohorts were assessed during the first semester of their M1 preclinical curriculum. Biometric data showed that nearly 30% of the students had frequent short sleep episodes (<6 hours of sleep for at least 30% of recorded days), and nearly 68% of students had at least one episode of three or more consecutive days of short sleep. Students that had consecutive short sleep episodes had higher rates of stress (8.3%) and depression (5.4%) symptoms and decreased academic efficiency (1.72%).

CONCLUSIONS

Biometric data were shown to significantly predict psychological health and academic experiences in medical students. Biometrically assessed sleep is poor in medical students, and consecutive days of short sleep duration are particularly impactful as it relates to other measures of wellness. Longitudinal, biometric data tracking is feasible and can provide students the ability to self-monitor health behaviors and allow for low-intensity health interventions.

Secure and privacy-preserving automated machine learning operations into end-to-end integrated IoT-edge-artificial intelligence-blockchain monitoring system for diabetes mellitus prediction

Diabetes Mellitus, one of the leading causes of death worldwide, has no cure to date and can lead to severe health complications, such as retinopathy, limb amputation, cardiovascular diseases, and neuronal disease, if left untreated. Consequently, it becomes crucial to be able to monitor and predict the incidence of diabetes. Machine learning approaches have been proposed and evaluated in the literature for diabetes prediction. This paper proposes an IoT-edge-Artificial Intelligence (AI)-blockchain system for diabetes prediction based on risk factors. The proposed system is underpinned by blockchain to obtain a cohesive view of the risk factors data from patients across different hospitals and ensure security and privacy of the user's data. We provide a comparative analysis of different medical sensors, devices, and methods to measure and collect the risk factors values in the system. Numerical experiments and comparative analysis were carried out within our proposed system, using the most accurate random forest (RF) model, and the two most used state-of-the-art machine learning approaches, Logistic Regression (LR) and Support Vector Machine (SVM), using three real-life diabetes datasets. The results show that the proposed system predicts diabetes using RF with 4.57% more accuracy on average in comparison with the other models LR and SVM, with 2.87 times more execution time. Data balancing without feature selection does not show significant improvement. When using feature selection, the performance is improved by 1.14% for PIMA Indian and 0.02% for Sylhet datasets, while it is reduced by 0.89% for MIMIC III.

Assessing sleep in primary brain tumor patients using smart wearables and patient-reported data: Feasibility and interim analysis of an observational study

Background

Sleep-wake disturbances are common and disabling in primary brain tumor (PBT) patients but studies exploring longitudinal data are limited. This study investigates the feasibility and relationship between longitudinal patient-reported outcomes (PROs) and physiologic data collected via smart wearables.

Methods

Fifty-four PBT patients ≥ 18 years wore Fitbit smart-wearable devices for 4 weeks, which captured physiologic sleep measures (eg, total sleep time, wake after sleep onset [WASO]). They completed PROs (sleep hygiene index, PROMIS sleep-related impairment [SRI] and Sleep Disturbance [SD], Morningness-Eveningness Questionnaire [MEQ]) at baseline and 4 weeks. Smart wearable use feasibility (enrollment/attrition, data missingness), clinical characteristics, test consistency, PROs severity, and relationships between PROs and physiologic sleep measures were assessed.

Results

The majority (72%) wore their Fitbit for the entire study duration with 89% missing < 3 days, no participant withdrawals, and 100% PRO completion. PROMIS SRI/SD and MEQ were all consistent/reliable (Cronbach's alpha 0.74-0.92). Chronotype breakdown showed 39% morning, 56% intermediate, and only 6% evening types. Moderate-severe SD and SRI were reported in 13% and 17% at baseline, and with significant improvement in SD at 4 weeks (P = .014). Fitbit-recorded measures showed a correlation at week 4 between WASO and SD (r = 0.35, P = .009) but not with SRI (r = 0.24, P = .08).

Conclusions

Collecting sleep data with Fitbits is feasible, PROs are consistent/reliable, > 10% of participants had SD and SRI that improved with smart wearable use, and SD was associated with WASO. The skewed chronotype distribution, risk and impact of sleep fragmentation mechanisms warrant further investigation.

Trial Registration

NCT04 669 574.

Predicting long-term sleep deprivation using wearable sensors and health surveys

Sufficient sleep is essential for individual well-being. Inadequate sleep has been shown to have significant negative impacts on our attention, cognition, and mood. The measurement of sleep from in-bed physiological signals has progressed to where commercial devices already incorporate this functionality. However, the prediction of sleep duration from previous awake activity is less studied. Previous studies have used daily exercise summaries, actigraph data, and pedometer data to predict sleep during individual nights. Building upon these, this article demonstrates how to predict a person's long-term average sleep length over the course of 30 days from Fitbit-recorded physical activity data alongside self-report surveys. Recursive Feature Elimination with Random Forest (RFE-RF) is used to extract the feature sets used by the machine learning models, and sex differences in the feature sets and performances of different machine learning models are then examined. The feature selection process demonstrates that previous sleep patterns and physical exercise are the most relevant kind of features for predicting sleep. Personality and depression metrics were also found to be relevant. When attempting to classify individuals as being long-term sleep-deprived, good performance was achieved across both the male, female, and combined data sets, with the highest-performing model achieving an AUC of 0.9762. The best-performing regression model for predicting the average nightly sleep time achieved an R-squared of 0.6861, with other models achieving similar results. When attempting to predict if a person who previously was obtaining sufficient sleep would become sleep-deprived, the best-performing model obtained an AUC of 0.9448.

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.

Circadian dysfunction and cardio-metabolic disorders in humans

The topic of human circadian rhythms is not only attracting the attention of clinical researchers from various fields but also sparking a growing public interest. The circadian system comprises the central clock, located in the suprachiasmatic nucleus of the hypothalamus, and the peripheral clocks in various tissues that are interconnected; together they coordinate many daily activities, including sleep and wakefulness, physical activity, food intake, glucose sensitivity and cardiovascular functions. Disruption of circadian regulation seems to be associated with metabolic disorders (particularly impaired glucose tolerance) and cardiovascular disease. Previous clinical trials revealed that disturbance of the circadian system, specifically due to shift work, is associated with an increased risk of type 2 diabetes mellitus. This review is intended to provide clinicians who wish to implement knowledge of circadian disruption in diagnosis and strategies to avoid cardio-metabolic disease with a general overview of this topic.

Sleep Patterns, Pain, and Emotional Functioning in Youth with Inflammatory Bowel Disease

Objective

Youth with inflammatory bowel disease (IBD) may be at increased risk for sleep difficulties due to the painful and inflammatory nature of their disease. Moreover, children and adolescents with IBD experience impairment across a variety of psychosocial domains. However, researchers have yet to investigate the complex interplay between sleep, disease-related symptoms, and psychosocial factors in this population. The purpose of this study was to examine sleep patterns, pain, and mood in pediatric IBD.

Methods

A sample of 25 children and adolescents with IBD (Mage = 14.24, Range = 10-18 years; 56% male) were recruited from a pediatric gastroenterology clinic. Youth wore an actigraphy watch and completed daily measures of affect and pain over the course of 14 days. Statistical analyses involved repeated measures general estimating equations.

Results

No significant association for sleep with negative affect was demonstrated. Despite majority of this sample being in disease remission, results revealed that increased sleep onset latency was associated with presence of next day pain and pain was associated with better next night sleep efficiency.

