Philips Actiwatch Research Articles

Continuous glucose monitoring (CGM) for 308 older-age participants in an English birth cohort: variability and correlates

Introduction & BackgroundEpochs of hyperglycaemia and hypoglycaemia may each increase risk of common chronic diseases and impair both cognitive and physical function even in people without diabetes. Older people may have greater frequency of adverse glycaemic excursions, partly due to disordered autonomic function and sleep quality. Data for older, non-diabetic people are however scant. Objectives & Approach1) To describe blood glucose variability (completed) and 2) its socio-demographic and lifestyle correlates in a predominantly non-diabetic cohort of older adults (planned). Participants were recruited during 2021-2023 from an English birth cohort (the 1946 National Survey for Health and Development Study). They wore a continuous glucose monitor (Freestyle libre Abbott), which measured circulating glucose four times/hour, for seven days. Summary statistics and time outside range (4.4-7.8mmol/L) were calculated. Further information on glycaemic excursions and day-to-day variability will be gleaned using the R “iglu” package. For all CGM summary and excursion measures, future analyses will investigate: associations with HbA1c, socio-demographics, body composition, physical activity, diet and alcohol use. Results will be stratified by sleep/ wake time periods estimated from simultaneous actigraphy (Philips Actiwatch Spectrum Plus). Sensitivity analyses will exclude people taking hypo/ hyperglycaemic medications and those with diabetes. Relevance to Digital FootprintsDerived summary measures can be used by future studies to give insights into glycaemic variability as a population-level risk factor. This work will bring together multiple data sources, i.e. from CGM, actigraphy and baseline cohort data. ResultsParticipants were aged 75-76 years, 45% female and 10% had diagnosed diabetes; median (IQR) BMI was 26.8 (24.6-29.2) kg/m2. CGM data from 308 participants was collected, for a median (IQR) of 6.9 (6.7-7.6) days. Average glucose over the recording period was 5.7mmol/L (5.3-6.2mmol/L), standard deviation was 1.0mmol/L (0.8-1.3mmol/L), time outside range was 12.8% (6.2-24.7%) and 16% of participants spent ≥1 hour/day above and ≥1 hour/day below range. Conclusions & ImplicationsCGM was feasible for this cohort of older adults, and demonstrated high levels of time outside range for a predominantly non-diabetic group. Future analysis will determine whether enhanced characterisation of glycaemic variability is a potentially more accurate tool for predicting future disease risk than isolated glucose measurements.

Improving physical movement during stroke rehabilitation: investigating associations between sleep measured by wearable actigraphy technology, fatigue, and key biomarkers

BackgroundSleep disturbance and fatigue are common in individuals undergoing inpatient rehabilitation following stroke. Understanding the relationships between sleep, fatigue, motor performance, and key biomarkers of inflammation and neuroplasticity could provide valuable insight into stroke recovery, possibly leading to personalized rehabilitation strategies. This study aimed to investigate the influence of sleep quality on motor function following stroke utilizing wearable technology to obtain objective sleep measurements. Additionally, we aimed to determine if there were relationships between sleep, fatigue, and motor function. Lastly, the study aimed to determine if salivary biomarkers of stress, inflammation, and neuroplasticity were associated with motor function or fatigue post-stroke.MethodsEighteen individuals who experienced a stroke and were undergoing inpatient rehabilitation participated in a cross-sectional observational study. Following consent, participants completed questionnaires to assess sleep patterns, fatigue, and quality of life. Objective sleep was measured throughout one night using the wearable Philips Actiwatch. Upper limb motor performance was assessed on the following day and saliva was collected for biomarker analysis. Correlation analyses were performed to assess the relationships between variables.ResultsParticipants reported poor sleep quality, frequent awakenings, and difficulties falling asleep following stroke. We identified a significant negative relationship between fatigue severity and both sleep quality (r=-0.539, p = 0.021) and participants experience of awakening from sleep (r=-0.656, p = 0.003). A significant positive relationship was found between grip strength on the non-hemiplegic limb and salivary gene expression of Brain-derived Neurotrophic Factor (r = 0.606, p = 0.028), as well as a significant negative relationship between grip strength on the hemiplegic side and salivary gene expression of C-reactive Protein (r=-0.556, p = 0.048).ConclusionThe findings of this study emphasize the importance of considering sleep quality, fatigue, and biomarkers in stroke rehabilitation to optimize recovery and that interventions may need to be tailored to the individual. Future longitudinal studies are required to explore these relationships over time. Integrating wearable technology for sleep and biomarker analysis can enhance monitoring and prediction of outcomes following stroke, ultimately improving rehabilitation strategies and patient outcomes.

Comparative Analysis and Conversion Between Actiwatch and ActiGraph Open-Source Counts

Body-worn sensors have contributed to a rich and growing body of literature in public health and clinical research in the last decades. A major challenge in sensor research is the lack of consistency and standardization of the collection and reporting of the sensor data. The algorithms used to derive these activity counts can be vastly different between manufactures and not always transparent to the researchers. With Philips, one of the major research-grade wearable device manufacturers, discontinuing this product line, many researchers are left in need of alternative solutions and at the risk of not being able to relate their historical data using the Philips Actiwatch 2 devices to future findings with other devices. We herein provide a comparison analysis and conversion method that can be used to convert activity counts from Philips to those from ActiGraph, another major manufacturer who provide both raw acceleration data and count data based on their open-source algorithm to the research community. This work provides an approach to maximize the scientific value of historical actigraphy data collected by the Actiwatch devices to support research continuity in this community. The conversion, however, is not perfect and only offers an approximation, due to the intrinsic difference in the count algorithms between the two accelerometers, and the permanent information loss during data reduction. We encourage future research using body-worn sensors to retain the raw sensor data to ensure data consistency, comparability, and the ability to leverage future algorithm improvement.

