University Archives

Investigating how physical activity influences exhaled breath composition provides valuable insights into short-term metabolic and physiological responses to exercise. This study applies data mining techniques to analyze pre- and post-workout breath samples collected from 148 student athletes at the University of Detroit Mercy through the Center for Augmenting Intelligence (CAI). A total of 210 samples were compiled from raw time-series sensor files, each summarized using mean, minimum, maximum, standard deviation, and median values across eight gas and environmental features, including Methane, Hydrogen, Acetone, CO₂, Pressure, and Humidity. After normalization and aggregation, descriptive and comparative analyses were performed to identify shifts in gas distributions following physical exertion. Results indicate that most measured gases—particularly Methane, Hydrogen, and Acetone—maintained stable medians and narrow ranges across pre- and post-exercise samples, suggesting limited immediate metabolic changes under the tested conditions. In contrast, Humidity exhibited noticeably greater variability after exercise, likely reflecting increased respiration rate and body heat dissipation. These findings demonstrate that while exhaled gas concentrations remain relatively stable in short-term post-exercise states, environmental and physiological parameters such as humidity respond dynamically to physical exertion. The processed dataset establishes a foundation for future predictive modeling to classify pre- and post-workout breath patterns and explore broader links between respiration, metabolism, and wellness indicators.