Objectives To systematically review and analyse studies of high amounts of physical activity and mortality risk in the general population.
Eligibility criteria Inclusion criteria related to follow-up (minimum 2 years), outcome (mortality from all causes, cancer, cardiovascular disease (CVD) or coronary heart disease), exposure (eg, a category of >1000 metabolic equivalent of task (MET) min/week), study design (prospective cohort, nested case control or case-cohort) and reports of cases and person years of exposure categories.
Information sources Systematic searches were conducted in Embase and Pubmed from database inception to 2 March 2019.
Risk of bias The quality of the studies was assessed with the Newcastle–Ottawa scale.
Included studies From 31 368 studies identified, 48 were included. Two authors independently extracted outcome estimates and assessed study quality.
Synthesis of results We estimated hazard ratios (HRs) using random effect restricted cubic spline dose–response meta-analyses. Compared with the recommended level of physical activity (750 MET min/week), mortality risk was lower at physical activity levels exceeding the recommendations, at least until 5000 MET min/week for all cause mortality (HR=0.86, 95% CI 0.78 to 0.94) and for CVD mortality (HR=0.73, 95% CI 0.56 to 0.95).
Strengths and limitations of evidence The strengths of this study include the detailed dose–response analyses, inclusion of 48 studies and examination of sources of heterogeneity. The limitations include the observational nature of the included studies and the inaccurate estimations of amount of physical activity.
Interpretation Compared with the recommended level, mortality risk was lower at physical activity levels well above the recommended target range. Further, there was no threshold beyond which lifespan was compromised.
Registration PROSPERO CRD42017055727.
- physical activity
Statistics from Altmetric.com
Contributors KB and AG conceived and designed the study. KB and CFB selected the articles. KB, CFB and MR-L extracted the data. KB, CFB and AG assessed the quality of the included studies. KB and AC performed the data analysis. KB and AG wrote the first draft of the manuscript. KB, AG, MR-L and AC interpreted the data. KB, AG, AC, MR-L and CFB contributed to the writing of the final version of the manuscript. All authors agreed with the results and conclusions of this article.
Funding This work was supported by grants from the Lundbeck Foundation (grant No R151-2013-14641) to AG; the Danish Council for Independent Research (grant No DFF-4004-00111) to AG; the European Research Council (grant No 716657] to AG. The Centre for Physical Activity Research (CFAS) is supported by a grant from TrygFonden. The funder of the study had no role in study design, data collection, data analysis, data interpretation or writing of the report. The corresponding author had full access to all of the data in the study and had final responsibility for the decision to submit for publication.
Competing interests MR-L: grants from the Danish Diabetes Academy, funded by the Novo Nordic Foundation, outside the submitted work.
Patient consent for publication Not required
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Code available from the corresponding author (firstname.lastname@example.org). Dataset can also be provided.