Sleep and circadian measures at age 73 are related to brain macrostructure, microstructure, and amyloid pathology

AbstractBackgroundSleep disturbance is associated with an increased risk of developing Alzheimer’s disease, yet the relationship between sleep metrics and brain pathology remains poorly understood. Complementing ‘traditional’ metrics such as total sleep time (TST) and sleep efficiency (SE), the midpoint of sleep, interdaily stability and intradaily variability provide important insights into chronotype and circadian rest activity rhythms.MethodsWe used data from dementia‐free participants from the Insight 46 neuroscience sub‐study of the MRC National Survey of Health and Development (NSHD, British 1946 Birth Cohort). Hippocampal, whole brain, and white matter hyperintensity volume (WMHV), and amyloid status (positive/negative) were derived from volumetric T1 and FLAIR MRI, and 18F‐florbetapir amyloid‐PET at age 73. TST, SE, midpoint of sleep, interdaily stability, and intradaily variability were calculated from 6 nights of wrist actigraphy (Philips Actiwatch) at the same time point. Normal appearing white matter (NAWM) microstructure was assessed by diffusion weighted imaging at age 70. Mean z‐scores for each participant’s neurite density index (NDI) and orientation dispersion index (ODI) were generated relative to normal controls.The relationship between sleep and imaging metrics was examined using generalised additive models to allow a flexible functional form for the sleep metrics. Where there was no evidence of a nonlinear effect, results are presented from linear regression (continuous outcomes) or logistic regression (binary outcomes). All analyses were controlled for TST, SE, age, gender, socioeconomic status, intracranial volume, and interval from imaging to actigraphy, in addition to blood pressure age 53, and body mass index for WMHV analyses.ResultsThere was a significant nonlinear relationship between TST and both hippocampal and brain volume, with ∼ 7 hours sleep appearing optimal (Table 1, Figure 1). Increasing TST and decreasing SE were associated with decreased NDI. Unexpectedly, increased interdaily stability was associated with an increased risk of being amyloid positive (P = 0.003, Figure 2).ConclusionSleep and circadian measures at age 73 were associated with brain microstructure, macrostructure, and amyloid pathology. Efficient sleep of around 7 hours is associated with optimal brain health. Ongoing longitudinal follow‐up will help to explore causality.

35 Correlations Between Sleep and Cognitive Functioning in Healthy, Older Adults

Objective:Alzheimer’s disease (AD), a leading cause of dementia worldwide, affected an estimated 47 million people in 2015, placing a burden of over $1 trillion on health systems. Subclinical markers of AD pathology are seen many years before the clinical onset of dementia, suggesting that steps could be taken to prevent progression to disease in healthy individuals. Sleep optimizes cognition by creating a window of opportunity to consolidate memories, prune synaptic networks, and clear waste products. Studies that characterize the relationship between sleep and cognitive function prior to the onset of clinical AD could guide research into effective methods of delaying AD onset or preventing it altogether. The objective of our study is to describe how sleep quality and quantity correlate with performance on cognitive assessments within a healthy, aging population.Participants and Methods:Seventeen participants, between 62-82 years of age enrolled in an ongoing clinical trial assessing the effects of melatonin (5mg daily) versus placebo, were included in our study. Participants were observed over a 2-month period, during which no experimental interventions were administered. At study entry, participants underwent a comprehensive neuropsychological evaluation evaluating cognitive domains of attention, memory, speed of information processing, language, executive functioning, and mood. Afterwards, all participants wore a watch that measured actigraphy and light data (Philips Actiwatch Spectrum Pro actigraphy monitor) for 8 weeks to evaluate their sleep habits. Pearson and Spearman partial correlations were used to evaluate relationships between objective sleep parameters and baseline cognitive function test scores.Results:Aberrations of sleep length, sleep fragmentation, and daytime activity measures significantly correlated with cognitive performance on memory, language, visuospatial skills, and speed of processing tests (p = <0.05). Greater variability of awakenings at nighttime associated with better scores on memory tests but worse scores on language tests. Longer sleep times associated with worse language scores, while greater variability in daily activity correlated with poorer scores on visuospatial skills tests and speed of processing tests.Conclusions:This study establishes a framework for obtaining longitudinal sleep data in conjunction with serial cognitive function testing, encouraging further exploration into how sleep metrics affect specific domains of cognitive function. Findings suggest that having a less consistent sleep routine correlates with poorer cognitive function across multiple domains. The authors recommend broader analysis of actigraphy and cognitive function testing as objective measures of sleep and cognition in research and clinical practice.

