Wearable Light-and-Motion Dataloggers for Sleep/Wake Research: A Review

Performance evaluation of a machine learning-based methodology using dynamical features to detect nonwear intervals in actigraphy data in a free-living setting

GOAL AND AIMS

One challenge using wearable sensors is nonwear time. Without a nonwear (e.g., capacitive) sensor, actigraphy data quality can be biased by subjective determinations confounding sleep/wake classification. We developed and evaluated a machine learning algorithm supplemented by dynamic features to discern wear/nonwear episodes.

FOCUS TECHNOLOGY

Actigraphy data from wrist actigraph (Spectrum, Philips-Respironics).

REFERENCE TECHNOLOGY

The built-in nonwear sensor as "ground truth" to classify nonwear periods using other data, mimicking features of Actiwatch 2.

SAMPLE

Data were collected over 1week from employed adults (n = 853).

DESIGN

Extreme gradient boosting (XGBoost), a tree-based classifier algorithm, was used to classify wear/nonwear, supplemented by dynamic features calculated over various time windows.

CORE ANALYTICS

The performance of the proposed algorithm was tested over 30-second epochs. Additional analytics and exploratory analyses: Evaluation of the SHapley Additive exPlanations (SHAP) values to find the effectiveness of the dynamic features.

CORE OUTCOMES

The XGBoost classifier yielded substantial improvements in balanced accuracy, sensitivity, and specificity, including dynamic features and comparison to default actiwatch classification algorithms.

IMPORTANT SUPPLEMENTAL OUTCOMES

The proposed classifier effectively distinguished between valid and invalid days, and the duration of contiguous periods of nonwear correctly identified.

CORE CONCLUSION

Our findings highlight the potential of XGBoost using dynamic features of varying activity levels across the time series to provide insights on wear/nonwear classification using a large dataset. The methodology provides an alternative to laborious manual benchmarking of the data for similar devices that do not have a nonwear sensor.

Light Environment of Arctic Solstices is Coupled With Melatonin Phase-Amplitude Changes and Decline of Metabolic Health

Light environment in the Arctic differs widely with the seasons. Studies of relationships between objectively measured circadian phase and amplitude of light exposure and melatonin in community-dwelling Arctic residents are lacking. This investigation combines cross-sectional (n = 24-62) and longitudinal (n = 13-27) data from week-long actigraphy (with light sensor), 24-h salivary melatonin profiles, and proxies of metabolic health. Data were collected within the same week bracketing spring equinox (SE), and winter/summer solstices (WS/SS). Drastic seasonal differences in blue light exposure (BLE) corresponded to seasonal changes in the 24-h pattern of melatonin, which was phase delayed and reduced in normalized amplitude (NA) during WS/SS compared to SE. The extent of individual melatonin's acrophase and Dim Light Melatonin Onset (DLMO) change from SE to WS correlated with that from SE to SS. Although similar in extent and direction, melatonin phase changes versus SE were linked to morning BLE deficit in WS, contrasting to evening BLE excess in SS. Seasonal changes in sleep characteristics were closely associated with changes in the phases of BLE and melatonin. Proxies of metabolic health included triglycerides (TG), high-density lipoprotein cholesterol (HDL), TG/HDL ratio, and cortisol. Adverse seasonal changes in these proxies were associated with delayed acrophases of BLE and melatonin during WS and SS. TG and TG/HDL were higher in WS and SS than in SE, and cross-sectionally correlated with later melatonin and BLE acrophases, while lower HDL was associated with later BLE onset and later melatonin acrophase. Overall, this study shows that optimal 24-h patterns of light exposure during SE is associated with an earlier acrophase and a larger 24-h amplitude of melatonin, and that both features are linked to better metabolic health. Improving light hygiene, in particular correcting winter morning light deficit and summer evening light excess may help maintain metabolic health at high latitudes. Novel solutions for introducing proper circadian light hygiene such as human-centric light technologies should be investigated to address these issues in future studies.

Advancements in Flexible Sensors for Monitoring Body Movements during Sleep: A Review

Sleep plays a role in maintaining our physical well-being. However, sleep-related issues impact millions of people globally. Accurate monitoring of sleep is vital for identifying and addressing these problems. While traditional methods like polysomnography (PSG) are commonly used in settings, they may not fully capture natural sleep patterns at home. Moreover, PSG equipment can disrupt sleep quality. In recent years, there has been growing interest in the use of sensors for sleep monitoring. These lightweight sensors can be easily integrated into textiles or wearable devices using technology. The flexible sensors can be designed for skin contact to offer continuous monitoring without being obtrusive in a home environment. This review presents an overview of the advancements made in flexible sensors for tracking body movements during sleep, which focus on their principles, mechanisms, and strategies for improved flexibility, practical applications, and future trends.

Blue Light and Temperature Actigraphy Measures Predicting Metabolic Health Are Linked to Melatonin Receptor Polymorphism

This study explores the relationship between the light features of the Arctic spring equinox and circadian rhythms, sleep and metabolic health. Residents (N = 62) provided week-long actigraphy measures, including light exposure, which were related to body mass index (BMI), leptin and cortisol. Lower wrist temperature (wT) and higher evening blue light exposure (BLE), expressed as a novel index, the nocturnal excess index (NEIbl), were the most sensitive actigraphy measures associated with BMI. A higher BMI was linked to nocturnal BLE within distinct time windows. These associations were present specifically in carriers of the MTNR1B rs10830963 G-allele. A larger wake-after-sleep onset (WASO), smaller 24 h amplitude and earlier phase of the activity rhythm were associated with higher leptin. Higher cortisol was associated with an earlier M10 onset of BLE and with our other novel index, the Daylight Deficit Index of blue light, DDIbl. We also found sex-, age- and population-dependent differences in the parametric and non-parametric indices of BLE, wT and physical activity, while there were no differences in any sleep characteristics. Overall, this study determined sensitive actigraphy markers of light exposure and wT predictive of metabolic health and showed that these markers are linked to melatonin receptor polymorphism.

Special Issue on “Research on Circadian Rhythms in Health and Disease”

Despite rigorous investigation of circadian rhythms in humans and animal models in the past, basic chronobiologic principles have not yet entered clinical practice [...]

Availability of Physical Activity Tracking Data from Wearable Devices for Glaucoma Patients

Physical activity has been found to potentially modulate glaucoma risk, but the evidence remains inconclusive. The increasing use of wearable physical activity trackers may provide longitudinal and granular data suitable to address this issue, but little is known regarding the characteristics and availability of these data sources. We performed a scoping review and query of data sources on the availability of wearable physical activity data for glaucoma patients. Literature databases (PubMed and MEDLINE) were reviewed with search terms consisting of those related to physical activity trackers and those related to glaucoma, and we evaluated results at the intersection of these two groups. Biomedical databases were also reviewed, for which we completed database queries. We identified eight data sources containing physical activity tracking data for glaucoma, with two being large national databases (UK BioBank and All of Us) and six from individual journal articles providing participant-level information. The number of glaucoma patients with physical activity tracking data available, types of glaucoma-related data, fitness devices utilized, and diversity of participants varied across all sources. Overall, there were limited analyses of these data, suggesting the need for additional research to further investigate how physical activity may alter glaucoma risk.