A 2017 review in Progress in Cardiovascular Diseases reported that runners had a 25 to 40 percent lower risk of premature death and lived about three years longer than non-runners. Those figures summarize observational research, not a lifelong experiment that assigned people to run.
The useful finding is less about athletic achievement than regular participation. In a large US cohort, adults reporting short, slow runs had much of the mortality advantage associated with running more often, farther or faster.
We are writers, not clinicians. This is a reading of population research, not medical or exercise advice.
The 25 to 40 percent range is a research summary
Duck-chul Lee and colleagues drew the headline range from several prospective cohorts. In this kind of study, researchers record people’s running habits and follow them for deaths. The review’s numbers describe a pattern across studies, not one fixed effect produced by a randomized trial.
That leaves a familiar problem. Runners can differ from non-runners in smoking, weight, income, diet, existing illness and access to health care. Statistical adjustments narrow those differences but cannot guarantee that running caused the whole mortality gap. The finding is worth taking seriously, but it should not be read as the final word.
The three-year estimate came from one large cohort
The central evidence was a 2014 Journal of the American College of Cardiology study, also led by Lee. It followed 55,137 adults for an average of 15 years. There were 3,413 deaths from all causes, including 1,217 from cardiovascular causes, and about 24 percent of participants reported running.
After adjustment, runners had a 30 percent lower all-cause mortality risk and a 45 percent lower cardiovascular mortality risk than non-runners. The researchers converted their statistical model into a three-year life-expectancy difference. Participants were not observed in a controlled experiment until each person died three years earlier or later.
The sample was mostly non-Hispanic white, college educated and from middle-to-upper socioeconomic groups attending the Cooper Clinic in Dallas. Running was self-reported from activity during the previous three months. Those details make the estimate less universal than the clean headline number suggests.
Short, slow running carried much of the association
The cohort divided runners into five groups by weekly time. Even the lowest group, below 51 minutes a week, had lower mortality than non-runners. The association also appeared below six miles a week, at one or two runs a week and at speeds below six miles per hour.
In absolute terms, the age-, sex- and examination-year-adjusted death rate was 45.9 per 10,000 person-years among non-runners. It ranged from 29.7 to 33.8 across the five running-time groups. After broader adjustment, the estimated all-cause mortality risk was 23 to 33 percent lower across those groups.
More running did not produce a clear, steadily larger mortality advantage within this dataset.
That is the basis for saying the habit appeared more important than speed or distance. It does not establish that any tiny dose is enough for every person, or that pace and mileage never matter for fitness, performance, injury exposure and enjoyment.
A later meta-analysis found a similar pattern
A systematic review led by Zeljko Pedisic combined 14 studies from six prospective cohorts: 232,149 people, 25,951 deaths and follow-up periods ranging from 5.5 to 35 years. Any running was associated with 27 percent lower all-cause mortality, 30 percent lower cardiovascular mortality and 23 percent lower cancer mortality than no running.
The authors found no statistically significant dose-response trend for weekly duration, frequency, pace or total volume. Only three cohorts contributed to those dose calculations, running definitions varied and activity was self-reported. Failure to detect a trend is not proof that dose makes no difference.
A broader 2024 review of different sports put running’s association with all-cause mortality at 23 percent, with substantial variation between studies. The direction is consistent; the exact size is not.
These studies cannot fully separate running from runners
Long follow-up and thousands of recorded deaths give the evidence weight. Its central weakness remains participant choice. Runners in the 2014 cohort were younger, leaner, less likely to smoke and fitter at baseline. Models accounted for many measured differences, including other physical activity, but unmeasured behavior and health can remain.
Reverse causation matters too. Some people may not run because illness, pain or declining fitness has already raised their mortality risk. Excluding people with certain serious diseases at baseline and deaths in the first year reduced that problem without eliminating it.
That broader context is why an earlier Science Blog report on several leisure activities is useful: running is not the only form of movement associated with lower mortality.
The mortality finding does not replace activity guidance
The largest contrast in these datasets was between some running and none. That does not mean a few weekly minutes provide every benefit addressed by public-health recommendations. The current US guideline for adults calls for at least 75 minutes of vigorous aerobic activity or 150 minutes of moderate activity each week, plus muscle-strengthening work on two days. It also says some activity is better than none.
People with chronic conditions, concerning symptoms or a long period of inactivity may need individualized guidance from a clinician before taking up vigorous exercise.
A stronger next step would repeatedly measure running and other activity with wearable devices, track changes in health and recruit a more representative population. A randomized trial lasting long enough to measure deaths would be difficult, so more precise answers will probably continue to depend on better observational studies.