0026 Novel home-based circadian phase assessment tool to enhance accessible quality care of circadian rhythm disorders- Circadia Study

Abstract Introduction Circadian rhythm sleep-wake disorders (CRSWDs), shifted timing of physiology with the external environments. CRSWDs, often undiagnosed, can increase risk of cardiometabolic and psychiatric disorders, and may have substantial financial and social impacts. Dim light melatonin onset assay (DLMO), the current gold-standard diagnostic test for CRSWDs is expensive/difficult. We aim to develop a more accessible melatonin assessment. Methods Our DLMO assay kit is low-cost, home-based and self-administered. It contains all necessary components for a home-based DLMO assay, including 18 time-stamped saliva samples, a Philips Actiwatch, and a lux meter. We tested the feasibility of the kit across two DLMO assessments, one-week apart, in CRSWDs patients (N=5) and controls (N=5). We concurrently assessed sleep with wrist actigraphy and sleep diary over four weeks through a self-guided study website. Results We enrolled participants with delayed (n=4) and advanced (n=1) sleep disorder and 5 control participants (mean 38.2 y.o., SD 11.66 y.o.; biological sex: 9 female, 1 male (control)). Participants were 100% compliant completing questionnaires and sleep logging with little intervention and no in-person guidance. 5/10 participants completed 100% of sleep diary days. Average duration of study involvement, from consent to study completion, was 42.4 days. Participants successfully set up their collection space and collected saliva samples, remaining compliant to objectively-measured study protocols: 86.67% of samples were taken under sufficiently low light, and 82% were taken within five minutes of scheduled collection. We obtained and averaged two DLMO times for 7/10 participants (delayed sleep phase disorder (DPSD): 12:05 AM, controls: 9:56 PM). DLMO times were on average 3:18 hours earlier than self-reported sleep times. Conclusion Our pilot study, with both CRSWDs patients (N=5) and controls (N=5), found the at-home kits feasible and reproducible (5/6 participants had their DLMO times within an hour of each other). Using the kit, we confirmed their previous DSPD diagnoses and correctly classified controls through melatonin samples. We used objective measures (Actigraph) to verify the testing procedure was followed. These procedures will be tested on a larger scale in the Circadia Study (circadiastudy.org), set to launch in February 2023. Support (if any) This work was supported by the National Institutes of Health [DHHS] (grant number 1R35GM146839-01).

Differences in sleep patterns between patients with anorexia nervosa and healthy controls: a cross-sectional study

BackgroundSleep difficulties are common in patients with anorexia nervosa (AN), but objective assessments have mostly been performed in hospital and laboratory settings. We aimed to identify differences in sleep patterns between patients with AN and healthy controls (HC) in their free-living environments, and potential associations between sleep patterns and clinical symptoms in patients with AN.MethodsThis cross-sectional study analyzed 20 patients with AN prior to them starting outpatient treatment and 23 HC. Sleep patterns were measured objectively using an accelerometer (Philips Actiwatch 2) for 7 consecutive days. Average sleep onset, sleep offset, total sleep time, sleep efficiency, wake after sleep onset (WASO) and mid-sleep awakenings lasting ≥ 5 min were compared between patients with AN and HC using nonparametric statistical analyses. Associations of sleep patterns with body mass index, eating-disorder symptoms, eating-disorder-associated impairment, and symptoms of depression were assessed in the patient group.ResultsCompared with HC, patients with AN had shorter WASO [median (interquartile range(IQR)): 33 vs. 42 min], but a longer average duration of mid-sleep awakenings lasting ≥ 5 min [median (IQR): 9 vs. 6 min, p = 0.006] and had more nights with no sleep (six nights in four patients with AN vs. zero nights in HC). There were no differences between patients with AN and HC regarding other sleep parameters and no significant correlations between sleep patterns and clinical parameters in patients with AN. However, HC presented a Intraindividual variability pattern that was closer to a normal distribution, whereas patients with AN tended to either have very regular or large variability in sleep onset time (AN; n = 7 < 25th percentile and n = 8 > 75th percentile vs. HC; n = 4 < 25 percentile and n = 3 > 75th percentile) during the week of sleep recordings.ConclusionPatients with AN seem to spend more time awake during the night and have more nights without sleep than do HC, even though their average weekly sleep duration did not differ from that in HC. The intraindividual variability in sleep pattern seems to be an important parameter that should be assessed when studying sleep in patients with AN.Trial registration ClinicalTroals.gov. Identifier: NCT02745067. Registered: April 20, 2016.

Measuring Sleep Quality in the Hospital Environment with Wearable and Non-Wearable Devices in Adults with Stroke Undergoing Inpatient Rehabilitation

Sleep disturbances are common after stroke and may affect recovery and rehabilitation outcomes. Sleep monitoring in the hospital environment is not routine practice yet may offer insight into how the hospital environment influences post-stroke sleep quality while also enabling us to investigate the relationships between sleep quality and neuroplasticity, physical activity, fatigue levels, and recovery of functional independence while undergoing rehabilitation. Commonly used sleep monitoring devices can be expensive, which limits their use in clinical settings. Therefore, there is a need for low-cost methods to monitor sleep quality in hospital settings. This study compared a commonly used actigraphy sleep monitoring device with a low-cost commercial device. Eighteen adults with stroke wore the Philips Actiwatch to monitor sleep latency, sleep time, number of awakenings, time spent awake, and sleep efficiency. A sub-sample (n = 6) slept with the Withings Sleep Analyzer in situ, recording the same sleep parameters. Intraclass correlation coefficients and Bland-Altman plots indicated poor agreement between the devices. Usability issues and inconsistencies were reported between the objectively measured sleep parameters recorded by the Withings device compared with the Philips Actiwatch. While these findings suggest that low-cost devices are not suitable for use in a hospital environment, further investigations in larger cohorts of adults with stroke are needed to examine the utility and accuracy of off-the-shelf low-cost devices to monitor sleep quality in the hospital environment.

