Students running on two Beijing campuses completed distance requirements significantly slower when temperatures dropped to 5 °C, according to a retrospective observational study.
Researchers examined 651,606 quality-screened runs completed by 19,255 students across six semesters from 2023 to 2025. They linked these records to hourly environmental data while applying daily means for timing analyses.
Compared with a baseline of 15 °C, time per kilometer was 1.28% longer at 5 °C. This difference amounted to approximately 4.8 seconds per kilometer at the observed median pace, with a 95% confidence interval ranging from 0.97–1.59%.
Pace also shifted when conditions warmed. Time per kilometer was 1.05% shorter at 25 °C compared to 15 °C, with a confidence interval spanning 0.72–1.38%.
The global temperature-pace test remained statistically significant after Holm adjustment. However, the primary test evaluating the relationship between PM2.5 levels and running pace remained inconclusive with a Holm P value of 0.140.
Temperature changes also altered running distance and scheduling choices. Total distance covered was 3.60% shorter at 5 °C than at 15 °C, with a confidence interval between 2.83–4.36%.
Among 636,181 observed student-campus-days, students were less likely to begin workouts late in the day when the air turned cold. The proportion of first starts in the evening at 18:00 or later was 6.87 percentage points lower at 5 °C than at 15 °C.
Higher background air pollution also influenced evening workouts. Daily PM2.5 levels of 75 micrograms per cubic meter versus 25 micrograms per cubic meter associated with a 2.23-percentage-point lower evening proportion.
Both global timing tests remained significant after exploratory adjustment. Analyses of common samples examined alternative environmental assignments and timing definitions, confirming that these modest associations described recorded campus activity while varying with model adjustment.
