ABSTRACT
BACKGROUND: Disturbed sleep is frequently identified in adult patients with cancer and their caregivers, with detrimental impact on physical health. Less known is the extent to which self-reported and actigraph-measured sleep patterns are similar between patients and their sleep-partner caregivers, and how these different modes of sleep measurements are related to physical health. METHODS: Patients diagnosed with colorectal cancer and their sleep-partner caregivers (81 dyads) completed a questionnaire for physical functioning and collected saliva samples for seven consecutive days, from which cortisol slope was quantified. Additionally, participants completed a daily sleep diary and wore actigraph for 14 consecutive days, from which sleep duration, sleep onset latency (SOL), and duration of wake after sleep onset (WASO) were calculated. RESULTS: Participants reported sleep patterns that fell within or close to the optimal range, which were similar between patients and their caregivers. Self-reported and actigraph-measured sleep duration had moderate levels of agreement (ICC = 0.604), whereas SOL and WASO had poor agreement (ICC = 0.269). Among patients, longer self-reported WASO was associated with poorer physical health and flatter cortisol slope (p ≤ 0.013). Among caregivers, longer self-reported SOL was associated with poorer physical functioning, actigraph-measured WASO was associated with steeper cortisol slope, and longer self-reported sleep markers studied than actigraph-measured were associated with poorer physical functioning (p ≤ 0.042). CONCLUSION: Findings suggest that employing multiple assessment modes for sleep and physical health is vital for comprehensive understanding of sleep health. Furthermore, when addressing patients' sleep health, it may be beneficial to include their sleep-partner caregivers who may experience similar disturbed sleep.
Subject(s)
Actigraphy , Caregivers , Colorectal Neoplasms , Hydrocortisone , Saliva , Self Report , Sleep , Humans , Caregivers/psychology , Male , Female , Colorectal Neoplasms/psychology , Middle Aged , Aged , Hydrocortisone/analysis , Hydrocortisone/metabolism , Sleep/physiology , Saliva/chemistry , Surveys and Questionnaires , Health Status , Adult , Sleep Wake Disorders/psychology , Diaries as Topic , Sleep QualityABSTRACT
BACKGROUND: Sleep is impaired in children with attention-deficit/hyperactivity disorder (ADHD). However, population-based examination of indicators of sleep insufficiency and bedtime irregularity is limited. This investigation examined associations between ADHD, weeknight sleep insufficiency, and bedtime irregularity in a nationally-representative child sample, and indicators of these sleep outcomes in ADHD. METHODS: Parents of children aged 3-17 years with ADHD (n = 7671) were surveyed through the 2020-2021 National Survey of Children's Health. Inverse probability of treatment weighting generated a weighted matched control sample (n = 51,572). Weighted generalized linear models were performed without and with age-stratification to examine associations between ADHD and sleep, adjusting for sociodemographics in the full sample, and between nineteen sociodemographic and clinical variables and sleep in ADHD. RESULTS: Having ADHD was associated with increased odds of sleep insufficiency and bedtime irregularity relative to controls, even after adjusting for sociodemographic variables. In ADHD, older age was associated with lower sleep insufficiency and greater bedtime irregularity. Black race, increased poverty, higher ADHD severity, depression, and increased screen time were associated with greater sleep insufficiency and bedtime irregularity. Adverse childhood experiences (ACEs) were associated with greater sleep insufficiency. Behavioral/conduct problems, female sex, and absence of both ADHD medication use and ASD diagnosis were associated with poorer bedtime irregularity. Age-stratified results are reported in text. CONCLUSIONS: Children with ADHD face heightened risk for insufficient sleep and irregular bedtimes. Findings suggest intervention targets (e.g., Black race, poverty, depression, screen time) to improve both sleep insufficiency and bedtime irregularity. Results highlight ACEs and behavioral/conduct problems as targets to improve sleep insufficiency and bedtime regularity, respectively. Age-stratified findings are discussed.
