Association between Ambient PM2·5 and Under- Five, Infant, and Child Mortality in Latin America, 2010- 2015: A Longitudinal Analysis in 337 Cities

Background. Highly urbanized and rapidly urbanizing countries are facing critical air pollution problems. In Latin American cities, the impact of ambient PM2·5 on child health has not been extensively explored. Objectives. To examine association between ambient PM2·5 and under-5 mortality (U5MR) infant mortality (IMR) and child mortality (CMR) in Latin American cities. Methods. We estimated U5MR, IMR, and CMR from 2010 to 2015 for 1,152 sub-city units clustered in 337 cities from Argentina, Brazil, Chile, Colombia, Costa Rica, Guatemala, and Mexico. We retrieved satellite annual PM2·5 for each sub-city unit. Using linear mixed-effects models, we estimated the percent change in U5MR, IMR, and CMR associated with a 1 μg/m3 increase in annual PM2·5, adjusted for city and municipality-level predictors (population growth, GDP, living conditions, service provision, population educational attainment, and mass transit availability). Results. Over the study period, mean annual PM2·5 was 12·7 μg/m3 (SD 6·3), U5MR and IMR were 14·2 and 12·1 deaths per 1,000 live births, respectively. CMR was 4·8 deaths per 10,000 children. A 1 μg/m3 annual increase in PM2·5 was associated with 0·4% increase in U5MR (95% CI 0.1 to 0.7) and 0·5% increase in IMR (95% CI 0.2 to 0.9) We found no significant association between PM2·5 increases and changes in CMR. Interpretation. Infants appear to be particularly vulnerable to the deleterious effect of PM2·5. Assuring a clean air environment is important to promote infant survival and health. Funding: The project is funded by Wellcome Trust Funds, Grant 205177/Z/16/Z. Declaration of Interest: We declare no competing interests. Ethical Approval: The SALURBAL study protocol has ethical approval from the Drexel University IRB with ID # 161 200 5035.