Conclusions

Findings of the current study suggest youth with IBD experience poor sleep quality, which is significantly related to the pain they experience. Consequently, healthcare providers should screen for and address sleep quality to optimize outcomes in their pediatric patients. Objectively assessing sleep patterns (e.g., actigraphy) may prove useful for pediatric IBD samples; however, additional research is needed to determine actigraphy's feasibility and efficacy in assessing sleep patterns in real world settings (e.g., pediatric medical clinics).

Feasibility of wearable activity tracking devices to measure physical activity and sleep change among adolescents with chronic pain-a pilot nonrandomized treatment study

Purpose

Personal informatics devices are being used to measure engagement in health behaviors in adults with chronic pain and may be appropriate for adolescent use. The aim of this study was to evaluate the utilization of a wearable activity tracking device to measure physical activity and sleep among adolescents attending a three-week, intensive interdisciplinary pain treatment (IIPT) program. We also assessed changes in physical activity and sleep from baseline to the treatment phase.

Methods

Participants (57.1% female, average age 15.88, SD = 1.27) wore an activity tracking device three weeks prior to starting and during the treatment program.

Results

Of 129 participants contacted, 47 (36.4%) agreed to participate. However, only 30 (64%) complied with the instructions for using the device prior to programming and during program participation. Preliminary analyses comparing averages from 3-weeks pre-treatment to 3-weeks during treatment indicated increases in daily overall activity minutes, daily step counts, and minutes of moderate to vigorous physical activity (by 353%), as well as a corresponding decrease in sedentary minutes. There was more missing data for sleep than anticipated.

Conclusions

Wearable activity tracking devices can be successfully used to measure adolescent physical activity in-person, with more difficulty obtaining this information remotely. Adolescents with chronic pain experience improvements in objective measurements of physical activity over the course of a 3-week IIPT program. Future studies may want to spend more time working with pediatric patients on their understanding of how to use trackers for sleep and physical activity.

Telepsychiatry in the Treatment of Major Depressive Disorders

This chapter explores the transformative role of telepsychiatry in managing major depressive disorders (MDD). Traversing geographical barriers and reducing stigma, this innovative branch of telemedicine leverages digital platforms to deliver effective psychiatric care. We investigate the evolution of telepsychiatry, examining its diverse interventions such as videoconferencing-based psychotherapy, medication management, and mobile applications. While offering significant advantages like increased accessibility, cost-effectiveness, and improved patient engagement, challenges in telepsychiatry include technological barriers, privacy concerns, ethical and legal considerations, and digital literacy gaps. Looking forward, emerging technologies like virtual reality, artificial intelligence, and precision medicine hold immense potential to personalize and enhance treatment effectiveness. Recognizing its limitations and advocating for equitable access, this chapter underscores telepsychiatry's power to revolutionize MDD treatment, making quality mental healthcare a reality for all.

Attention to Innate Circadian Rhythm and the Impact of Its Disruption on Diabetes

Novel strategies are required to reduce the risk of developing diabetes and/or clinical outcomes and complications of diabetes. In this regard, the role of the circadian system may be a potential candidate for the prevention of diabetes. We reviewed evidence from animal, clinical, and epidemiological studies linking the circadian system to various aspects of the pathophysiology and clinical outcomes of diabetes. The circadian clock governs genetic, metabolic, hormonal, and behavioral signals in anticipation of cyclic 24-hour events through interactions between a "central clock" in the suprachiasmatic nucleus and "peripheral clocks" in the whole body. Currently, circadian rhythmicity in humans can be subjectively or objectively assessed by measuring melatonin and glucocorticoid levels, core body temperature, peripheral blood, oral mucosa, hair follicles, rest-activity cycles, sleep diaries, and circadian chronotypes. In this review, we summarized various circadian misalignments, such as altered light-dark, sleep-wake, rest-activity, fasting-feeding, shift work, evening chronotype, and social jetlag, as well as mutations in clock genes that could contribute to the development of diabetes and poor glycemic status in patients with diabetes. Targeting critical components of the circadian system could deliver potential candidates for the treatment and prevention of type 2 diabetes mellitus in the future.

Comparison of Polysomnography, Single-Channel Electroencephalogram, Fitbit, and Sleep Logs in Patients With Psychiatric Disorders: Cross-Sectional Study

BACKGROUND

Sleep disturbances are core symptoms of psychiatric disorders. Although various sleep measures have been developed to assess sleep patterns and quality of sleep, the concordance of these measures in patients with psychiatric disorders remains relatively elusive.

OBJECTIVE

This study aims to examine the degree of agreement among 3 sleep recording methods and the consistency between subjective and objective sleep measures, with a specific focus on recently developed devices in a population of individuals with psychiatric disorders.

METHODS

We analyzed 62 participants for this cross-sectional study, all having data for polysomnography (PSG), Zmachine, Fitbit, and sleep logs. Participants completed questionnaires on their symptoms and estimated sleep duration the morning after the overnight sleep assessment. The interclass correlation coefficients (ICCs) were calculated to evaluate the consistency between sleep parameters obtained from each instrument. Additionally, Bland-Altman plots were used to visually show differences and limits of agreement for sleep parameters measured by PSG, Zmachine, Fitbit, and sleep logs.

RESULTS

The findings indicated a moderate agreement between PSG and Zmachine data for total sleep time (ICC=0.46; P<.001), wake after sleep onset (ICC=0.39; P=.002), and sleep efficiency (ICC=0.40; P=.006). In contrast, Fitbit demonstrated notable disagreement with PSG (total sleep time: ICC=0.08; wake after sleep onset: ICC=0.18; sleep efficiency: ICC=0.10) and exhibited particularly large discrepancies from the sleep logs (total sleep time: ICC=-0.01; wake after sleep onset: ICC=0.05; sleep efficiency: ICC=-0.02). Furthermore, subjective and objective concordance among PSG, Zmachine, and sleep logs appeared to be influenced by the severity of the depressive symptoms and obstructive sleep apnea, while these associations were not observed between the Fitbit and other sleep instruments.

CONCLUSIONS

Our study results suggest that Fitbit accuracy is reduced in the presence of comorbid clinical symptoms. Although user-friendly, Fitbit has limitations that should be considered when assessing sleep in patients with psychiatric disorders.

Development and validation of a sleep questionnaire, SNoRE 3.0, to evaluate sleep in companion dogs

Disturbances in the sleep-wake cycle are a debilitating, yet rather common condition not only in humans, but also in family dogs. While there is an emerging need for easy-to-use tools to document sleep alterations (in order to ultimately treat and/or prevent them), the veterinary tools which yield objective data (e.g. polysomnography, activity monitors) are both labor intensive and expensive. In this study, we developed a modified version of a previously used sleep questionnaire (SNoRE) and determined criterion validity in companion dogs against polysomnography and physical activity monitors (PAMs). Since a negative correlation between sleep time and cognitive performance in senior dogs has been demonstrated, we evaluated the correlation between the SNoRE scores and the Canine Dementia Scale (CADES, which includes a factor concerning sleep). There was a significant correlation between SNoRE 3.0 questionnaire scores and polysomnography data (latency to NREM sleep, ρ = 0.507, p < 0.001) as well as PAMs' data (activity between 1:00 and 3:00 AM, p < 0.05). There was a moderate positive correlation between the SNoRE 3.0 scores and the CADES scores (ρ = 0.625, p < 0.001). Additionally, the questionnaire structure was validated by a confirmatory factor analysis, and it also showed an adequate test-retest reliability. In conclusion the present paper describes a valid and reliable questionnaire tool, that can be used as a cost-effective way to monitor dog sleep in clinical settings.

