Using Physiological Big Data to Predict Cross Country Performance

Abstract

Sleep quality and heart rate variability are hypothesized in research to be indicators of improved or impaired athletic performance. This is especially relevant for the sport of endurance running. Very little research has been done on the associations between these physiological variables and endurance running performance in race conditions. The purpose of this study was to investigate whether continuously collected physiological data using wearable devices can be used to predict endurance running performance. Sleep and heart rate data from members of the Harvard men’s varsity cross country team was collected using the WHOOP performance optimization system during their season. This study observed that sleep duration had significant associations with endurance performance, especially two nights before a race; the idea that two nights before a race is the most important night of sleep is always inferred but has never been supported by research. Specifically, it was observed that increased REM sleep duration had significant associations with better cross country performance, while increased sleep latency had a significant association with worse performance. Further research suggests that REM sleep duration could be a marker for consistent sleep patterns. No significant associations were observed between HRV and race performance. It was concluded that physiological data collected from wearable devices can be used to indicate better or worse endurance performance.