Physical Activity Researcher · paresearcher

/Highlights/ Interesting Ideas How to Analyse Sleep and Physical Activity Data - Dr Christina Reynolds (Pt3)

·14 min·1 clip

Dr. Reynolds warns against getting lost in data without first asking what questions you want to answer.

Host Oli Tikkanen interviews Dr. Christina Reynolds, a researcher with an astrophysics background now applying data analysis to sleep and physical activity studies. This episode focuses on practical methodologies for analyzing long-term activity and sleep data collected from wearable devices. Dr. Reynolds works with the SHARP Lab at Portland VA Hospital and Oregon Health & Science University under Dr. Miranda Lim. She advocates for using intraday stability and intraday variability as key metrics for analyzing patterns of active versus rest states. Reynolds references a specific study using a Taiwan-made accelerometer that classifies sitting, standing, walking, and cycling, paired with a bed sensor for sleep, collecting three months of data. She suggests integrating external datasets like local weather information and length of daylight into such analyses. The conversation explores analyzing the impact of a bad night's sleep on subsequent recovery cycles and activity patterns. A case study involves workers on a demanding schedule of three consecutive 12-hour shifts followed by four days off, examining their health struggles. Reynolds proposes calculating daily intraday stability and variability scores to visualize how the rigid shift days contrast with recovery days. She highlights the research question of determining the optimal sleep strategy on the first recovery day to hasten recuperation. An interesting insight is that actigraphy data, while not perfect for measuring absolute sleep, still provides valuable long-term behavioral pattern data unobtrusively. Reynolds challenges common advice, like avoiding intense exercise before bed, by suggesting data can quantify the actual impact of activity timing on sleep quality. She emphasizes the importance of defining clear research questions before diving into a large dataset to avoid unproductive analytical rabbit holes. The discussion reveals how circadian rhythm research can benefit from examining not just the amount, but the timing of physical activity. Dr. Reynolds recommends interdisciplinary collaboration, specifically naming mathematicians, sociologists, and historians as valuable partners for innovation. She recounts a positive past collaboration with a mathematician who brought exciting new analytical perspectives to physiological data. The tone is relaxed, conversational, and educational, mirroring the podcast's aim to discuss research practicalities outside formal publications. Listeners interested in wearable device data, sleep research methodology, or shift work health impacts will find this episode valuable. Researchers seeking inspiration for interdisciplinary collaboration will appreciate Reynolds's perspective. Those looking for definitive conclusions or narrative storytelling might find the technical, methodological focus less engaging.

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Physical Activity Researcher