ABSTRACT
Mental health is independently influenced by the inclination to sleep at specific times (chronotype) and the actual sleep timing (behavior). Chronotype and timing of actual sleep are, however, often misaligned. This study aims to determine how chronotype, sleep timing, and the alignment between the two impact mental health. In a community-dwelling cohort of middle- and older-aged adults (UK Biobank, n = 73,888), we examined the impact of chronotype (questionnaire-based), the timing of behavior (determined with 7-day accelerometry), and the alignment between the two on mental, behavioral, neurodevelopmental disorders (MBN), depression, and anxiety, as assessed through ICD-10 codes. As compared to morning types with early behavior (aligned), morning types with late behavior (misaligned) had an increased risk of having MBN, depression, and anxiety (p's<0.001). As compared to evening-types with late behavior (aligned), however, evening-types with early behavior (misaligned) had a decreased risk of depression (p < 0.01), with a trend for MBN (p = 0.04) and anxiety (p = 0.05). Longitudinal analyses, in which the likelihood of developing de novo mental health disorders was associated with chronotype, behavioral timing, and alignment between the two, confirmed cross-sectional findings. To age healthily, individuals should start sleeping before 1AM, despite chronobiological preferences.
Subject(s)
Anxiety , Circadian Rhythm , Depression , Independent Living , Sleep , Humans , Male , Female , Middle Aged , Aged , Circadian Rhythm/physiology , Sleep/physiology , Mental Health , Adult , Time Factors , Cohort Studies , AccelerometryABSTRACT
OBJECTIVE: To achieve consensus on whether screen-based digital media (1) in general, (2) via prebedtime content, and (3) via prebedtime light impairs sleep health in (a) childhood, (b) adolescence, and (c) adulthood. Furthermore, to address whether employing behavioral strategies and interventions may reduce the potential negative effects of screens on sleep health. METHODS: The National Sleep Foundation convened a 16-person multidisciplinary expert panel ("Panel"). Panelists met virtually 5 times throughout 2023, during which they followed a modified Delphi RAND/UCLA Appropriateness Method to reach consensus. RESULTS: The Panel conducted a literature review starting with 2209 articles, narrowed down to 522 relevant empirical articles and 52 relevant review articles. The search was refined to include 35 experimental/intervention studies that examined whether there was a causal link between screen-based digital media and sleep. In addition, panelists reviewed 5 recent relevant systematic review articles. After reviewing the summarized current literature, panelists voted on 10 candidate statements about whether screen use impairs sleep health. The Panel met virtually to discuss the results of the first round of votes, which was then followed by a second round of voting, ultimately achieving consensus on 5 out of the 10 statements. CONCLUSIONS: The Panel achieved consensus that (1) in general, screen use impairs sleep health among children and adolescents, (2) the content of screen use before sleep impairs sleep health of children and adolescents, and (3) behavioral strategies and interventions may attenuate the negative effects of screen use on sleep health.
Subject(s)
Screen Time , Humans , Adolescent , Child , Sleep , Consensus , Foundations , Adult , LongevityABSTRACT
Obstructive sleep apnea (OSA) is a common, underdiagnosed sleep-related breathing disorder with serious health implications Objective - We propose a deep transfer learning approach for sleep stage classification and sleep apnea (SA) detection using wrist-worn consumer sleep technologies (CST). Methods - Our model is based on a deep convolutional neural network (DNN) utilizing accelerometers and photo-plethysmography signals from nocturnal recordings. The DNN was trained and tested on internal datasets that include raw data from clinical and wrist-worn devices; external validation was performed on a hold-out test dataset containing raw data from a wrist-worn CST. Results - Training on clinical data improves performance significantly, and feature enrichment through a sleep stage stream gives only minor improvements. Raw data input outperforms feature-based input in CST datasets. The system generalizes well but performs slightly worse on wearable device data compared to clinical data. However, it excels in detecting events during REM sleep and is associated with arousal and oxygen desaturation. We found; cases that were significantly underestimated were characterized by fewer of such event associations. Conclusion - This study showcases the potential of using CSTs as alternate screening solution for undiagnosed cases of OSA. Significance - This work is significant for its development of a deep transfer learning approach using wrist-worn consumer sleep technologies, offering comprehensive validation for data utilization, and learning techniques, ultimately improving sleep apnea detection across diverse devices.