Rest-Activity Rhythm Characteristics Associated With Depression Symptoms in Stroke Survivors

OBJECTIVE

To examine which 24-hour rest-activity rhythm (RAR) characteristics are associated with depression symptoms in stroke survivors.

DESIGN

Cross-sectional observational study examining associations of RAR characteristics with the presence of depression symptoms adjusting for age, sex, race, and medical comorbidity.

SETTING

Community setting.

PARTICIPANTS

Stroke survivors: (1) recruited locally (N women=35, N men=28) and (2) a nationally representative probability sample (the National Health and Nutrition Examination Survey [NHANES]; N women=156, N men=124).

INTERVENTIONS

None.

MEASUREMENTS

Objective RAR characteristics derived from accelerometer recordings including activity onset/offset times and non-parametric measures of RAR strength (relative amplitude), stability (interdaily stability), and fragmentation (intradaily variability). The presence of depression symptoms was categorized using Patient Health Questionnaire scores.

RESULTS

In both samples, the only RAR characteristic associated with depression symptoms was intradaily variability (fragmentation): local sample, odds ratio=1.96 [95% confidence interval=1.05-3.63]; NHANES sample, odds ratio=1.34, [95% confidence interval=1.01-1.78]). In the NHANES sample, which included both mild and moderate/severe depression, the association between 24-hour sleep-wake fragmentation and depression symptoms was driven by moderate-to-severe cases.

CONCLUSIONS

Stroke survivors with higher levels of RAR fragmentation were more likely to have depression symptoms in both samples. These findings have implications, given prior studies in general samples linking RAR fragmentation with future depression and dementia risk. Research is needed to establish the potential consequences, mechanisms, and modifiability of RAR fragmentation in stroke survivors.

The Effect of Night Float Rotation on Resident Sleep, Activity, and Well-Being

BACKGROUND

Night float call systems are becoming increasingly common at training programs with the goal of reducing fatigue related to sleep deprivation and sleep disturbance. Previous studies have shown that trainees obtain less sleep during the night float rotation and have decreased sleep efficiency for several days after the rotation. The impact on physical and emotional well-being has not been documented.

METHODS

Twenty-seven anesthesia residents were enrolled in a study using wearable sleep and activity trackers and National Institutes of Health Patient-Reported Outcome Measurement Information System (NIH PROMIS) surveys for sleep disturbance, fatigue, and positive affect to record data the week before ("baseline"), during ("night float"), and 1 week after ("recovery") their night float rotation. Each subject's data during the night float week and recovery week were compared to his or her own baseline week data using a paired, nonparametric analysis. The primary outcome variable was the change in average daily sleep hours during the night float week compared to the baseline week. Average daily rapid eye movement (REM) sleep, daily steps, and NIH PROMIS scores comparing night float and recovery weeks to baseline week were prespecified secondary outcomes. NIH PROMIS scores range from 0 to 100 with 50 as the national mean and more of the construct having a higher score.

RESULTS

There was no difference in average daily sleep hours between the night float and the baseline weeks (6.7 [5.9-7.8] vs 6.7 [5.5-7.7] hours, median [interquartile range]; P = .20). Residents had less REM sleep during the night float compared to the baseline weeks (1.1 [0.7-1.5] vs 1.4 [1.1-1.9] hours, P = .002). NIH PROMIS fatigue scores were higher during the night float than the baseline week (58.8 [54.6-65.1] vs 48.6 [46.0-55.1], P = .0004) and did not return to baseline during the recovery week (51.0 [48.6-58.8], P = .029 compared to baseline). Sleep disturbance was not different among the weeks. Positive affect was reduced after night float compared to baseline (39.6 [35.0-43.5] vs 44.8 [40.1-49.6], P = .0009), but returned to baseline during the recovery week (43.6 [39.6-48.2], P = .38).

CONCLUSIONS

The residents slept the same number of total hours during their night float week but had less REM sleep, were more fatigued, and had less positive affect. All of these resolved to baseline except fatigue, that was still greater than the baseline week. This methodology appears to robustly capture psychophysiological data that might be useful for quality initiatives.

Machine Learning-Based Prediction of Attention-Deficit/Hyperactivity Disorder and Sleep Problems With Wearable Data in Children

IMPORTANCE

Early detection of attention-deficit/hyperactivity disorder (ADHD) and sleep problems is paramount for children's mental health. Interview-based diagnostic approaches have drawbacks, necessitating the development of an evaluation method that uses digital phenotypes in daily life.

OBJECTIVE

To evaluate the predictive performance of machine learning (ML) models by setting the data obtained from personal digital devices comprising training features (ie, wearable data) and diagnostic results of ADHD and sleep problems by the Kiddie Schedule for Affective Disorders and Schizophrenia Present and Lifetime Version for Diagnostic and Statistical Manual of Mental Disorders, 5th edition (K-SADS) as a prediction class from the Adolescent Brain Cognitive Development (ABCD) study.

DESIGN, SETTING, AND PARTICIPANTS

In this diagnostic study, wearable data and K-SADS data were collected at 21 sites in the US in the ABCD study (release 3.0, November 2, 2020, analyzed October 11, 2021). Screening data from 6571 patients and 21 days of wearable data from 5725 patients collected at the 2-year follow-up were used, and circadian rhythm-based features were generated for each participant. A total of 12 348 wearable data for ADHD and 39 160 for sleep problems were merged for developing ML models.

MAIN OUTCOMES AND MEASURES

The average performance of the ML models was measured using an area under the receiver operating characteristics curve (AUC), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). In addition, the Shapley Additive Explanations value was used to calculate the importance of features.

RESULTS

The final population consisted of 79 children with ADHD problems (mean [SD] age, 144.5 [8.1] months; 55 [69.6%] males) vs 1011 controls and 68 with sleep problems (mean [SD] age, 143.5 [7.5] months; 38 [55.9%] males) vs 3346 controls. The ML models showed reasonable predictive performance for ADHD (AUC, 0.798; sensitivity, 0.756; specificity, 0.716; PPV, 0.159; and NPV, 0.976) and sleep problems (AUC, 0.737; sensitivity, 0.743; specificity, 0.632; PPV, 0.036; and NPV, 0.992).

CONCLUSIONS AND RELEVANCE

In this diagnostic study, an ML method for early detection or screening using digital phenotypes in children's daily lives was developed. The results support facilitating early detection in children; however, additional follow-up studies can improve its performance.

CovidRhythm: A Deep Learning Model for Passive Prediction of Covid-19 Using Biobehavioral Rhythms Derived From Wearable Physiological Data

Goal: To investigate whether a deep learning model can detect Covid-19 from disruptions in the human body's physiological (heart rate) and rest-activity rhythms (rhythmic dysregulation) caused by the SARS-CoV-2 virus. Methods: We propose CovidRhythm, a novel Gated Recurrent Unit (GRU) Network with Multi-Head Self-Attention (MHSA) that combines sensor and rhythmic features extracted from heart rate and activity (steps) data gathered passively using consumer-grade smart wearable to predict Covid-19. A total of 39 features were extracted (standard deviation, mean, min/max/avg length of sedentary and active bouts) from wearable sensor data. Biobehavioral rhythms were modeled using nine parameters (mesor, amplitude, acrophase, and intra-daily variability). These features were then input to CovidRhythm for predicting Covid-19 in the incubation phase (one day before biological symptoms manifest). Results: A combination of sensor and biobehavioral rhythm features achieved the highest AUC-ROC of 0.79 [Sensitivity = 0.69, Specificity = 0.89, F[Formula: see text] = 0.76], outperforming prior approaches in discriminating Covid-positive patients from healthy controls using 24 hours of historical wearable physiological. Rhythmic features were the most predictive of Covid-19 infection when utilized either alone or in conjunction with sensor features. Sensor features predicted healthy subjects best. Circadian rest-activity rhythms that combine 24 h activity and sleep information were the most disrupted. Conclusions: CovidRhythm demonstrates that biobehavioral rhythms derived from consumer-grade wearable data can facilitate timely Covid-19 detection. To the best of our knowledge, our work is the first to detect Covid-19 using deep learning and biobehavioral rhythms features derived from consumer-grade wearable data.