A comparison of agreement between actigraphy and polysomnography for assessing sleep during posttraumatic amnesia.

Sleep disturbance often emerges in the early recovery phase following a moderate to severe traumatic brain injury, known as posttraumatic amnesia. Actigraphy is commonly employed to assess sleep, as it is assumed that patients in posttraumatic amnesia (who display confusion, restlessness, and agitation) would better tolerate this measure over gold-standard polysomnography (PSG). This study evaluated the agreement between PSG and actigraphy for determining (sleep/wake time, sleep efficiency, sleep latency, and awakenings) in patients experiencing posttraumatic amnesia. It also compared the epoch-by-epoch sensitivity, specificity, and accuracy between the Actigraph device's 4 wake threshold settings (low, medium, high, and automatic) to PSG. The sample consisted of 24 inpatients recruited from a traumatic brain injury inpatient rehabilitation unit. Ambulatory PSG was recorded overnight at bedside and a Philips Actiwatch was secured to each patient's wrist for the same period. There were poor correlations between PSG and actigraphy for all parameters (Lin's concordance correlation coefficient = < 0.80). The low threshold displayed the highest correlation with PSG for wake and sleep time, albeit still low. Actigraphy displayed low specificity (ranging from 17.1% to 36.6%). There appears to be a greater disparity between actigraphy and PSG for patients with increased wake time. Actigraphy, while convenient, demonstrated poorer performance in determining sleep-wake parameters in patients with significantly disturbed sleep. Ambulatory PSG can provide a clearer understanding of the extent of sleep disturbances in these patients with reduced mobility during early rehabilitation. Study findings can help design future protocols of sleep assessment during posttraumatic amnesia and optimize treatment. Fedele B, McKenzie D, Williams G, Giles R, Olver J. A comparison of agreement between actigraphy and polysomnography for assessing sleep during posttraumatic amnesia. J Clin Sleep Med. 2022;18(11):2605-2616.

Direct assessment of overnight parent-child proximity in children with behavioral insomnia: Extending models of operant and classical conditioning

ABSTRACT Introduction Explanatory models of behavioral insomnia typically draw on operant learning theory with behavioral techniques focused on altering parent-child interactions to improve sleep. However, there are no data describing parent-child interactions overnight beyond parent report. In this study we used radio frequency identification technology to quantify parent-child proximity overnight in two groups at elevated risk of behavioral insomnia, Angelman syndrome (AS) and Smith-Magenis syndrome (SMS). Materials and Methods Nineteen children aged 4–15 years (8 with AS, 11 with SMS) participated in a week-long at-home assessment of sleep and overnight parent-child proximity. Sleep parameters were recorded using the Philips Actiwatch 2 and proximity data were recorded using custom-built radio frequency identification watches. Results Three patterns of proximity data between parent-child dyads overnight were evident: “checking” (six with AS, five with SMS), “co-sleeping” (four with SMS) and those who had “no proximity” overnight (two with AS, two with SMS). In the AS group, 25.45% of actigraphy-defined wakes resulted in a parent-child interaction. In the SMS group, 39.34% of wakes resulted in a parent-child interaction. Children who interacted with their parents when settling to sleep were not significantly more likely to interact at waking. Discussion The novel application of radio frequency identification technology is a feasible method for studying overnight parent-child proximity. Profiles of proximity between participants that are not closely aligned with operant models of behavioral insomnia were evident. These results have significant implications for the etiology of poor sleep and the application of behavioral sleep interventions.

0090 Performance of a Multisensor Ring to Evaluate Sleep: In-Lab Evaluation Relative to PSG and Actigraphy: Importance of Generalized Versus Personalized Scoring

