Pedisic et al. 2020, Running and mortality meta-analysis

Systematic review and meta-analysis of 14 prospective cohorts (232,149 people, 25,951 deaths). Any running was associated with 27% lower all-cause mortality, 30% lower cardiovascular mortality and 23% lower cancer mortality versus no running. No clear additional benefit from higher doses, but no signal of harm either. Observational, so association not proof of cause. Moderate.