Association of 24-Hour Activity Pattern Phenotypes With Depression Symptoms and Cognitive Performance in Aging

Importance

Evidence regarding the nature and prevalence of 24-hour activity pattern phenotypes in older adults, especially those related to depression symptoms and cognition, is needed to guide the development of targeted mechanism research and behavioral interventions.

Objectives

To identify subgroups of older adults with similar 24-hour activity rhythm characteristics and characterize associated depression symptoms and cognitive performance.

Design, Setting, and Participants

From January to March 2022, a cross-sectional analysis of the 2011-2014 National Health and Nutrition Examination and Survey (NHANES) accelerometer study was conducted. The NHANES used a multistage probability sample that was designed to be representative of noninstitutionalized adults in the US. The main analysis included participants 65 years or older who had accelerometer and depression measures weighted to represent approximately 32 million older adults.

Exposures

Latent profile analysis identified subgroups with similar 24-hour activity pattern characteristics as measured using extended-cosine and nonparametric methods.

Main Outcomes and Measures

Covariate-adjusted sample-weighted regressions assessed associations of subgroup membership with (1) depression symptoms defined as 9-Item Patient Health Questionnaire (PHQ-9) scores of 10 or greater (PHQ-9) and (2) having at least psychometric mild cognitive impairment (p-MCI) defined as scoring less than 1 SD below the mean on a composite cognitive performance score.

Results

The actual clustering sample size was 1800 (weighted: mean [SD] age, 72.9 [7.3] years; 57% female participants). Clustering identified 4 subgroups: (1) 677 earlier rising/robust (37.6%), (2) 587 shorter active period/less modelable (32.6%), (3) 177 shorter active period/very weak (9.8%), and (4) 359 later settling/very weak (20.0%). The prevalence of a PHQ-9 score of 10 or greater differed significantly across groups (cluster 1, 3.5%; cluster 2, 4.7%; cluster 3, 7.5%; cluster 4, 9.0%; χ2 P = .004). The prevalence of having at least p-MCI differed significantly across groups (cluster 1, 7.2%; cluster 2, 12.0%; cluster 3, 21.0%; cluster 4, 18.0%; χ2 P < .001). Five of 9 depression symptoms differed significantly across subgroups.

Conclusions and Relevance

In this cross-sectional study, findings indicate that approximately 1 in 5 older adults in the US may be classified in a subgroup with weak activity patterns and later settling, and approximately 1 in 10 may be classified in a subgroup with weak patterns and shorter active duration. Future research is needed to investigate the biologic processes related to these behavioral phenotypes, including why earlier and robust activity patterns appear protective, and whether modifying disrupted patterns improves outcomes.

A Computational Framework for Modeling Biobehavioral Rhythms from Mobile and Wearable Data Streams

This paper presents a computational framework for modeling biobehavioral rhythms - the repeating cycles of physiological, psychological, social, and environmental events - from mobile and wearable data streams. The framework incorporates four main components: mobile data processing, rhythm discovery, rhythm modeling, and machine learning. We evaluate the framework with two case studies using datasets of smartphone, Fitbit, and OURA smart ring to evaluate the framework’s ability to (1) detect cyclic biobehavior, (2) model commonality and differences in rhythms of human participants in the sample datasets, and (3) predict their health and readiness status using models of biobehavioral rhythms. Our evaluation demonstrates the framework’s ability to generate new knowledge and findings through rigorous micro- and macro-level modeling of human rhythms from mobile and wearable data streams collected in the wild and using them to assess and predict different life and health outcomes.

External Auditory Stimulation as a Non-Pharmacological Sleep Aid

The increased demand for well-being has fueled interest in sleep. Research in technology for monitoring sleep ranges from sleep efficiency and sleep stage analysis to sleep disorder detection, centering on wearable devices such as fitness bands, and some techniques have been commercialized and are available to consumers. Recently, as interest in digital therapeutics has increased, the field of sleep engineering demands a technology that helps people obtain quality sleep that goes beyond the level of monitoring. In particular, interest in sleep aids for people with or without insomnia but who cannot fall asleep easily at night is increasing. In this review, we discuss experiments that have tested the sleep-inducing effects of various auditory stimuli currently used for sleep-inducing purposes. The auditory stimulations were divided into (1) colored noises such as white noise and pink noise, (2) autonomous sensory meridian response sounds such as natural sounds such as rain and firewood burning, sounds of whispers, or rubbing various objects with a brush, and (3) classical music or a preferred type of music. For now, the current clinical method of receiving drugs or cognitive behavioral therapy to induce sleep is expected to dominate. However, it is anticipated that devices or applications with proven ability to induce sleep clinically will begin to appear outside the hospital environment in everyday life.

Greater variability in daily sleep efficiency predicts depression and anxiety in young adults: Estimation of depression severity using the two-week sleep quality records of wearable devices

OBJECTIVES

Sleep disturbances are associated with both the onset and progression of depressive disorders. It is important to capture day-to-day variability in sleep patterns; irregular sleep is associated with depressive symptoms. We used sleep efficiency, measured with wearable devices, as an objective indicator of daily sleep variability.

MATERIALS AND METHODS

The total sample consists of 100 undergraduate and graduate students, 60% of whom were female. All were divided into three groups (with major depressive disorder, mild depressive symptoms, and controls). Self-report questionnaires were completed at the beginning of the experiment, and sleep efficiency data were collected daily for 2 weeks using wearable devices. We explored whether the mean value of sleep efficiency, and its variability, predicted the severity of depression using dynamic structural equation modeling.

RESULTS

More marked daily variability in sleep efficiency significantly predicted levels of depression and anxiety, as did the average person-level covariates (longer time in bed, poorer quality of life, lower extraversion, and higher neuroticism).

CONCLUSION

Large swings in day-to-day sleep efficiency and certain clinical characteristics might be associated with depression severity in young adults.

Feasibility of a wearable biosensor device to characterize exercise and sleep in neurology residents

BACKGROUND

Research suggests optimizing sleep, exercise and work-life balance may improve resident physician burnout. Wearable biosensors may allow residents to detect and correct poor sleep and exercise habits before burnout develops. Our objectives were to evaluate the feasibility of a wearable biosensor to characterize exercise/sleep in neurology residents and examine its relationship to self-reported, validated survey measures. We also assessed the device's impact on well-being and barriers to use.

METHODS

This prospective cohort study evaluated the WHOOP Strap 2.0 in neurology residents. Participants completed regular online surveys, including self-reported hours of sleep/exercise, and validated sleep/exercise scales at 3-month intervals. Autonomic, exercise, and sleep measures were obtained from WHOOP. Changes were evaluated over time via linear regression. Survey and WHOOP metrics were compared using Pearson correlations.