Subject(s)
Deep Learning , Polysomnography , Signal Processing, Computer-Assisted , Sleep Stages , Wearable Electronic Devices , Humans , Polysomnography/instrumentation , Polysomnography/methods , Sleep Stages/physiology , Male , Wrist , Adult , Middle Aged , Female , Sleep Apnea Syndromes/diagnosis , Sleep Apnea Syndromes/physiopathology , Aged , Accelerometry/instrumentation , Accelerometry/methodsABSTRACT
BACKGROUND: Sleep-wake regulating circuits are affected during prodromal stages in the pathological progression of both Alzheimer's disease (AD) and Parkinson's disease (PD), and this disturbance can be measured passively using wearable devices. Our objective was to determine whether accelerometer-based measures of 24-h activity are associated with subsequent development of AD, PD, and cognitive decline. METHODS: This study obtained UK Biobank data from 82,829 individuals with wrist-worn accelerometer data aged 40 to 79 years with a mean (± SD) follow-up of 6.8 (± 0.9) years. Outcomes were accelerometer-derived measures of 24-h activity (derived by cosinor, nonparametric, and functional principal component methods), incident AD and PD diagnosis (obtained through hospitalization or primary care records), and prospective longitudinal cognitive testing. RESULTS: One hundred eighty-seven individuals progressed to AD and 265 to PD. Interdaily stability (a measure of regularity, hazard ratio [HR] per SD increase 1.25, 95% confidence interval [CI] 1.05-1.48), diurnal amplitude (HR 0.79, CI 0.65-0.96), mesor (mean activity; HR 0.77, CI 0.59-0.998), and activity during most active 10 h (HR 0.75, CI 0.61-0.94), were associated with risk of AD. Diurnal amplitude (HR 0.28, CI 0.23-0.34), mesor (HR 0.13, CI 0.10-0.16), activity during least active 5 h (HR 0.24, CI 0.08-0.69), and activity during most active 10 h (HR 0.20, CI 0.16-0.25) were associated with risk of PD. Several measures were additionally predictive of longitudinal cognitive test performance. CONCLUSIONS: In this community-based longitudinal study, accelerometer-derived metrics were associated with elevated risk of AD, PD, and accelerated cognitive decline. These findings suggest 24-h rhythm integrity, as measured by affordable, non-invasive wearable devices, may serve as a scalable early marker of neurodegenerative disease.
Subject(s)
Alzheimer Disease , Cognitive Dysfunction , Neurodegenerative Diseases , Parkinson Disease , Humans , Parkinson Disease/epidemiology , Alzheimer Disease/epidemiology , Alzheimer Disease/complications , Longitudinal Studies , Prospective Studies , Cognitive Dysfunction/psychologyABSTRACT
Introduction: High variability in response and retention rates for posttraumatic stress disorder (PTSD) treatment highlights the need to identify "personalized" or "precision" medicine factors that can inform optimal intervention selection before an individual commences treatment. In secondary analyses from a non-inferiority randomized controlled trial, behavioral and physiological emotion regulation were examined as non-specific predictors (that identify which individuals are more likely to respond to treatment, regardless of treatment type) and treatment moderators (that identify which treatment works best for whom) of PTSD outcome. Methods: There were 85 US Veterans with clinically significant PTSD symptoms randomized to 6 weeks of either cognitive processing therapy (CPT; n = 44) or a breathing-based yoga practice (Sudarshan kriya yoga; SKY; n = 41). Baseline self-reported emotion regulation (Difficulties in Emotion Regulation Scale) and heart rate variability (HRV) were assessed prior to treatment, and self-reported PTSD symptoms were assessed at baseline, end-of-treatment, 1-month follow-up, and 1-year follow-up. Results: Greater baseline deficit in self-reported emotional awareness (similar to alexithymia) predicted better overall PTSD improvement in both the short- and long-term, following either CPT or SKY. High self-reported levels of emotional response non-acceptance were associated with better PTSD treatment response with CPT than with SKY. However, all significant HRV indices were stronger moderators than all self-reported emotion regulation scales, both in the short- and long-term. Veterans with lower baseline HRV had better PTSD treatment response with SKY, whereas Veterans with higher or average-to-high baseline HRV had better PTSD treatment response with CPT. Conclusions: To our knowledge, this is the first study to examine both self-reported emotion regulation and HRV, within the same study, as both non-specific predictors and moderators of PTSD treatment outcome. Veterans with poorer autonomic regulation prior to treatment had better PTSD outcome with a yoga-based intervention, whereas those with better autonomic regulation did better with a trauma-focused psychological therapy. Findings show potential for the use of HRV in clinical practice to personalize PTSD treatment. Clinical trial registration: ClinicalTrials.gov identifier, NCT02366403.
