Arem · JAMA internal medicine 2015 · pooled prospective cohort study · n=661137

Leisure time physical activity and mortality: a detailed pooled analysis of the dose-response relationship.

Cited 1517 times in the scientific literature.

Level 3 - non-randomized controlled study

Pooled analysis of prospective cohort studies

PubMed 25844730 · doi:10.1001/jamainternmed.2015.0533 · record verified 2026-08-28

What was done

Data were pooled from 6 prospective cohort studies in the National Cancer Institute Cohort Consortium (baseline 1992–2003) across the United States and Europe. The analysis included 661,137 adults (median age 62 years, range 21–98) followed for a median of 14.2 years, during which 116,686 deaths occurred. Self-reported moderate- to vigorous-intensity leisure time physical activity was quantified in metabolic-equivalent (MET) hours per week and evaluated against the 2008 physical activity guideline minimum (7.5 MET-hours/week) using multivariable-adjusted, cohort-stratified Cox proportional hazards models.

What was found

Compared with no leisure time physical activity, performing less than the recommended minimum was associated with a 20% lower mortality risk (HR 0.80; 95% CI, 0.78–0.82). Meeting 1 to 2 times the minimum yielded a 31% reduction (HR 0.69; 95% CI, 0.67–0.70), and 2 to 3 times yielded a 37% reduction (HR 0.63; 95% CI, 0.62–0.65). The maximum benefit threshold occurred at 3 to 5 times the recommendation (HR 0.61; 95% CI, 0.59–0.62; 39% reduction). Activity at 10 or more times the recommended minimum showed no elevated mortality risk (HR 0.69; 95% CI, 0.59–0.78). Similar dose-response curves were observed for cardiovascular and cancer mortality.

Why it matters

This study establishes that even sub-target activity levels confer meaningful survival benefits, while very high volumes of physical activity plateau in benefit rather than causing excess mortality risk.

Limits

Physical activity was based entirely on self-report questionnaires, creating susceptibility to misclassification and recall bias. The observational design cannot eliminate residual confounding, and baseline activity measurements did not capture changes in exercise habits over the 14-year follow-up period. The sample was restricted to cohorts from the United States and Europe.

Cited by