Air pollution, dementia, and related outcomes: an updated systematic review

Abstract

Background/Aim: Since our prior review in 2016, the epidemiologic evidence examining the association between long-term exposure to air pollution and dementia and dementia-related outcomes has grown substantially. Our goal was to update this review and summarize the state of the evidence, highlighting consistencies, new developments, and continued methodological challenges.Methods: We systematically searched PubMed and EMBASE to identify new articles published through June 19, 2019. Systematic review methods followed those in the prior review. We collectively evaluated both newly identified, eligible studies and studies from our original review. Results: Based on our eligibility criteria, we included 36 reports (18 original + 18 new) in our review. Newer articles were more likely to report on associations with cognitive change, incident cognitive impairment, and neuroimaging outcomes. They also considered a broader range of pollutants and used exposure assessments with finer spatial resolution. Higher NO2/NOx exposures were consistently associated with lower cognitive level, without strong evidence of disparate impacts by cognitive domain. Analyses of PM2.5 and PM10 were more mixed, although the weight of the evidence still suggests either a null or adverse association between particulate matter and cognitive health. Selection bias, timing of exposure ascertainment relative to outcome, and reliance on claims or medical records to identify dementia remain the most common threats to study validity. Conclusions: The literature on air pollution and late-life cognitive health has grown substantially since 2016, and continues to support the hypothesis that air pollution impacts late-life cognitive health. To strengthen confidence in these conclusions, additional studies are needed that attend to key methodological issues, especially robust dementia ascertainment. Future work should also recognize the heterogeneity of dementia pathogenesis, consider associations with biomarkers of pre-clinical dementia, and consider the impact of air pollution mixtures.