RESULTS

Sixteen (72.7%) of 22 eligible participants enrolled. Eleven (68.8%) met the minimum usage requirement (6+ months) and were classified as 'consecutive wearers.' Significant increases were found in sleep duration and exercise intensity. Moderate-to-low correlations were found between survey responses and WHOOP measures. Most (73%) participants reported a positive impact on well-being. Barriers to use included 'Forgetting to wear' (20%) and 'not motivational' (23.3%).

CONCLUSION

Wearable biosensors may be a feasible tool to evaluate sleep/exercise in residents.

Characterisation of ICU sleep by a commercially available activity tracker and its agreement with patient-perceived sleep quality

Background

A low-cost, quantitative method to evaluate sleep in the intensive care unit (ICU) that is both feasible for routine clinical practice and reliable does not yet exist. We characterised nocturnal ICU sleep using a commercially available activity tracker and evaluated agreement between tracker-derived sleep data and patient-perceived sleep quality.

Patients and methods

A prospective cohort study was performed in a 40-bed ICU at a community teaching hospital. An activity tracker (Fitbit Charge 2) was applied for up to 7 ICU days in English-speaking adults with an anticipated ICU stay ≥2 days and without mechanical ventilation, sleep apnoea, delirium, continuous sedation, contact isolation or recent anaesthesia. The Richards-Campbell Sleep Questionnaire (RCSQ) was administered each morning by a trained investigator.

Results

Available activity tracker-derived data for each ICU study night (20:00–09:00) (total sleep time (TST), number of awakenings (#AW), and time spent light sleep, deep sleep and rapid eye movement (REM) sleep) were downloaded and analysed. Across the 232 evaluated nights (76 patients), TST and RCSQ data were available for 232 (100%), #AW data for 180 (78%) and sleep stage data for 73 (31%). Agreement between TST (349±168 min) and RCSQ Score was moderate and significant (r=0.34; 95% CI 0.18 to 0.48). Agreement between #AW (median (IQR), 4 (2–9)) and RCSQ Score was negative and non-significant (r=−0.01; 95% CI −0.19 to 0.14). Agreement between time (min) spent in light (259 (182 to 328)), deep (43±29), and REM (47 (28–72)) sleep and RCSQ Score was moderate but non-significant (light (r=0.44, 95% CI −0.05 to 0.36); deep sleep (r=0.44, 95% CI −0.11 to 0.15) and REM sleep (r=0.44; 95% CI −0.21 to 0.21)).

Conclusions

A Fitbit Charge 2 when applied to non-intubated adults in an ICU consistently collects TST data but not #AW or sleep stage data at night. The TST moderately correlates with patient-perceived sleep quality; a correlation between either #AW or sleep stages and sleep quality was not found.

Relationship between Sport-Related Concussion and Sleep Based on Self-Report and Commercial Actigraph Measurement

Sleep-wake disturbance (SWD) results from sport-related concussion (SRC) and may increase risk of protracted post-injury symptoms. However, methodological limitations in the extant literature limit our understanding of the role of SWD in SRC. This study examined the association between acute/subacute SRC and two sleep behaviors—sleep duration and efficiency—as measured by self-report and commercially available actigraphy (CA) in a sample of football players enrolled in a larger prospective longitudinal study of concussion. Fifty-seven high school and Division 3 male football players with SRC (mean [M] age = 18.00 years, standard deviation [SD] = 1.44) and 26 male teammate controls (M age = 18.54 years, SD = 2.21) were enrolled in this prospective pilot study. Sleep duration and sleep efficiency were recorded nightly for 2 weeks (starting 24–48 h post-injury in the SRC group) via CA and survey delivered via mobile application. There was no significant relationship between SRC and objectively recorded sleep measures, a null finding. However, the SRC group reported a brief (3-day) reduction in sleep efficiency after injury (M SRC = 82.18, SD = 12.24; M control = 89.2, SD = 4.25; p = 0.013; Cohen's d = 0.77), with no change in sleep duration. Self-reported and actigraph-assessed hours of sleep were weakly and insignificantly correlated in the SRC group (r = −0.21, p = 0.145), whereas they were robustly correlated in the non-injured control group (r = 0.65, p = 0.004). SWD post-SRC was not observed in objectively measured sleep duration or sleep efficiency and was modest and time-limited based on self-reported sleep efficiency. The weak correlation between self-reported and objective sleep behavior measures implies that subjective experience of SWD post-SRC may be due to factors other than actual changes in these observable sleep behaviors. Clinically, SWD in the early-subacute stages of recovery from SRC may not be adequately measurable via current CA. Subjective SWD may require alternative methods of evaluation (e.g., clinical actigraph or sleep study).

Intra- and Inter-Model Variability of Light Detection Using a Commercially Available Light Sensor

The veracity of claims made by researchers and clinicians when reporting the impact of lighting on vision and other biological mechanisms is, in part, reliant on accurate and valid measurement devices. We aim to quantify the intra- and inter-watch variability of a commercially available light sensor device which has been widely used in vision and other photobiological research. Intra- and inter-watch differences were investigated between four Actiwatch Spectrum Pro devices. The devices were used to obtain measurements on two separate occasions, under three different controlled light conditions; the Gretag Macbeth Judge II lightbox was used to produce Simulated Daylight (D65), Illuminant A (A) and Cool White Fluorescent (CWF) lighting. Significant inter-watch differences were noted when considering tricolour (red, green, blue) and the white sensor outputs under each of the three illuminants (p < 0.01). A significant interaction was also found between tricolour sensor and watch used (p < 0.01). Intra-watch differences were noted for the tricolour and for the white sensor outputs under the three illuminants (≤0.05), for all but one watch which showed no significant intra-watch difference for the white 'sensor output' under the D65 illuminant. Use of spectral sensitivity devices is an evolving field. Before drawing causal relationships between light and other biological processes, researchers should acknowledge the limitations of the instruments used, their validation, and the resultant data. The outcomes of the study indicate caution must be exercised in longitudinal data collection and the mixing of watches amongst study participants should be avoided.

Impact of a Pharmacist-Led Intensive Care Unit Sleep Improvement Protocol on Sleep Duration and Quality

BACKGROUND

Sleep improvement protocols are recommended for use in the intensive care unit (ICU) despite questions regarding which interventions to include, whether sleep quality or duration will improve, and the role of pharmacists in their development and implementation.

OBJECTIVE

To characterize the impact of a pharmacist-led, ICU sleep improvement protocol on sleep duration and quality as evaluated by a commercially available activity tracker and patient perception.

METHODS

Critical care pharmacists from a 40-bed, mixed ICU at a large community hospital led the development and implementation of an interprofessional sleep improvement protocol. It included daily pharmacist medication review to reduce use of medications known to disrupt sleep or increase delirium and guideline-based recommendations on both environmental and nonpharmacological sleep-focused interventions. Sleep duration and quality were compared before (December 2018 to December 2019) and after (January to June 2019) protocol implementation in non-mechanically ventilated adults using both objective (total nocturnal sleep time [TST] measured by an activity tracker (Fitbit Charge 2) and subjective (patient-perceived sleep quality using the Richards-Campbell Sleep Questionnaire [RCSQ]) measures.

RESULTS

Groups before (n = 48) and after (n = 29) sleep protocol implementation were well matched. After protocol implementation, patients had a longer TST (389 ± 123 vs 310 ± 147 minutes; P = 0.02) and better RCSQ-perceived sleep quality (63 ± 18 vs 42 ± 24 mm; P = 0.0003) compared with before implementation.

CONCLUSION AND RELEVANCE

A sleep protocol that incorporated novel elements led to objective and subjective improvements in ICU sleep duration and quality. Application of this study may result in increased utilization of sleep protocols and pharmacist involvement.

