Running sounds hard? Good news!

Meta-analysis of cohort studies. 232,149 people. follow-up 5.5 to 35 years across the included cohorts; 25,951 deaths.

Many studies, added up. People were watched, not told what to do. A clue, not proof.

Who: Adults from 6 prospective cohorts (14 publications) in the USA (Aerobics Center Longitudinal Study; NHANES; 50+ Runners Association vs Stanford controls), Denmark (Copenhagen City Heart Study), UK (Health Survey for England + Scottish Health Survey) and China (Shanghai Men's Health Study, men only); non-clinical populations; runners were ~10% of the pooled sample

Nobody was told to run.

More to know, from our read of the paper:

• Meta-analysis of observational cohorts: association, not causation. Authors: reverse causation 'cannot be ruled out' (sick people are less likely to run); residual confounding possible; only 4 studies met all their adjustment requirements
• Running was self-reported in every included study, with differently worded questions across cohorts
• Few studies per analysis: all-cause pooled from 5 cohorts, cardiovascular from 3 (substantial heterogeneity, I2 63%, CI up to 0.98), cancer from 4, women from only 2; dose-response from only 3 cohorts. Publication bias could not be assessed
• One included study (NHANES) found a non-significant association but could not be pooled, so the pooled HR 'might be somewhat overestimated' (authors)
• In one cohort (Aerobics Center, Lee et al.) the running-mortality link disappeared after adjusting for cardiorespiratory fitness, i.e. fitness may be the pathway

Scientists added up 14 studies of 232,149 grown-ups in the USA, the UK, Denmark and China, who said themselves whether and how much they ran, then checked who died over 5 to 35 years. People who ran at all had about a 27% lower risk of dying than people who never ran, and even once a week or less (under 50 minutes a week) showed up. Running more showed no clear extra benefit (that doesn't mean more is bad). Nobody was told to run, so this is a clue, not proof (a meta-analysis of observational cohort studies): healthier people may simply run more.

We read the abstract and the full text of this paper and checked every number on this page against it (19 facts, checked on 2026-10-03). The study design is named above. Cohort and other observational studies are described as linked, not proven. Found an error? Email theis@onestudyaday.com.