ABSTRACT
OBJECTIVE: Sleepiness and fatigue are common complaints among individuals with sleep disorders. The two concepts are often used interchangeably, causing difficulty with differential diagnosis and treatment decisions. The current study investigated sleep disorder patients to determine which factors best differentiated sleepiness from fatigue. METHODS: The study used a subset of participants from a multi-site study (n = 606), using a cross-sectional study design. We selected 60 variables associated with either sleepiness or fatigue, including demographic, mental health, and lifestyle factors, medical history, sleep questionnaires, rest-activity rhythms (actigraphy), polysomnographic (PSG) variables, and sleep diaries. Fatigue was measured with the Fatigue Severity Scale and sleepiness was measured with the Epworth Sleepiness Scale. A Random Forest machine learning approach was utilized for analysis. RESULTS: Participants' average age was 47.5 years (SD 14.0), 54.6% female, and the most common sleep disorder diagnosis was obstructive sleep apnea (67.4%). Sleepiness and fatigue were moderately correlated (r = 0.334). The model for fatigue (explained variance 49.5%) indicated depression was the strongest predictor (relative explained variance 42.7%), followed by insomnia severity (12.3%). The model for sleepiness (explained variance 17.9%), indicated insomnia symptoms was the strongest predictor (relative explained variance 17.6%). A post hoc receiver operating characteristic analysis indicated depression could be used to discriminate fatigue (AUC = 0.856) but not sleepiness (AUC = 0.643). CONCLUSIONS: The moderate correlation between fatigue and sleepiness supports previous literature that the two concepts are overlapping yet distinct. Importantly, depression played a more prominent role in characterizing fatigue than sleepiness, suggesting depression could be used to differentiate the two concepts.
Subject(s)
Disorders of Excessive Somnolence , Sleep Initiation and Maintenance Disorders , Sleep Wake Disorders , Humans , Female , Middle Aged , Male , Cross-Sectional Studies , Sleep Initiation and Maintenance Disorders/diagnosis , Sleep Initiation and Maintenance Disorders/complications , Sleepiness , Fatigue/diagnosis , Fatigue/etiology , Sleep Wake Disorders/complications , Surveys and Questionnaires , Disorders of Excessive Somnolence/diagnosisABSTRACT
OBJECTIVES: To explore how the blood plasma proteome fluctuates across the 24-hour day and identify a subset of proteins that show endogenous circadian rhythmicity. METHODS: Plasma samples from 17 healthy adults were collected hourly under controlled conditions designed to unmask endogenous circadian rhythmicity; in a subset of 8 participants, we also collected samples across a day on a typical sleep-wake schedule. A total of 6916 proteins were analyzed from each sample using the SomaScan aptamer-based multiplexed platform. We used differential rhythmicity analysis based on a cosinor model with mixed effects to identify a subset of proteins that showed circadian rhythmicity in their abundance. RESULTS: One thousand and sixty-three (15%) proteins exhibited significant daily rhythmicity. Of those, 431 (6.2%) proteins displayed consistent endogenous circadian rhythms on both a sleep-wake schedule and under controlled conditions: it included both known and novel proteins. When models were fitted with two harmonics, an additional 259 (3.7%) proteins exhibited significant endogenous circadian rhythmicity, indicating that some rhythmic proteins cannot be solely captured by a simple sinusoidal model. Overall, we found that the largest number of proteins had their peak levels in the late afternoon/evening, with another smaller group peaking in the early morning. CONCLUSIONS: This study reveals that hundreds of plasma proteins exhibit endogenous circadian rhythmicity in humans. Future analyses will likely reveal novel physiological pathways regulated by circadian clocks and pave the way for improved diagnosis and treatment for patients with circadian disorders and other pathologies. It will also advance efforts to include knowledge about time-of-day, thereby incorporating circadian medicine into personalized medicine.
