Ekelund · BMJ (Clinical research ed.) 2019 · systematic review and harmonised meta-analysis · n=8 studies (36,383 participants)

Dose-response associations between accelerometry measured physical activity and sedentary time and all cause mortality: systematic review and harmonised meta-analysis.

Cited 1793 times in the scientific literature.

Level 3 - non-randomized controlled study

Systematic review and meta-analysis of prospective observational cohort studies

PubMed 31434697 · doi:10.1136/bmj.l4570 · record verified 2026-08-29

What was done

A systematic review and individual participant data harmonised meta-analysis was conducted across prospective cohort studies indexed in PubMed, PsycINFO, Embase, Web of Science, and Sport Discus up to July 31, 2018. Studies were included if they measured physical activity and sedentary time via accelerometry and reported all-cause mortality. Physical activity and sedentary time were categorised into quarters at the study level. Cox proportional hazards regression models were run per study and pooled using random-effects meta-analysis across 8 eligible cohorts (n=36,383; mean age 62.6 years; 72.8% women; median follow-up 5.8 years; 2,149 deaths).

What was found

Any physical activity, regardless of intensity, showed a non-linear dose-response reduction in all-cause mortality. Compared to the least active quarter (Q1), hazard ratios (HR) for total physical activity were 0.48 (95% CI 0.43 to 0.54) in Q2, 0.34 (0.26 to 0.45) in Q3, and 0.27 (0.23 to 0.32) in Q4. For light physical activity, HRs were 0.60 (0.54 to 0.68) for Q2, 0.44 (0.38 to 0.51) for Q3, and 0.38 (0.28 to 0.51) for Q4. For moderate-to-vigorous activity, HRs were 0.64 (0.55 to 0.74) for Q2, 0.55 (0.40 to 0.74) for Q3, and 0.52 (0.43 to 0.61) for Q4. Conversely, higher sedentary time was associated with increased mortality: HR 1.28 (1.09 to 1.51) for Q2, 1.71 (1.36 to 2.15) for Q3, and 2.63 (1.94 to 3.56) for Q4 compared to the least sedentary quarter.

Why it matters

This study provides device-measured evidence that physical activity of any intensity, including light activity, substantially lowers all-cause mortality, while high sedentary time markedly increases risk.

Limits

The findings derive from observational cohorts, precluding causal conclusions and leaving potential for residual confounding or reverse causation. The sample was predominantly female (72.8%) with a mean age of 62.6 years, limiting generalisability to younger populations. Accelerometry was typically measured at a single baseline point, and three candidate studies were excluded due to harmonisation barriers such as differing wear locations.

Cited by