Monitoring Sleep and Glucose Among University Students

This study aims to characterise associations between day-to-day sleep, activity, meal schedules, well-being and continuous glucose profiles in a cohort of free-living healthy, young adults. Multi-day data will be collected using wearables and smartphone-based measures in field settings.

There are two iterations of this study. In the first iteration (METWI1), wearables and smartphone-based measures are used to characterise free-living sleep, activity, meal schedules, well-being and continuous glucose profiles in a cohort of healthy, young Chinese university students for 4 weeks during the normal school term. While undergoing glucose monitoring (2 weeks), participants consume a standardised meal plan catered by the laboratory to reduce added variance from dietary intake. Examining relationships between sleep and behavioural characteristics and glucose profiles may contribute to the identification of phenotypes at higher risk of developing metabolic disorders. Data collected in this study may furthermore aid the identification of changes in sleep patterns associated with closer proximity to academic assessments, when students are predicted to experience increased academic workload and stress. Delays and more irregularity in sleep timing, shorter sleep durations and reduced sleep quality are expected closer to assessment dates. These in turn are predicted to result in higher glucose levels and glycemic variability. In the second iteration (METWI2), in addition to the above measures, participants undergo an oral glucose tolerance test following a night of moderate sleep restriction and baseline sleep (without sleep restriction). This allows us to examine effects of moderate, at-home sleep restriction on glucose tolerance and insulin sensitivity. In terms of sleep monitoring, we additionally aim to validate passive WiFi sensing against measurement of sleep using a commercial sleep and activity tracker (Oura ring), smartphone touchscreen interactions (tappigraphy-based sleep estimation) and sleep diary logs in students who are residing in dormitories. Studying this sample affords a convenient, and privacy protecting way of obtaining WiFi data. This can contribute to establishing whether a combination of multiple data sources for sleep detection can improve accuracy of sleep detection, incorporating the influence of device usage in the peri-sleep period. The secondary goal of this sleep study is the triangulation of sleep detection techniques for long term sleep monitoring on university campus. The hope is to access a larger population of students to infer sleep behaviours and sleep health, and eventually, to develop interventions to improve population health using individualised sleep data.