Background: Few studies have investigated the association between exposure to fine particulate matter (PM2.5) and infant mortality in developing countries, especially for the health effects of specific PM2.5 constituents. We aimed to examine the association of long-term exposure to specific PM2.5 constituents with infant mortality in 15 African countries from 2005 to 2015. Methods: Based on the Demographic and Health Surveys (DHS) dataset, we included birth history records from 15 countries in Africa and conducted a multi-country cross-sectional study to examine the associations between specific PM2.5 constituents and infant mortality. We used satellite-based models to predict annual residential exposure to PM2.5 and its six constituents, including organic carbon (OC), black carbon (BC), sulfate (SO>sub>42−) , nitrate (NO3−), ammonium (NH4+), and soil dust (DUST). Multivariable logistic regression analysis was employed by fitting single-constituent models, the constituent-PM2.5 models, and the constituent-residual models. We also conducted stratified analyses by potential effect modifiers and examined the specific associations for each country. Findings: We found positive and significant associations between PM2.5 total mass and most of its constituents with infant mortality. In the single-constituent model, for an IQR increase in pollutant concentrations, the odds ratio (OR) of infant mortality was 1·03 (95%CI; 1·01, 1·06) for PM2.5 total mass, and was 1·04 (95%CI: 1·02, 1·06), 1·04 (95%CI: 1·02, 1·05), 1·02 (95%CI: 1·00, 1·03), 1·04 (1·01, 1·06) for BC, OC, SO42-, and DUST, respectively. The associations of BC, OC, and SO42- remained significant in the other two models. We observed larger estimates in subgroups with older maternal age, living in urban areas, using unclean cooking energy, and with access to piped water. The associations varied among countries, and by different constituents. Interpretation: The carbonaceous fractions and sulfate play a major important role among PM2.5 constituents on infant mortality, suggesting the need for a targeted controlling strategy on corresponding sources of emissions. Funding: HK was supported by the National Natural Science Foundation of China (92043301 and 82030103). CL was supported by the China Postdoctoral Science Foundation (2020M681185) and China National Postdoctoral Program for Innovative Talents (BX20200088). Declaration of Interests: The authors have declared that no competing interests exist.
Read full abstract