ABSTRACT
Isolated rapid-eye-movement (REM) sleep behavior disorder (iRBD) is caused by motor disinhibition during REM sleep and is a strong early predictor of Parkinson's disease. However, screening questionnaires for iRBD lack specificity due to other sleep disorders that mimic the symptoms. Nocturnal wrist actigraphy has shown promise in detecting iRBD by measuring sleep-related motor activity, but it relies on sleep diary-defined sleep periods, which are not always available. Our aim was to precisely detect iRBD using actigraphy alone by combining two actigraphy-based markers of iRBD - abnormal nighttime activity and 24-hour rhythm disruption. In a sample of 42 iRBD patients and 42 controls (21 clinical controls with other sleep disorders and 21 community controls) from the Stanford Sleep Clinic, the nighttime actigraphy model was optimized using automated detection of sleep periods. Using a subset of 38 iRBD patients with daytime data and 110 age-, sex-, and body-mass-index-matched controls from the UK Biobank, the 24-hour rhythm actigraphy model was optimized. Both nighttime and 24-hour rhythm features were found to distinguish iRBD from controls. To improve the accuracy of iRBD detection, we fused the nighttime and 24-hour rhythm disruption classifiers using logistic regression, which achieved a sensitivity of 78.9%, a specificity of 96.4%, and an AUC of 0.954. This study preliminarily validates a fully automated method for detecting iRBD using actigraphy in a general population.Clinical relevance- Actigraphy-based iRBD detection has potential for large-scale screening of iRBD in the general population.
Subject(s)
Parkinson Disease , REM Sleep Behavior Disorder , Humans , Actigraphy , REM Sleep Behavior Disorder/diagnosis , Parkinson Disease/diagnosis , Sleep, REM , Surveys and QuestionnairesABSTRACT
In humans, exposure to continuous light is typically used to change the timing of the circadian clock. This study examines the efficiency of a sequence of light flashes ("flash therapy") applied during sleep to shift the clock. Healthy participants (n = 10) took part in two 36-h laboratory stays, receiving a placebo (goggles, no light) during one visit and the intervention (goggles, 2-ms flashes broad-spectrum light for 60 min, delivered every 15 s, starting 30 min after habitual sleep onset) during the other. Circadian phase shift was assessed with changes in salivary dim light melatonin onset (DLMO). Sleep, measured with polysomnography, was analyzed to assess changes in sleep architecture and spectral power. After 1 h of flashes, DLMO showed a substantial delay (1.13 ± 1.27 h) compared to placebo (12 ± 20 min). Two individuals exhibited very large shifts of 6.4 and 3.1 h. There were no substantive differences in sleep architecture, but some evidence for greater instability in sleep. 1 h of flash therapy during sleep evokes large changes in circadian timing, up to 6 h, and does so with only minimal, if any, impact on sleep. Flash therapy may offer a practical option to delay the circadian clock in shift workers and jet travelers.
Subject(s)
Melatonin , Retinal Diseases , Humans , Light , Phototherapy , Polysomnography , SleepABSTRACT
STUDY OBJECTIVES: This study evaluated the effects of early time-restricted eating (eTRE) on shifting the timing of sleep among late sleepers. Primary outcomes included actigraphy- and sleep diary-derived sleep onset, midsleep phase, and wake time with total sleep time as a secondary outcome. METHODS: Fifteen healthy adults with habitual late sleep timing were randomized to receive either eTRE or sleep and nutrition hygiene (control) via a single 30-minute synchronous video session. Participants completed an initial 1-week baseline phase followed by a 2-week intervention phase. Measures included continuous sleep monitoring and sleep and nutrition diaries. RESULTS: Linear mixed-effects modeling demonstrated that eTRE significantly advanced sleep timing compared with controls. Self-reported sleep onset (56.1 [95% confidence interval: 20.5, 91.7] minutes), midpoint (19.5 [7.2, 31.9] minutes), and offset (42.2 [2.9, 81.5] minutes) each moved earlier in eTRE as compared with controls. Similarly, objectively determined sleep onset (66.5 [29.6, 103.4] minutes), midpoint (21.9 [9.1, 34.7] minutes), and offset (39.3 [1.3, 77.3] minutes) each moved earlier in eTRE as compared with controls. Total sleep time showed a nonsignificant increase in the eTRE group as compared with controls. CONCLUSIONS: Late sleepers who were instructed in a single session about eTRE significantly advanced their sleep timing, especially sleep onset. eTRE shows potential as a clinical strategy for advancing sleep timing in late sleepers. CLINICAL TRIAL REGISTRATION: Registry: Chinese Clinical Trial Registry; Name: FAST Asleep: It's All About Timing; URL: https://www.chictr.org.cn/showproj.html?proj=122504; Identifier: ChiCTR2100043691. CITATION: Blum DJ, Hernandez B, Zeitzer JM. Early time-restricted eating advances sleep in late sleepers: a pilot randomized controlled trial. J Clin Sleep Med. 2023;19(12):2097-2106.