Optimizing Behavioral Sleep Strategies

Patients are increasingly looking to optimize sleep as a health and wellness strategy. Sleep health is often individualized based on three elements that correspond to overall physical and mental well-being: (1) sleep quality, which refers to the continuity and depth of sleep as well as a feeling of restoration upon awakening; (2) sleep quantity, which refers to the duration of sleep that is appropriate for a given age group; and (3) timing of the sleep window, which refers to the positioning of sleep that is aligned with an individual's circadian rhythm for sleep or an ideal circadian zone. In the past, prescribing hypnotic medications was considered the primary approach for improving sleep. However, there has been a recent paradigm shift to favor behavioral approaches, particularly in the case of insomnia where cognitive-behavioral therapy has been shown to have a more favorable benefit-to-harm profile than medications. The clinical vignette is presented here as a springboard for discussion regarding the latest evidence and efficacy for sleep behavior techniques and consumer monitoring devices developed to improve sleep health and awareness for clinicians to consider when educating their patients on maximizing sleep health behaviors.

Effectiveness of a Smartphone App With a Wearable Activity Tracker in Preventing the Recurrence of Mood Disorders: Prospective Case-Control Study

BACKGROUND

Smartphones and wearable devices can be used to obtain diverse daily log data related to circadian rhythms. For patients with mood disorders, giving feedback via a smartphone app with appropriate behavioral correction guides could play an important therapeutic role in the real world.

OBJECTIVE

We aimed to evaluate the effectiveness of a smartphone app named Circadian Rhythm for Mood (CRM), which was developed to prevent mood episodes based on a machine learning algorithm that uses passive digital phenotype data of circadian rhythm behaviors obtained with a wearable activity tracker. The feedback intervention for the CRM app consisted of a trend report of mood prediction, H-score feedback with behavioral guidance, and an alert system triggered when trending toward a high-risk state.

METHODS

In total, 73 patients with a major mood disorder were recruited and allocated in a nonrandomized fashion into 2 groups: the CRM group (14 patients) and the non-CRM group (59 patients). After the data qualification process, 10 subjects in the CRM group and 33 subjects in the non-CRM group were evaluated over 12 months. Both groups were treated in a similar manner. Patients took their usual medications, wore a wrist-worn activity tracker, and checked their eMoodChart daily. Patients in the CRM group were provided with daily feedback on their mood prediction and health scores based on the algorithm. For the CRM group, warning alerts were given when irregular life patterns were observed. However, these alerts were not given to patients in the non-CRM group. Every 3 months, mood episodes that had occurred in the previous 3 months were assessed based on the completed daily eMoodChart for both groups. The clinical course and prognosis, including mood episodes, were evaluated via face-to-face interviews based on the completed daily eMoodChart. For a 1-year prospective period, the number and duration of mood episodes were compared between the CRM and non-CRM groups using a generalized linear model.

RESULTS

The CRM group had 96.7% fewer total depressive episodes (n/year; exp β=0.033, P=.03), 99.5% shorter depressive episodes (total; exp β=0.005, P<.001), 96.1% shorter manic or hypomanic episodes (exp β=0.039, P<.001), 97.4% fewer total mood episodes (exp β=0.026, P=.008), and 98.9% shorter mood episodes (total; exp β=0.011, P<.001) than the non-CRM group. Positive changes in health behaviors due to the alerts and in wearable device adherence rates were observed in the CRM group.

CONCLUSIONS

The CRM app with a wearable activity tracker was found to be effective in preventing and reducing the recurrence of mood disorders, improving prognosis, and promoting better health behaviors. Patients appeared to develop a regular habit of using the CRM app.

TRIAL REGISTRATION

ClinicalTrials.gov NCT03088657; https://clinicaltrials.gov/ct2/show/NCT03088657.

Feasibility of Using a Wearable Biosensor Device in Patients at Risk for Alzheimer's Disease Dementia

BACKGROUND

Alzheimer's disease (AD) is the most common and most costly chronic neurodegenerative disease globally. AD develops over an extended period prior to cognitive symptoms, leaving a "window of opportunity" for targeted risk-reduction interventions. Further, this pre-dementia phase includes early physiological changes in sleep and autonomic regulation, for which wearable biosensor devices may offer a convenient and cost-effective method to assess AD-risk.

METHODS

Patients with a family history of AD and no or minimal cognitive complaints were recruited from the Alzheimer's Prevention Clinic at Weill Cornell Medicine and New York-Presbyterian. Of the 40 consecutive patients screened, 34 (85%) agreed to wear a wearable biosensor device (WHOOP). One subject (2.5%) lost the device prior to data collection. Of the remaining subjects, 24 were classified as normal cognition and were asymptomatic, 6 were classified as subjective cognitive decline, and 3 were amyloid-positive (one with pre-clinical AD, one with pre-clinical Lewy-Body Dementia, and one with mild cognitive impairment due to AD). Sleep-cycle, autonomic (heart rate variability [HRV]) and activity measures were collected via WHOOP. Blood biomarkers and neuropsychological testing sensitive to cognitive changes in pre-clinical AD were obtained. Participants completed surveys assessing their sleep-patterns, exercise habits, and attitudes towards WHOOP. The goal of this prospective observational study was to determine the feasibility of using a wrist-worn biosensor device in patients at-risk for AD dementia. Unsupervised machine learning was performed to first separate participants into distinct phenotypic groups using the multivariate biometric data. Additional statistical analyses were conducted to examine correlations between individual biometric measures and cognitive performance.

RESULTS

27 (81.8%) participants completed the follow-up surveys. Twenty-four participants (88.9%) were satisfied with WHOOP after six months, and twenty-three (85.2%) wanted to continue wearing WHOOP. K-means clustering separated participants into two groups. Group 1 was older, had lower HRV, and spent more time in slow-wave sleep (SWS) than Group 2. Group 1 performed better on two cognitive tests assessing executive function: Flanker Inhibitory Attention/Control (FIAC) (p=.031), and Dimensional Change Card Sort (DCCS) (p=.061). In Group 1, DCCS was correlated with SWS (ρ=.68, p=0.024) and HRV (ρ=.6, p=0.019). In Group 2, DCCS was correlated with HRV (ρ=.55, p=0.018). There were no significant differences in blood biomarkers between the two groups.

CONCLUSIONS

Wearable biosensor devices may be a feasible tool to assess AD-related physiological changes. Longitudinal collection of sleep and HRV data may potentially be a non-invasive method for monitoring cognitive changes related to pre-clinical AD. Further study is warranted in larger populations.

Wearable devices for remote vital signs monitoring in the outpatient setting: an overview of the field

Early detection of physiological deterioration has been shown to improve patient outcomes. Due to recent improvements in technology, comprehensive outpatient vital signs monitoring is now possible. This is the first review to collate information on all wearable devices on the market for outpatient physiological monitoring.

A scoping review was undertaken. The monitors reviewed were limited to those that can function in the outpatient setting with minimal restrictions on the patient’s normal lifestyle, while measuring any or all of the vital signs: heart rate, ECG, oxygen saturation, respiration rate, blood pressure and temperature.

A total of 270 papers were included in the review. Thirty wearable monitors were examined: 6 patches, 3 clothing-based monitors, 4 chest straps, 2 upper arm bands and 15 wristbands. The monitoring of vital signs in the outpatient setting is a developing field with differing levels of evidence for each monitor. The most common clinical application was heart rate monitoring. Blood pressure and oxygen saturation measurements were the least common applications. There is a need for clinical validation studies in the outpatient setting to prove the potential of many of the monitors identified.