Abstract Introduction Multisensor sleep wearable devices have demonstrated utility for research and relative accuracy for discerning sleep-wake patterns at home and in the laboratory. Additional sensors and more complex scoring algorithms may improve the ability of wearables to assess sleep health. Methods Thirty-six healthy adults completed assessment while wearing the experimental device (Happy Ring), as well as Philips Actiwatch, Fitbit, Oura, and Whoop devices. Evaluations occurred in the laboratory (Alice 6 polysomnogram). The Happy Ring includes sensors for accelerometry, photoplethysmography, electrodermal activity, and skin temperature. Epoch-by-epoch analyses compared the Happy Ring to lab polysomnography, as well as other sleep-tracking devices. Scoring was accomplished using two machine-learning-derived algorithms: a “generalized” algorithm which was static and applied to all users (like those used for other devices) and a “personalized” algorithm where parameters are personalized, dynamic, and change based on data collected across different parts of the night of sleep. Results Compared to in-lab polysomnography, the generalized algorithm using data from the Happy Ring demonstrated good sensitivity (94%) and specificity (70%). The personalized algorithm also performed well with good sensitivity (93%) and specificity (83%). Other devices also demonstrated good sensitivity, ranging from 89% (Fitbit) to 94% (Actiwatch); specificity however, was more variable, ranging from 19% (Actiwatch) to 54% (Whoop). Overall accuracy was 91% for generalized and 92% for personalized, compared to 88% for Oura, 86% for Whoop, 84% for Fitbit, and 85% for Actiwatch. Measurement of sleep stage accuracy was 67%, 85%, and 85% for light, deep, and REM sleep, respectively, for the Happy generalized algorithm. For the Happy personalized algorithm, accuracy for sleep stages were 81%, 95%, and 92%, for light, deep and REM sleep, respectively. Post-hoc analyses showed that the Happy personalized algorithm demonstrated better specificity than all other modalities (p&amp;lt;0.001). Kappa scores were 0.45 for generalized and 0.68 for personalized, compared to 0.32 for the Oura Ring, 0.32 for Whoop Strap, and 0.37 for Fitbit wristband. Conclusion The multisensory Happy ring demonstrated good sensitivity and specificity for the detection of sleep in the laboratory. The personalized approach outperformed all others, representing a potential innovation for improving detection accuracy. Support (If Any) Dr. Grandner is supported by R01DA051321 and R01MD011600. This work was supported by Happy Health, Inc.

0101 Performance of a Multisensor Ring to Evaluate Sleep At-Home Relative to PSG and Actigraphy: Importance of Generalized Versus Personalized Scoring

Abstract Introduction Multisensor sleep wearable devices have demonstrated utility for research and relative accuracy for discerning sleep-wake patterns at home and in the laboratory. Additional sensors and more complex scoring algorithms may improve the ability of wearables to assess sleep health. Methods Thirty-six healthy adults completed assessment while wearing the experimental device (Happy Ring), as well as Philips Actiwatch, Fitbit, Oura, and Whoop devices. Evaluations at home were conducted using the Dreem headband as an at-home polysomnography reference. The experimental Happy Ring device includes accelerometry, photoplethysmography, electrodermal activity, and skin temperature. Epoch-by-epoch analyses compared the Happy Ring to home polysomnography, as well as other sleep-tracking wearable devices. Scoring was accomplished using two machine-learning-derived algorithms: a “generalized” algorithm, similar to that used in other devices, which was static and applied to all users, and a “personalized” algorithm where parameters are personalized, dynamic, and change based on data collected across different parts of the night of sleep. Results Compared to home polysomnography, the Happy generalized algorithm demonstrated good sensitivity (94%) and specificity (67%), and the Happy personalized algorithm also performed well (93% and 75%, respectively). Other devices demonstrated good sensitivity, ranging from 91% (Whoop) to 96% (Oura). However, specificity was more variable, ranging from 41% (Actiwatch) to 60% (Fitbit). Overall accuracy using the Happy Ring was 91% for generalized and 92% for personalized algorithms, compared to 92% for Oura, 89% for Whoop, 89% for Fitbit, and 89% for Actiwatch. Regarding sleep stages, accuracy for the Happy Ring was 66%, 83%, and 78% for light, deep, and REM sleep, respectively, for the generalized algorithm. For the personalized algorithm accuracy was 78%, 92%, and 95%, for light, deep and REM sleep, respectively. Post-hoc analyses showed that the Happy personalized algorithm demonstrated better specificity than all other modalities (p&amp;lt;0.001). Kappa scores were 0.42 for generalized and 0.60 for personalized, compared to 0.45 for Oura, 0.47 for Whoop, and 0.48 for Fitbit. Conclusion The multisensory Happy ring demonstrated good sensitivity and specificity for the detection of sleep at home. The personalized approach outperformed all others, representing a potential innovation for improving detection accuracy. Support (If Any) Dr. Grandner is supported by R01DA051321 and R01MD011600. This work was supported by Happy Health, Inc.

The Influence of Self‐Reported and Actigraphy‐Derived Sleep Quality on Cardiometabolic Health in a Diverse Cohort of Healthy Adults