Subject(s)
Sleep Wake Disorders , Sleep , Adult , Humans , Pilot Projects , Time , ActigraphyABSTRACT
PURPOSE: Circadian rest-Activity Rhythm Disorders (CARDs) are common in patients with cancer, particularly in advanced disease. CARDs are associated with increased symptom burden, poorer quality of life, and shorter survival. Research and reporting practices lack standardization, and formal diagnostic criteria do not exist. This electronic Delphi (e-Delphi) study aimed to formulate international recommendations for the assessment and diagnosis of CARDs in patients with cancer. METHODS: An international e-Delphi was performed using an online platform (Welphi). Round 1 developed statements regarding circadian rest-activity rhythms, diagnostic criteria, and assessment techniques. Rounds 2 and 3 involved participants rating their level of agreement with the statements and providing comments until consensus (defined internally as 67%) and stability between rounds were achieved. Recommendations were then created and distributed to participants for comments before being finalized. RESULTS: Sixteen participants from nine different clinical specialties and seven different countries, with 5-35 years of relevant research experience, were recruited, and thirteen participants completed all three rounds. Of the 164 generated statements, 66% achieved consensus, and responses were stable between the final two rounds. CONCLUSIONS: The e-Delphi resulted in international recommendations for assessing and diagnosing CARDs in patients with cancer. These recommendations should ensure standardized research and reporting practices in future studies.
ABSTRACT
Little is known about links of circadian rhythm alterations with neuropsychiatric symptoms and cognition in memory impaired older adults. Associations of actigraphic rest/activity rhythms (RAR) with depressive symptoms and cognition are examined using function-on-scalar regression (FOSR). Forty-four older adults with memory impairment (mean: 76.84 ± 8.15 years; 40.9% female) completed 6.37 ± 0.93 days of actigraphy, the Beck depression inventory-II (BDI-II), mini-mental state examination (MMSE) and consortium to establish a registry for Alzheimer's disease (CERAD) delayed word recall. FOSR models with BDI-II, MMSE, or CERAD as individual predictors adjusted for demographics (Models A1-A3) and all three predictors and demographics (Model B). In Model B, higher BDI-II scores are associated with greater activity from 12:00-11:50 a.m., 2:10-5:50 p.m., 8:40-9:40 p.m., 11:20-12:00 a.m., higher CERAD scores with greater activity from 9:20-10:00 p.m., and higher MMSE scores with greater activity from 5:50-10:50 a.m. and 12:40-5:00 p.m. Greater depressive symptomatology is associated with greater activity in midafternoon, evening, and overnight into midday; better delayed recall with greater late evening activity; and higher global cognitive performance with greater morning and afternoon activity (Model B). Time-of-day specific RAR alterations may affect mood and cognitive performance in this population.
Subject(s)
Alzheimer Disease , Cognition , Humans , Female , Male , Aged , Neuropsychological Tests , Circadian Rhythm , Memory Disorders/diagnosisABSTRACT
Controlled breathwork practices have emerged as potential tools for stress management and well-being. Here, we report a remote, randomized, controlled study (NCT05304000) of three different daily 5-min breathwork exercises compared with an equivalent period of mindfulness meditation over 1 month. The breathing conditions are (1) cyclic sighing, which emphasizes prolonged exhalations; (2) box breathing, which is equal duration of inhalations, breath retentions, and exhalations; and (3) cyclic hyperventilation with retention, with longer inhalations and shorter exhalations. The primary endpoints are improvement in mood and anxiety as well as reduced physiological arousal (respiratory rate, heart rate, and heart rate variability). Using a mixed-effects model, we show that breathwork, especially the exhale-focused cyclic sighing, produces greater improvement in mood (p < 0.05) and reduction in respiratory rate (p < 0.05) compared with mindfulness meditation. Daily 5-min cyclic sighing has promise as an effective stress management exercise.
