Exploring the association between ambient air pollution and COVID-19 risk: A comprehensive meta-analysis with meta-regression modelling

Abstract

Introduction: Air pollution is speculated to increase the risk of Coronavirus disease-2019 (COVID-19). Nevertheless, the results remain inconsistent and inconclusive. This study aimed to explore the association between ambient air pollution (AAP) and COVID-19 risks using a meta-analysis with meta-regression modelling.

Methods: The inclusion criteria were: original studies quantifying the association using effect sizes and 95 % confidence intervals (CIs); time-series, cohort, ecological or case-crossover peer-reviewed studies in English. Exclusion criteria encompassed non-original studies, animal studies, and data with common errors. PubMed, Web of Science, Embase and Google Scholar electronic databases were systemically searched for eligible literature, up to 31, March 2023. The risk of bias (ROB) was assessed following the Agency for Healthcare Research and Quality parameters. A random-effects model was used to calculate pooled risk ratios (RRs) and their 95 % CIs.

Results: A total of 58 studies, between 2020 and 2023, met the inclusion criteria. The global representation was skewed, with major contributions from the USA (24.1 %) and China (22.4 %). The distribution included studies on short-term (43.1 %) and long-term (56.9 %) air pollution exposure. Ecological studies constituted 51.7 %, time-series-27.6 %, cohorts-17.2 %, and case crossover-3.4 %. ROB assessment showed low (86.2 %) and moderate (13.8 %) risk. The COVID-19 incidences increased with a 10 μg/m³ increase in PM_2.5 [RR = 4.9045; 95 % CI (4.1548-5.7895)], PM₁₀ [RR = 2.9427: (2.2290-3.8850)], NO₂ [RR = 3.2750: (3.1420-3.4136)], SO₂ [RR = 3.3400: (2.7931-3.9940)], CO [RR = 2.6244: (2.5208-2.7322)] and O₃ [RR = 2.4008: (2.1859-2.6368)] concentrations. A 10 μg/m³ increase in concentrations of PM_2.5 [RR = 3.0418: (2.7344-3.3838)], PM₁₀ [RR = 2.6202: (2.1602-3.1781)], NO₂ [RR = 3.2226: (2.1411-4.8504)], CO [RR = 1.8021 (0.8045-4.0370)] and O₃ [RR = 2.3270 (1.5906-3.4045)] was significantly associated with COVID-19 mortality. Stratified analysis showed that study design, exposure period, and country influenced exposure-response associations. Meta-regression model indicated significant predictors for air pollution-COVID-19 incidence associations.

Conclusion: The study, while robust, lacks causality demonstration and focuses only on the USA and China, limiting its generalizability. Regardless, the study provides a strong evidence base for air pollution-COVID-19-risks associations, offering valuable insights for intervention measures for COVID-19.

Keywords: Air pollution; COVID-19; Effect estimates; Meta-analysis; Meta-regression.

Fig. 1

Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) flow diagram of identification, screening, eligibility and inclusion of studies.

Fig. 2

Bubble plots showing results of the meta-regression model for air pollution COVID-19 incidences. From left to right: Risk of bias (low or moderate), YOP (year of publication) and exposure window (long or short term).

Fig. 2

Bubble plots showing results of the meta-regression model for air pollution COVID-19 incidences. From left to right: Risk of bias (low or moderate), YOP (year of publication) and exposure window (long or short term).

Fig. 3

Bubble plots showing results of the meta-regression model for air pollution COVID-19 mortality. From left to right: exposure window (long or short term), Risk of bias (low or moderate) and YOP (year of publication).