Author(s) :

Niranjan Uday.

Volume/Issue :

Abstract :

Sleep is widely recognized as an important factor in physical health and recovery, yet the amount of sleep needed to support optimal physiological performance remains unclear. This study examined whether nightly sleep duration predicts exercise-related performance measures using a large wearablebased archival dataset consisting of over 10,000 adult participants. Single variable statistical tests, including Pearson correlations, were used alongside multivariable modeling through a neural network to evaluate relationships between sleep duration and performance indicators such as resting heart rate, workout duration, burned calories, steps taken, workout intensity, and mood after exercise. Across all analyses, no statistically significant or robust predictive relationships were identified between sleep duration and performance outcomes. These null findings suggest that wearable datasets may mask true sleep-performance relationships due to factors such as device variability, participant heterogeneity, self-reported measures, and unmeasured confounding variables. Despite the absence of significant associations, this study highlights important methodological challenges and provides direction for future research using more standardized sleep tracking and controlled sleep study designs.