PurposeInsufficient sleep is associated with adverse cardiometabolic health (CMH) including high blood pressure (BP) and insulin resistance, which increases cardiovascular disease risk. However, there are limited data on insufficient sleep and CMH in healthy adults, particularly in diverse cohorts. Thus, we sought to examine associations between actigraphy‐derived sleep, self‐reported sleep quality, and CMH in healthy adults.Methods99 individuals (50 female, age: 26 ± 12yrs, BMI: 25.5 ± 4.1 kg/m2, resting BP: 108±11/ 65 ± 8 mmHg, mean ± SD) were included in this analysis. We assessed objective sleep quality, including sleep duration and efficiency using wrist‐worn Philips Actiwatch Spectrum PLUS monitors (8.4 ± 2.7 days). Self‐reported sleep was characterized using the Pittsburgh Sleep Quality Index (0 [better] to 21 [worse]) and an average weekly score to the question “how rested do you feel” (0 [better] to 4 [worse]). We combined all four sleep variables to create a multidimensional sleep score (0 [better] to 4 [worse]). Following ≥10‐minutes of supine rest, we assessed oscillometric brachial BP in triplicate. We assessed estimated central BP with pulse wave analysis and arterial stiffness via carotid‐femoral pulse wave velocity (cf‐PWV) with SphygmoCor. We evaluated endothelial function using brachial artery flow mediated dilation (FMD) via ultrasound. We are collected venous blood samples to calculate homeostatic model assessment of insulin resistance (HOMA‐IR) via blood glucose and insulin concentration. We performed stepwise regressions controlling for sex, race, age, and BMI (model 1) and with the addition of individual sleep metrics (model 2) or the multidimensional sleep score (model 3) to determine whether sleep was associated with measures of CMH.ResultsAdding independent sleep variables or composite sleep score did not improve model fit for brachial BP (e.g., systolic BP Model 1: R2= 0.301, p &lt; 0.001, Model 3: R2= 0.262, p&lt;0.001). The addition of sleep variables improved model fit for central BP (e.g., systolic BP Model 1: R2= 0.280, p &lt; 0.001, Model 3: R2= 0.306, p &lt;0.001) and cf‐PWV (Model 1: R2= 0.153 p &lt; 0.001, Model 2: R2= 0.159, p = .014). Furthermore, the multidimensional sleep score was independently associated with central systolic BP (β± SE = ‐2.41 ± 0.961, p = 0.015) and central diastolic BP (β± SE = ‐1.97 ± 0.91, p = 0.035). Finally, the addition of the multidimensional sleep score improved model fit for FMD (Model 1: R2=0.113, p = 0.012, Model 3: R2= 0.132, p = 0.023) and HOMA‐IR (Model 1: R2= 0.229, p&lt;0.001, Model 3: R2= 0.246, p &lt;0.001).ConclusionOur preliminary data indicate that multidimensional sleep is independently associated with central BP and the addition of individual or multidimensional sleep adds upon the predictive value of sex, race age and BMI in predicting CMH in a diverse cohort of healthy adults.

P133 Backward Rotating Shifts are Associated with Real-Time Drowsiness During Daytime Drives in Heavy Vehicle Drivers

Abstract Purpose While 10–20% of heavy vehicle crashes (HVDs) are drowsiness-related, the contributions of subsequent shifts to chronic drowsiness in HVDs is largely unknown. Eye-blink parameters indicate driver drowsiness reliably. This study examined the association of consecutive shifts and real-time drowsiness in HVDs. Methods Habitual sleep-wake of HVDs (all males, aged 49.5 ± 8 years) was monitored objectively (Philips Actiwatch, N=15) for 5 weeks (5.75± 1.4 hours). Johns Drowsiness Score (JDS, a composite eye-blink parameter in one-min intervals) was monitored for 4 weeks in HVDs (N=14) using an infrared oculography (Optalert, Melbourne, Australia) device. We assessed the association of drowsiness events (JDS equal or larger than 2.6) with consecutive shift types via mixed linear regression models. Results Eigth consecutive shifts increased drowsiness by 1.06 times compared to 2 shifts (8.37 events/h vs 6.77 events/h, P= 0.03). Consecutive shift sequences included afternoons (9%), mornings (29%), nights (5%), mixed rotating shifts (28%), forward-rotating shifts (11%) and backward-rotating shifts (12%). Drowsiness event rates were 1.23 times greater during night consecutive shifts relative to afternoon shifts (8.37 events/h vs 6.67 events/h, P= 0.03). Backward-rotating shifts (morning-night-evening- afternoon) elevated daytime drowsiness between 10 am and 3 pm by 1.55 times (10.01 events/h vs 6.47 events/h, P= 0.016). Conclusions Regardless of the number of consecutive shifts, sequential night shifts increase real-time drowsiness in HVDs, with backward rotating shifts resulting in higher rates of drowsiness events during daytime. The interaction of schedule features should inform the work scheduling of HVDs to reduce the risk of drowsiness.

Clinical Utility of Wearable Sensors and Patient-Reported Surveys in Patients With Schizophrenia: Noninterventional, Observational Study.

BackgroundRelapse in schizophrenia may be preceded by early warning signs of biological, sensory, and clinical status. Early detection of warning signs may facilitate intervention and prevent relapses.ObjectiveThis study aims to investigate the feasibility of using wearable devices and self-reported technologies to identify symptom exacerbation correlates and relapse in patients with schizophrenia.MethodsIn this observational study, patients with schizophrenia were provided with remote sensing devices to continuously monitor activity (Garmin vivofit) and sleep (Philips Actiwatch), and smartphones were used to record patient-reported outcomes. Clinical assessments of symptoms (Positive and Negative Syndrome Scale and Brief Psychiatric Rating Scale) were performed biweekly, and other clinical scales on symptoms (Clinical Global Impression-Schizophrenia, Calgary Depression Scale), psychosocial functioning, physical activity (Yale Physical Activity Survey), and sleep (Pittsburgh Sleep Quality Index) were assessed every 4 weeks. Patients were observed for 4 months, and correlations between clinical assessments and aggregated device metrics data were assessed using a mixed-effect model. An elastic net model was used to predict the clinical symptoms based on the device features.ResultsOf the 40 patients enrolled, 1 patient relapsed after being stable with evaluable postbaseline data. Weekly patient-reported outcomes were moderately correlated with psychiatric symptoms (Brief Psychiatric Rating Scale total score, r=0.29; Calgary Depression Scale total score, r=0.37; and Positive and Negative Syndrome Scale total score, r=0.3). In the elastic net model, sleep and activity features derived from Philips Actigraph and Garmin vivofit were predictive of the sitting index of the Yale Physical Activity Survey and sleep duration component of the Pittsburgh Sleep Quality Index. On the basis of the combined patient data, a high percentage of data coverage and compliance (>80%) was observed for each device.ConclusionsThis study demonstrated that wearable devices and smartphones could be effectively deployed and potentially used to monitor patients with schizophrenia. Furthermore, metrics-based prediction models can assist in detecting earlier signs of symptom changes. The operational learnings from this study may provide insights to conduct future studies.Trial RegistrationClinicalTrials.gov NCT02224430; https://www.clinicaltrials.gov/ct2/show/NCT02224430