Research in this area is in its infancy. Future research should look at aggregating the results of validity and reliability and patient outcome studies for each monitor and between different devices. This would provide a more holistic overview of the potential for the clinical use of each device.

Field-based Measurement of Sleep: Agreement between Six Commercial Activity Monitors and a Validated Accelerometer

OBJECTIVE

To examine agreement between multiple commercial activity monitors (CAMs) and a validated actigraph to measure sleep.

METHODS

Thirty adults without sleep disorders wore an Actiwatch Spectrum (AW) and alternated wearing 6 CAMs for one 24-h period each (Fitbit Alta, Jawbone Up3, Misfit Shine 2, Polar A360, Samsung Gear Fit2, Xiaomi Mi Band 2). Total sleep time (TST) and wake after sleep onset (WASO) were compared between edited AW and unedited CAM outputs. Comparisons between AW and CAM data were made via paired t-tests, mean absolute percent error (MAPE) calculations, and intra-class correlations (ICC). Intra-model reliability was performed in 10 participants who wore a pair of each AW and CAM model.

RESULTS

Fitbit, Jawbone, Misfit, and Xiaomi overestimated TST relative to AW (53.7-80.4 min, P ≤ .001). WASO was underestimated by Fitbit, Misfit, Samsung and Xiaomi devices (15.0-27.9 min; P ≤ .004) and overestimated by Polar (27.7 min, P ≤ .001). MAPEs ranged from 5.1% (Samsung) to 25.4% (Misfit) for TST and from 36.6% (Fitbit) to 165.1% (Polar) for WASO. TST ICCs ranged from .00 (Polar) to .92 (Samsung), while WASO ICCs ranged from .38 (Misfit) to .69 (Samsung). Differences were similar between poor sleepers (Pittsburgh Sleep Quality Index global score >5; n = 10) and good sleepers. Intra-model reliability analyses revealed minimal between-pair differences and high ICCs.

CONCLUSIONS

Agreement between CAMs and AW varied by device, with greater agreement observed for TST than WASO. While reliable, variability in agreement across CAMs with traditional actigraphy may complicate the interpretation of CAM data obtained for clinical or research purposes.

Accuracy of Wristband Fitbit Models in Assessing Sleep: Systematic Review and Meta-Analysis

BACKGROUND

Wearable sleep monitors are of high interest to consumers and researchers because of their ability to provide estimation of sleep patterns in free-living conditions in a cost-efficient way.

OBJECTIVE

We conducted a systematic review of publications reporting on the performance of wristband Fitbit models in assessing sleep parameters and stages.

METHODS

In adherence with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement, we comprehensively searched the Cumulative Index to Nursing and Allied Health Literature (CINAHL), Cochrane, Embase, MEDLINE, PubMed, PsycINFO, and Web of Science databases using the keyword Fitbit to identify relevant publications meeting predefined inclusion and exclusion criteria.

RESULTS

The search yielded 3085 candidate articles. After eliminating duplicates and in compliance with inclusion and exclusion criteria, 22 articles qualified for systematic review, with 8 providing quantitative data for meta-analysis. In reference to polysomnography (PSG), nonsleep-staging Fitbit models tended to overestimate total sleep time (TST; range from approximately 7 to 67 mins; effect size=-0.51, P<.001; heterogenicity: I2=8.8%, P=.36) and sleep efficiency (SE; range from approximately 2% to 15%; effect size=-0.74, P<.001; heterogenicity: I2=24.0%, P=.25), and underestimate wake after sleep onset (WASO; range from approximately 6 to 44 mins; effect size=0.60, P<.001; heterogenicity: I2=0%, P=.92) and there was no significant difference in sleep onset latency (SOL; P=.37; heterogenicity: I2=0%, P=.92). In reference to PSG, nonsleep-staging Fitbit models correctly identified sleep epochs with accuracy values between 0.81 and 0.91, sensitivity values between 0.87 and 0.99, and specificity values between 0.10 and 0.52. Recent-generation Fitbit models that collectively utilize heart rate variability and body movement to assess sleep stages performed better than early-generation nonsleep-staging ones that utilize only body movement. Sleep-staging Fitbit models, in comparison to PSG, showed no significant difference in measured values of WASO (P=.25; heterogenicity: I2=0%, P=.92), TST (P=.29; heterogenicity: I2=0%, P=.98), and SE (P=.19) but they underestimated SOL (P=.03; heterogenicity: I2=0%, P=.66). Sleep-staging Fitbit models showed higher sensitivity (0.95-0.96) and specificity (0.58-0.69) values in detecting sleep epochs than nonsleep-staging models and those reported in the literature for regular wrist actigraphy.

CONCLUSIONS

Sleep-staging Fitbit models showed promising performance, especially in differentiating wake from sleep. However, although these models are a convenient and economical means for consumers to obtain gross estimates of sleep parameters and time spent in sleep stages, they are of limited specificity and are not a substitute for PSG.

A validation study of a consumer wearable sleep tracker compared to a portable EEG system in naturalistic conditions

OBJECTIVE

To compare a wearable device, the Fitbit Versa (FV), to a validated portable single-channel EEG system across multiple nights in a naturalistic environment.

METHODS

Twenty participants (10 men and 10 women) aged 25-67 years were recruited for the present study. Study duration was 14 days during which participants were asked to wear the FV daily and nightly. The study intended to reproduce free-living conditions; thus, no guidelines for sleep or activity were imposed on the participants. A total of 138 person-nights, equivalent to 76,539 epochs, were used in the validation process. Sleep measures were compared between the FV and portable EEG using Bland-Altman plots, paired t-tests and epoch-by-epoch (EBE) analyses.

RESULTS

The FV showed no significant bias with the EEG for the global sleep measures time in bed (TIB) and total sleep time (TST), and for calculated sleep efficiency (cSE = [TST/TIB] x 100). The FV had 92.1% sensitivity, 54.1% specificity, and 88.5% accuracy with a Cohen's kappa of 0.41, but a prevalence- and bias adjusted kappa of 0.77. The predictive values for sleep (PVS; positive predictive value) and wakefulness (PVW; negative predictive value) were 95.0% and 42.0%, respectively. The FV showed significant bias compared to the portable EEG for time spent in specific sleep stages, for SE as provided by FV, for sleep onset latency, sleep period time, and wake after sleep onset.

CONCLUSIONS

The consumer sleep tracker could be a useful tool for measuring sleep duration in longitudinal epidemiologic naturalistic studies albeit with some limitations in specificity.

Digital phenotyping by consumer wearables identifies sleep-associated markers of cardiovascular disease risk and biological aging

Sleep is associated with various health outcomes. Despite their growing adoption, the potential for consumer wearables to contribute sleep metrics to sleep-related biomedical research remains largely uncharacterized. Here we analyzed sleep tracking data, along with questionnaire responses and multi-modal phenotypic data generated from 482 normal volunteers. First, we compared wearable-derived and self-reported sleep metrics, particularly total sleep time (TST) and sleep efficiency (SE). We then identified demographic, socioeconomic and lifestyle factors associated with wearable-derived TST; they included age, gender, occupation and alcohol consumption. Multi-modal phenotypic data analysis showed that wearable-derived TST and SE were associated with cardiovascular disease risk markers such as body mass index and waist circumference, whereas self-reported measures were not. Using wearable-derived TST, we showed that insufficient sleep was associated with premature telomere attrition. Our study highlights the potential for sleep metrics from consumer wearables to provide novel insights into data generated from population cohort studies.