Subject(s)
Meditation , Humans , Affect , Anxiety/therapy , Respiration , ArousalABSTRACT
Wrist-worn consumer sleep technologies (CST) that contain accelerometers (ACC) and photoplethysmography (PPG) are increasingly common and hold great potential to function as out-of-clinic (OOC) sleep monitoring systems. However, very few validation studies exist because raw data from CSTs are rarely made accessible for external use. We present a deep neural network (DNN) with a strong temporal core, inspired by U-Net, that can process multivariate time series inputs with different dimensionality to predict sleep stages (wake, light-, deep-, and REM sleep) using ACC and PPG signals from nocturnal recordings. The DNN was trained and tested on 3 internal datasets, comprising raw data both from clinical and wrist-worn devices from 301 recordings (PSG-PPG: 266, Wrist-worn PPG: 35). External validation was performed on a hold-out test dataset containing 35 recordings comprising only raw data from a wrist-worn CST. An accuracy = 0.71 ± 0.09, 0.76 ± 0.07, 0.73 ± 0.06, and κ = 0.58 ± 0.13, 0.64 ± 0.09, 0.59 ± 0.09 was achieved on the internal test sets. Our experiments show that spectral preprocessing yields superior performance when compared to surrogate-, feature-, raw data-based preparation. Combining both modalities produce the overall best performance, although PPG proved to be the most impactful and was the only modality capable of detecting REM sleep well. Including ACC improved model precision to wake and sleep metric estimation. Increasing input segment size improved performance consistently; the best performance was achieved using 1024 epochs (â¼8.5 hrs.). An accuracy = 0.69 ± 0.13 and κ = 0.58 ± 0.18 was achieved on the hold-out test dataset, proving the generalizability and robustness of our approach to raw data collected with a wrist-worn CST.
Subject(s)
Deep Learning , Photoplethysmography , Sleep , Sleep Stages , Accelerometry , Heart RateABSTRACT
BACKGROUND: Older men with the worse alignment of activity and light may have lower levels of cognition and increased rates of cognitive decline. METHODS: This cohort consisted of 1 036 older men (81.1 ± 4.6 years) from the MrOS Sleep Study (2009-2012). Light and activity levels were gathered by wrist actigraphy. Phasor analysis was used to quantify the alignment of light-dark and rest-activity patterns (magnitude) and their temporal relationship (angle). Global cognitive function (Modified Mini-Mental State examination [3MS]) and executive function (Trails B test) were measured, then repeated 4.2 ± 0.8 years later. Linear regression models examined the associations of phasor magnitude and angle with cognition and cognitive decline. Models were adjusted for age, clinic, race, education, and season. RESULTS: Smaller phasor magnitude (worse aligned light and activity patterns) was associated with lower initial level and increased decline in executive function. Compared to those with higher phasor magnitude, those with lower magnitude took an average of 11.1 seconds longer to complete the Trails B test (quartile 1 vs quartile 4, p = .02). After follow-up, Trails B completion time increased an average of 5.5 seconds per standard deviation decrease in phasor magnitude (95% confidence interval [CI] 0.7-10.4, p = .03). There were no associations with phasor angle, and none with magnitude and global cognition (3MS). CONCLUSION: Among older men, worse alignment of light and activity patterns was associated with worse initial performance and increased decline in executive function, but not related to global cognition. Interventions that improve the alignment of light and activity may slow cognitive decline in older adults.
Subject(s)
Cognition Disorders , Cognitive Dysfunction , Male , Humans , Aged , Cognitive Dysfunction/etiology , Polysomnography , Cognition , SleepABSTRACT
Unlike light input for forming images, non-image-forming retinal pathways are optimized to convey information about the total light environment, integrating this information over time and space. In a variety of species, discontinuous light sequences (flashes) can be effective stimuli, notably impacting circadian entrainment. In this study, we examined the extent to which this temporal integration can occur. A group of healthy, young (n = 20) individuals took part in a series of 16-day protocols in which we examined the impact of different lengths of light flash sequences on circadian timing. We find a significant phase change of -0.70 h in response to flashes that did not differ by duration; a 15-min sequence could engender as much change in circadian timing as 3.5-h sequences. Acute suppression of melatonin was also observed during short (15-min) exposures, but not in exposures over one hour in length. Our data are consistent with the theory that responses to light flashes are mediated by the extrinsic, rod/cone pathway, and saturate the response of this pathway within 15 min. Further excitation leads to no greater change in circadian timing and an inability to acutely suppress melatonin, indicating that this pathway may be in a refractory state following this brief light stimulation.