A Temporal Threshold for Distinguishing Off-Wrist from Inactivity Periods: A Retrospective Actigraphy Analysis.

(1) Background. To facilitate accurate actigraphy data analysis, inactive periods have to be distinguished from periods during which the device is not being worn. The current analysis investigates the degree to which off-wrist and inactive periods can be automatically identified. (2) Methods. In total, 125 actigraphy records were manually scored for ‘off-wrist’ and ‘inactivity’ (99 collected with the Motionlogger AMI, 26 (sampling frequency of 60 (n = 20) and 120 (n = 6) s) with the Philips Actiwatch 2.) Data were plotted with cumulative frequency percentage and analyzed with receiver operating characteristic curves. To confirm findings, the thresholds determined in a subset of the Motionlogger dataset (n = 74) were tested in the remaining dataset (n = 25). (3) Results. Inactivity data lasted shorter than off-wrist periods, with 95% of inactive events being shorter than 11 min (Motionlogger), 20 min (Actiwatch 2; 60 s epochs) or 30 min (Actiwatch 2; 120 s epochs), correctly identifying 35, 92 or 66% of the off-wrist periods. The optimal accurate detection of both inactive and off-wrist periods for the Motionlogger was 3 min (Youden’s Index (J) = 0.37), while it was 18 (J = 0.89) and 16 min (J = 0.81) for the Actiwatch 2 (60 and 120 s epochs, respectively). The thresholds as determined in the subset of the Motionlogger dataset showed similar results in the remaining dataset. (4) Conclusion. Off-wrist periods can be automatically identified from inactivity data based on a temporal threshold. Depending on the goal of the analysis, a threshold can be chosen to favor inactivity data’s inclusion or accurate off-wrist detection.

PSG Validation of minute-to-minute scoring for sleep and wake periods in a consumer wearable device.

BackgroundActigraphs are wrist-worn devices that record tri-axial accelerometry data used clinically and in research studies. The expense of research-grade actigraphs, however, limit their widespread adoption, especially in clinical settings. Tri-axial accelerometer-based consumer wearable devices have gained worldwide popularity and hold potential for a cost-effective alternative. The lack of independent validation of minute-to-minute accelerometer data with polysomnographic data or even research-grade actigraphs, as well as access to raw data has hindered the utility and acceptance of consumer-grade actigraphs.MethodsSleep clinic patients wore a consumer-grade wearable (Huami Arc) on their non-dominant wrist while undergoing an overnight polysomnography (PSG) study. The sample was split into two, 20 in a training group and 21 in a testing group. In addition to the Arc, the testing group also wore a research-grade actigraph (Philips Actiwatch Spectrum). Sleep was scored for each 60-s epoch on both devices using the Cole-Kripke algorithm.ResultsBased on analysis of our training group, Arc and PSG data were aligned best when a threshold of 10 units was used to examine the Arc data. Using this threshold value in our testing group, the Arc has an accuracy of 90.3%±4.3%, sleep sensitivity (or wake specificity) of 95.5%±3.5%, and sleep specificity (wake sensitivity) of 55.6%±22.7%. Compared to PSG, Actiwatch has an accuracy of 88.7%±4.5%, sleep sensitivity of 92.6%±5.2%, and sleep specificity of 60.5%±20.2%, comparable to that observed in the Arc.ConclusionsAn optimized sleep/wake threshold value was identified for a consumer-grade wearable Arc trained by PSG data. By applying this sleep/wake threshold value for Arc generated accelerometer data, when compared to PSG, sleep and wake estimates were adequate and comparable to those generated by a clinical-grade actigraph. As with other actigraphs, sleep specificity plateaus due to limitations in distinguishing wake without movement from sleep. Further studies are needed to evaluate the Arc’s ability to differentiate between sleep and wake using other sources of data available from the Arc, such as high resolution accelerometry and photoplethysmography.