Experimental Assessment of Sleep-Related Parameters by Passive Infrared Sensors: Measurement Setup, Feature Extraction, and Uncertainty Analysis

A simple sleep monitoring measurement method is presented in this paper, based on a simple, non-invasive motion sensor, the Passive InfraRed (PIR) motion sensor. The easy measurement set-up proposed is presented and its performances are compared with the ones provided by a commercial, ballistocardiographic bed sensor, used as reference tool. Testing was conducted on 25 nocturnal acquisitions with a voluntary, healthy subject, using the PIR-based proposed method and the reference sensor, simultaneously. A dedicated algorithm was developed to correlate the bed sensor outputs with the PIR signal to extract sleep-related features: sleep latency (SL), sleep interruptions (INT), and time to wake (TTW). Such sleep parameters were automatically identified by the algorithm, and then correlated to the ones computed by the reference bed sensor. The identification of these sleep parameters allowed the computation of an important, global sleep quality parameter: the sleep efficiency (SE). It was calculated for each nocturnal acquisition and then correlated to the SE values provided by the reference sensor. Results show the correlation between the SE values monitored with the PIR and the bed sensor with a robust statistic confidence of 4.7% for the measurement of SE (coverage parameter k = 2), indicating the validity of the proposed, unobstructive approach, based on a simple, small, and low-cost sensor, for the assessment of important sleep-related parameters.

Consumer Wearable Deployments in Actigraphy Research: Evaluation of an Observational Study

BACKGROUND

Consumer wearables can provide a practical and accessible method of data collection in actigraphy research. However, as this area continues to grow, it is becoming increasingly important for researchers to be aware of the many challenges facing the capture of quality data using consumer wearables.

OBJECTIVE

This study aimed to (1) present the challenges encountered by a research team in actigraphy data collection using a consumer wearable and (2) present considerations for researchers to apply in the pursuit of robust data using this approach.

METHODS

The Nokia Go was deployed to 33 elite Gaelic footballers from a single team for a planned period of 14 weeks. A bring-your-own-device model was employed for this study where the Health Mate app was downloaded on participants' personal mobile phones and connected to the Nokia Go via Bluetooth. Retrospective evaluation of the researcher and participant experience was conducted through transactional data such as study logs and email correspondence. The participant experience of the data collection process was further explored through the design of a 34-question survey utilizing aspects of the Technology Acceptance Model.

RESULTS

Researcher challenges included device disconnection, logistics and monitoring, and rectifying of technical issues. Participant challenges included device syncing, loss of the device, and wear issues, particularly during contact sport. Following disconnection issues, the data collection period was defined as 87 days for which there were 18 remaining participants. Average wear time was 79 out of 87 days (90%) and 20.8 hours per day. The participant survey found mainly positive results regarding device comfort, perceived ease of use, and perceived usefulness.

CONCLUSIONS

Although this study did not encounter some of the common published barriers to wearable data collection, our experience was impacted by technical issues such as disconnection and syncing challenges, practical considerations such as loss of the device, issues with personal mobile phones in the bring-your-own-device model, and the logistics and resources required to ensure a smooth data collection with an active cohort. Recommendations for achieving high-quality data are made for readers to consider in the deployment of consumer wearables in research.

Mood Prediction of Patients With Mood Disorders by Machine Learning Using Passive Digital Phenotypes Based on the Circadian Rhythm: Prospective Observational Cohort Study

Background

Virtually, all organisms on Earth have their own circadian rhythm, and humans are no exception. Circadian rhythms are associated with various human states, especially mood disorders, and disturbance of the circadian rhythm is known to be very closely related. Attempts have also been made to derive clinical implications associated with mood disorders using the vast amounts of digital log that is acquired by digital technologies develop and using computational analysis techniques.

Objective

This study was conducted to evaluate the mood state or episode, activity, sleep, light exposure, and heart rate during a period of about 2 years by acquiring various digital log data through wearable devices and smartphone apps as well as conventional clinical assessments. We investigated a mood prediction algorithm developed with machine learning using passive data phenotypes based on circadian rhythms.

Methods

We performed a prospective observational cohort study on 55 patients with mood disorders (major depressive disorder [MDD] and bipolar disorder type 1 [BD I] and 2 [BD II]) for 2 years. A smartphone app for self-recording daily mood scores and detecting light exposure (using the installed sensor) were provided. From daily worn activity trackers, digital log data of activity, sleep, and heart rate were collected. Passive digital phenotypes were processed into 130 features based on circadian rhythms, and a mood prediction algorithm was developed by random forest.

Results

The mood state prediction accuracies for the next 3 days in all patients, MDD patients, BD I patients, and BD II patients were 65%, 65%, 64%, and 65% with 0.7, 0.69, 0.67, and 0.67 area under the curve (AUC) values, respectively. The accuracies of all patients for no episode (NE), depressive episode (DE), manic episode (ME), and hypomanic episode (HME) were 85.3%, 87%, 94%, and 91.2% with 0.87, 0.87, 0.958, and 0.912 AUC values, respectively. The prediction accuracy in BD II patients was distinctively balanced as high showing 82.6%, 74.4%, and 87.5% of accuracy (with generally good sensitivity and specificity) with 0.919, 0.868, and 0.949 AUC values for NE, DE, and HME, respectively.

Conclusions

On the basis of the theoretical basis of chronobiology, this study proposed a good model for future research by developing a mood prediction algorithm using machine learning by processing and reclassifying digital log data. In addition to academic value, it is expected that this study will be of practical help to improve the prognosis of patients with mood disorders by making it possible to apply actual clinical application owing to the rapid expansion of digital technology.

Modeling Biobehavioral Rhythms with Passive Sensing in the Wild

Biobehavioral rhythms are associated with numerous health and life outcomes. We study the feasibility of detecting rhythms in data that is passively collected from Fitbit devices and using the obtained model parameters to predict readmission risk after pancreatic surgery. We analyze data from 49 patients who were tracked before surgery, in hospital, and after discharge. Our analysis produces a model of individual patients' rhythms for each stage of treatment that is predictive of readmission. All of the rhythm-based models outperform the traditional approaches to readmission risk stratification that uses administrative data.

Validation of the Korean Version of the Biological Rhythms Interview of Assessment in Neuropsychiatry

OBJECTIVE

The Biological Rhythms Interview of Assessment in Neuropsychiatry (BRIAN) is a scale used to clinically evaluate disturbances in biological rhythm. In this study, we aimed to examine the reliability and validity of the Korean version of the BRIAN (K-BRIAN) in a Korean population.

METHODS

A total of 181 participants, including 141 outpatients with bipolar disorder (BD; type I, 62; type II, 79) and 40 controls, were recruited. Construct validity was tested by comparing the mean K-BRIAN scores of the BD patients and control subjects. Concurrent validity was tested by evaluating the association between the K-BRIAN and the Morningness-Eveningness Questionnaire (MEQ).

RESULTS

The mean K-BRIAN scores of the control subjects and patients with BD differed significantly (p<0.001). Particularly, the mean K-BRIAN score was considerably lower among control subjects (mean±standard deviation=35.00±8.88) than among patients with BD type I (41.19±12.10) and type II (50.18±13.73). The Cronbach's alpha for the K-BRIAN was 0.914. The K-BRIAN was found to correlate with the MEQ (r=-0.45, p<0.001).

CONCLUSION

The findings affirm that the K-BRIAN has good construct validity and internal consistency. This suggests that the K-BRIAN can be used to assess biological rhythms in the Korean population, especially for patients with mood disorder.