Ambient particulate matter air pollution and stillbirth in Ghana: A difference-in-differences approach

Abstract

Sub-Saharan African countries, including Ghana, are known hotspots for fine particulate matter air pollution (PM2.5) and stillbirths but lacked epidemiologic evidence. We investigated the association between PM2.5 and stillbirth in Ghana. District-level stillbirth data were obtained from the Ghana Health Service for all 260 local districts from 2012 to 2019 for a total of 5,229,338 births, including 81,611 stillbirths. Spatiotemporal datasets, including satellite-derived PM2.5, temperature, population density, and gross domestic product were linked with the birth data. We applied a variant difference-in-differences design with conditional quasi-Poisson regression to estimate the risk of stillbirth associated with annual PM2.5 concentrations. We adjusted for relevant environmental and sociodemographic factors and performed subgroup analyses by population density and household air pollution. The average district-level annual stillbirth incidence was 29 (standard deviation = 55) per 1000 births. The annual average PM2.5 concentration was 59.97 μg/m3 (standard deviation = 9.75). Every 10 μg/m3 increment in annual average PM2.5 was associated with a 3% risk of stillbirth (RR); 1.03 (95% CI: 0.97, 1.09) for all-source PM2.5 and 2% risk each for anthropogenic (RR = 1.02, 95% CI: 0.96, 1.07) and natural (RR = 1.02, 95% CI: 0.94, 1.11) sources. The association was higher for moderate or high subgroup, relative to low subgroup and higher in natural than anthropogenic sources of PM2.5 exposures. Thus, there was some evidence for an adverse association between PM2.5 exposure and stillbirth but estimates were less precise. Given that the district-level variation may be underpowered, stronger risk is expected in future high-quality individual-level longitudinal cohort studies in Ghana.