Gestational exposure to landscape fire increases under-5 child death via reducing birthweight: A risk assessment based on mediation analysis in low- and middle-income countries

Abstract

BackgroundExposure to landscape fire smoke (LFS) is linked to child mortality and birthweight. It is unknown whether gestational exposure to LFS affects child survival rate. We aimed to link under-five death (U5D) to gestational LFS exposure by performing a causal mediation analysis based on birthweight. MethodWe conducted a sibling-matched case-control study of children under 5 years of age who were affiliated with the same mothers from Demographic and Health Surveys in 54 low- and middle-income countries, during the period from 2000 to 2014. LFS exposure was quantified as the surface concentration of fine particulate matter (PM2.5) attributable to landscape fires, estimated using a global atmospheric model. Three pairwise associations between fire-sourced PM2.5, birthweight, and U5D were assessed using fixed-effects regressions. We used a bootstrap-based mediation test of regression coefficients to examine whether the LFS-birthweight-U5D pathway was statistically significant. We also conducted three pairwise exposure-response functions using nonlinear models and used them to estimate the pathway-specific disease burden from 2000 to 2014. ResultsAfter adjustments for multiple confounders, each 1-µg/m3 increase in gestational exposure to fire-sourced PM2.5 was associated with a reduction of 2.179 (95% confidence interval [CI]: −3.777, −0.580) g in birthweight. Each 1-g birthweight reduction was associated with a 0.072% (95% CI: 0.065%, 0.078%) increase in U5D. Furthermore, each increase in exposure to fire-sourced PM2.5 was associated with a 2.853% (95% CI: 0.835%, 4.911%) increase in U5D; 7.294% (95% CI: 0.710%, 24.254%) of the linkage was explained by LFS-attributable birthweight reduction. Based on the estimated exposure-response functions, from 2000 to 2014, global exposure to fire-sourced PM2.5 contributed a mean birthweight reduction of 10.30 (95% CI: 2.93, 19.47) g, contributing to 60,350 (18,111, 106,619) premature U5Ds annually. ConclusionIn low- and middle-income countries, gestational exposure to LFS can increase mortality during infancy; appropriate interventions are needed to promote health in childhood.