Practice parameters for the use of actigraphy in the military operational context: the Walter Reed Army Institute of Research Operational Research Kit-Actigraphy (WORK-A)

BackgroundThe Walter Reed Army Institute of Research (WRAIR) Operational Research Kit-Actigraphy (WORK-A) is a set of unique practice parameters and actigraphy-derived measures for the analysis of operational military sleep patterns. The WORK-A draws on best practices from the literature and comprises 15 additional descriptive variables. Here, we demonstrate the WORK-A with a sample of United States Army Reserve Officers’ Training Corps (ROTC) cadets (n = 286) during a month-long capstone pre-commissioning training exercise.MethodsThe sleep of ROTC cadets (n = 286) was measured by Philips Actiwatch devices during the 31-day training exercise. The preliminary effectiveness of the WORK-A was tested by comparing differences in sleep measures collected by Actiwatches as calculated by Philips Actiware software against WORK-A-determined sleep measures and self-report sleep collected from a subset of ROTC cadets (n = 140).ResultsActiware sleep summary statistics were significantly different from WORK-A measures and self-report sleep (all P ≤ 0.001). Bedtimes and waketimes as determined by WORK-A major sleep intervals showed the best agreement with self-report bedtime (22:21 ± 1:30 vs. 22:13 ± 0:40, P = 0.21) and waketime (04:30 ± 2:17 vs. 04:31 ± 0:47, P = 0.68). Though still significantly different, the discrepancy was smaller between the WORK-A measure of time in bed (TIB) for major sleep intervals (352 ± 29 min) and self-report nightly sleep duration (337 ± 57 min, P = 0.006) than that between the WORK-A major TIB and Actiware TIB (177 ± 42, P ≤ 0.001).ConclusionsDefault actigraphy methods are not the most accurate methods for characterizing soldier sleep, but reliable methods for characterizing operational sleep patterns is a necessary first step in developing strategies to improve soldier readiness. The WORK-A addresses this knowledge gap by providing practice parameters and a robust variety of measures with which to profile sleep behavior in service members.

0286 Schedule Characteristics of Heavy Vehicle Drivers are Associated with Eye-Blink Indicators of Real-Time Drowsiness on the Road

Abstract Introduction While up to 52% of heavy vehicle crashes are drowsiness-related, the contributions of schedule factors to real-time objective drowsiness in heavy vehicle drivers (HVDs) have not been studied. Eye-blink parameters are a reliable indicator of driver drowsiness. This study aimed to examine the relationship between work-related factors and objective drowsiness in HVDs. Methods HVDs (all males, aged 49.5 ± 8 years) undertook 5- weeks of sleep-wake monitoring (Philips Actiwatch, N=15), and 4-weeks of infrared oculography (Optalert, Melbourne, Australia) to monitor their eye-blink parameters (averaged each minute) while driving their own vehicle (N=12). Participants slept for 5.75± 1.4 hours before the drives. Drowsiness events were defined as any Johns Drowsiness Scores (JDS) scores larger than 2.6 based on prior research. The relationships of schedule factors and drowsiness events per hour were assessed via mixed linear regression models. Results Drowsiness event rates were 3–5 times greater between 22:00 and 03:00 hours compared to between 16:00 and 17:00 hours (17- 25 events/h vs 5 events/h, P= 0.0001 to 0.007). The frequency of drowsiness events at night varied with shift start time and time into shift (P= 0.0001 to 0.001). Compared to the first hour of driving, drowsiness event rates rose significantly during the 13th to the 21st hours into the shift (13- 59 events/h vs 5.5 events/h, P= 0.0001 to 0.007). During sequential night shifts drowsiness events were 1.8 times more common compared to 1–3 sequential day shifts (9 events/h vs 5 events/h, P= 0.012 to 0.019). Conclusion Driving at night, for more than 12 hours and sequential night shifts increase real-time drowsiness in HVDs, with these factors interacting resulting in even higher rates of drowsiness events. Longitudinal studies in larger populations will further define how these factors interact to inform the work scheduling of HVDs to reduce the risk of drowsiness. Support This research was supported by the CRC for Alertness, Safety and Productivity.

1201 Performance Evaluation Of A Novel Contactless Breathing Monitor And Machine Learning Algorithm For Sleep Stage Classification In A Healthy Population

Abstract Introduction Although polysomnography (PSG) remains the gold standard for sleep assessment in a lab setting, non-EEG signals such as respiration and motion are directly affected by sleep stages and can be used for sleep stage prediction. Importantly, these signals can be obtained in a low-cost and unobtrusive manner, allowing for large scale and longitudinal data collection in a home environment. The Circadia C100 System (FDA 510(k) clearance expected Q1 2020) is a novel ‘nearable’ device that uses radar for contactless monitoring of respiration and motion. The current study aims to validate the performance of the associated sleep analysis algorithm. Methods A total of 41 nights of sleep data were recorded from 33 healthy participants using the device, alongside PSG. Data were recorded both in a sleep lab and home environment. PSG data were scored by RPSGT-certified technicians. Respiration and movement features were extracted, and machine learning algorithms were developed to perform sleep stage classification and predict sleep metrics. Algorithms were trained and validated on PSG data using cross-validation. Results An epoch-by-epoch true positive rate of 56.2%, 79.4%, 55.5% and 72.6% was found for ‘Wake’, ‘REM’, ‘Light’ and ‘Deep’ respectively. No statistical differences in performance were found between home-recorded and lab-recorded contactless data. Mean absolute error of total sleep time (TST), wake after sleep onset (WASO), and sleep efficiency (SE) was 13.2 minutes, 11.3 minutes and 3%, respectively. The contactless monitor was found to outperform both medical grade and clinical grade actigraphy based devices: The Philips Actiwatch Spectrum Plus and the Fitbit Alta HR. Conclusion Current results are encouraging and suggest that the contactless monitor could be used for long-term sleep assessment and continuous evaluation of sleep therapy outcomes. Further clinical validation work is ongoing in subjects diagnosed with sleep disorders such as obstructive sleep apnea. Support -