0289 Computational Modeling of Sleep Opportunity in College Students Under Academic and Social Schedule Constraints

Abstract Introduction College students frequently report insufficient sleep, but the extent to which structural schedule constraints alone -independent of individual behavior- limit sleep opportunity is not well quantified. In this study, we developed a computational simulation model to estimate how academic schedules, evening obligations, and weekend variability shape the maximum achievable sleep window in a synthetic student population. Methods We constructed a Monte Carlo agent-based model generating 10,000 synthetic weekly schedules based on distributions of publicly available university class times. Morning constraints were drawn from a normal distribution centered at 9:30 (SD=1.2 h). Evening constraints (labs, extracurriculars) were sampled from empirical distributions with a 30% probability of extending past 20:00. Study blocks were assigned using a Poisson process (λ=1.8/day). Sleep opportunity was defined as the longest continuous interval between the day’s final constraint (plus a 45-min wind-down buffer) and the next morning’s earliest constraint (minus a 45-min preparation buffer). Weekend days removed academic constraints but preserved social obligations with higher variance. Simulations were stratified into STEM–like and non-STEM–like schedules based on evening course likelihood. Results Across all schedules, the mean weekday sleep opportunity was 6.41 ± 1.05 hours, compared to 8.12 ± 1.32 hours on weekends (p < .001). Only 14.7% of weekday nights allowed ≥8 hours of potential sleep. STEM-modeled schedules produced 22.9% fewer nights with ≥8 hours opportunity than non-STEM schedules (12.1% vs. 15.7%). Schedules containing evening labs delayed modeled bedtime by 1.08 hours, reducing sleep opportunity by 0.74 hours per affected day. The highest restriction occurred in students with both early-morning classes and ≥2 evening constraints, yielding an average weekday opportunity of 5.38 hours. Conclusion Structural features of typical college schedules substantially restrict sleep opportunity even before accounting for student behavior. These simulations show that many students may be potentially chronically limited by schedule architecture alone, suggesting a need for academic policy reforms. Computational modeling provides a scalable method for assessing population-level determinants of sleep health without requiring human-subjects data. Support (if any)