Reduction of preterm births (<37 completed weeks of gestation) would substantially reduce neonatal and infant mortality, and deleterious health effects in survivors. Maternal fine particulate matter (PM2.5) exposure has been identified as a possible risk factor contributing to preterm birth. The aim of this study was to produce the first estimates of ambient PM2.5-associated preterm births for 183 individual countries and globally. To do this, national, population-weighted, annual average ambient PM2.5 concentration, preterm birth rate and number of livebirths were combined to calculate the number of PM2.5-associated preterm births in 2010 for 183 countries. Uncertainty was quantified using Monte-Carlo simulations, and analyses were undertaken to investigate the sensitivity of PM2.5-associated preterm birth estimates to assumptions about the shape of the concentration-response function at low and high PM2.5 exposures, inclusion of provider-initiated preterm births, and exposure to indoor air pollution. Globally, in 2010, the number of PM2.5-associated preterm births was estimated as 2.7 million (1.8-3.5 million, 18% (12-24%) of total preterm births globally) with a low concentration cut-off (LCC) set at 10μgm-3, and 3.4 million (2.4-4.2 million, 23% (16-28%)) with a LCC of 4.3μgm-3. South and East Asia, North Africa/Middle East and West sub-Saharan Africa had the largest contribution to the global total, and the largest percentage of preterm births associated with PM2.5. Sensitivity analyses showed that PM2.5-associated preterm birth estimates were 24% lower when provider-initiated preterm births were excluded, 38-51% lower when risk was confined to the PM2.5 exposure range in the studies used to derive the effect estimate, and 56% lower when mothers who live in households that cook with solid fuels (and whose personal PM2.5 exposure is likely dominated by indoor air pollution) were excluded. The concentration-response function applied here derives from a meta-analysis of studies, most of which were conducted in the US and Europe, and its application to the areas of the world where we estimate the greatest effects on preterm births remains uncertain. Nevertheless, the substantial percentage of preterm births estimated to be associated with anthropogenic PM2.5 (18% (13%-24%) of total preterm births globally) indicates that reduction of maternal PM2.5 exposure through emission reduction strategies should be considered alongside mitigation of other risk factors associated with preterm births.
