Answers are grounded in this paper and cite their sources.
In this retrospective observational study, we examined associations of background temperature and PM2.5 with campus running under semester distance requirements. We linked 651,606 quality-screened runs by 19,255 students at two Beijing campuses across six semesters, 2023–2025 to hourly environmental data; timing analyses used daily means. Joint nonlinear models included student-by-campus fixed effects and weather and calendar covariates, with two-way clustering by student and date. Compared with 15 °C, time per kilometer was 1.28% longer at 5 °C (95% confidence interval 0.97–1.59%; approximately 4.8 s/km at the observed median pace) and 1.05% shorter at 25 °C (0.72–1.38%). The global temperature–pace test remained significant after Holm adjustment, whereas the primary PM2.5–pace test was inconclusive (Holm P = 0.140). Distance was 3.60% shorter at 5 °C than at 15 °C (2.83–4.36%). Among 636,181 observed student-campus-days, the proportion of first starts in the evening (≥ 18:00) was 6.87% points lower at 5 °C than at 15 °C (4.62–9.12 lower). Daily PM2.5 of 75 versus 25 µg/m³ was associated with a 2.23-percentage-point lower evening proportion (0.61–3.85 lower). Both global timing tests remained significant after exploratory adjustment. Analyses of common samples examined alternative environmental assignments and timing definitions. These modest associations describe recorded campus activity, and their magnitude varied with model adjustment.
No takes yet. Share an insight, caveat, or question.
Peng et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: