Effects of education on adult mortality: a global systematic review and meta-analysis

The positive effect of education on reducing all-cause adult mortality is known; however, the relative magnitude of this effect has not been systematically quantified. The aim of our study was to estimate the reduction in all-cause adult mortality associated with each year of schooling at a global level. In this systematic review and meta-analysis, we assessed the effect of education on all-cause adult mortality. We searched PubMed, Web of Science, Scopus, Embase, Global Health (CAB), EconLit, and Sociology Source Ultimate databases from Jan 1, 1980, to May 31, 2023. Reviewers (LD, TM, HDV, CW, IG, AG, CD, DS, KB, KE, and AA) assessed each record for individual-level data on educational attainment and mortality. Data were extracted by a single reviewer into a standard template from the Global Burden of Diseases, Injuries, and Risk Factors Study. We excluded studies that relied on case-crossover or ecological study designs to reduce the risk of bias from unlinked data and studies that did not report key measures of interest (all-cause adult mortality). Mixed-effects meta-regression models were implemented to address heterogeneity in referent and exposure measures among studies and to adjust for study-level covariates. This study was registered with PROSPERO (CRD42020183923). 17 094 unique records were identified, 603 of which were eligible for analysis and included data from 70 locations in 59 countries, producing a final dataset of 10 355 observations. Education showed a dose–response relationship with all-cause adult mortality, with an average reduction in mortality risk of 1·9% (95% uncertainty interval 1·8–2·0) per additional year of education. The effect was greater in younger age groups than in older age groups, with an average reduction in mortality risk of 2·9% (2·8–3·0) associated with each additional year of education for adults aged 18–49 years, compared with a 0·8% (0·6–1·0) reduction for adults older than 70 years. We found no differential effect of education on all-cause mortality by sex or Socio-demographic Index level. We identified publication bias (p