Answers are grounded in this paper and cite their sources.
Background With over 2.8 billion adults engaged in occupational physical activity globally and physical inactivity remaining the fourth leading mortality risk factor, understanding domain-specific PA effects is a public health priority. Evidence suggests domain-divergent effects LTPA is protective; OPA may confer adverse effects. No umbrella review has compared evidence across all three PA domains for all-cause mortality. Methods Umbrella review of systematic reviews with meta-analyses examining domain-specific PA (LTPA, OPA, transport) and all-cause mortality (PROSPERO: CRD42026142431). Five databases and Epistemonikos were searched to June 2026. Four Chinese-language databases were additionally searched. Quality was assessed via AMSTAR 2; evidence credibility was classified using the Ioannidis framework. Primary study overlap was quantified via Corrected Covered Area (CCA) analysis. CCA-informed descriptive synthesis with correlated-effects sensitivity analysis was complemented by leave-one-out and AMSTAR 2 quality-stratified sensitivity analyses. Results Twelve systematic reviews (>300 cohort studies, >32 million participants) were included. A consistent descriptive domain-gradient emerged: LTPA showed strong protective associations (HR/RR 0.50–0.81, Ioannidis Class II, GRADE Moderate); transport PA showed moderate protection (RR 0.77–0.92, Ioannidis Class III, GRADE Moderate); OPA showed null-to-elevated-risk associations after LTPA adjustment (pooled HR 1.12, 95% CI 1.05–1.19, Ioannidis Class IV, GRADE Low), from one evidence cluster (pairwise CCA = 89%, very high study-level overlap; this does not mean 89% of participants derive from one source, but rather that the two reviews share approximately 89% of their included primary studies). CCA analysis confirmed largely independent evidence bases between domains (cross-domain CCA = 6.3%). Illustrative population attributable fraction analyses estimated that insufficient LTPA accounts for 10.2% of all-cause mortality globally, assuming causality. Conclusions This first multi-domain umbrella review reveals a consistent descriptive domain-gradient: LTPA is protective, transport PA moderately protective, and OPA shows null-to-elevated-risk after LTPA adjustment, though this derives from a single evidence cluster (CCA = 89%) and requires independent replication. Large Chinese cohorts (~0.9 million [CKB evidence reported in, n ~ 512,000 overlapping]; post-hoc exploratory finding, not from the 12 included reviews) report protective OPA associations, suggesting context-dependence. Current evidence suggests that PA domains may differ in mortality risk associations and warrant domain-specific consideration, though OPA evidence remains preliminary and requires independent replication.
No takes yet. Share an insight, caveat, or question.
Lin et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: