RESUMEN
Air pollution has been linked to adverse neonatal outcomes, mainly in the case of prolonged exposures. This study focuses on the short-term effects on maternal health. We conducted a retrospective ecological time-series study in the Madrid Region covering the period 2013-2018. The independent variables were mean daily concentrations of tropospheric ozone (O3), particulate matter (PM10/PM2.5) and nitrogen dioxide (NO2), as well as noise levels. The dependent variables were daily emergency hospital admissions due to complications in pregnancy, childbirth and the puerperium. Poisson generalised linear regression models were fitted to quantify the relative and attributable risks, controlling for trend, seasonality, the autoregressive nature of the series, and a number of meteorological factors. There were 318 069 emergency hospital admissions due to obstetric complications across the 2191 days of study. Of this total: 13 164 (95%CI: 9930-16 398) admissions were attributable to exposure to O3, the only pollutant to show a statistically significant (p < 0.05) association with admissions due to hypertensive disorders; and 10 575 (95%CI: 3573-17 566) admissions were attributable to daytime noise levels, while admissions due to hyperemesis gravidarum and vomiting were related to exposure to night noise. Other pollutants which also displayed statistically significant associations were: NO2 concentrations, with admissions due to vomiting and preterm labour; PM10 concentrations, with premature rupture of membranes: and PM2.5 concentrations, with total complications. Exposure to a range of air pollutants, and ozone in particular, is associated with a higher number of emergency hospital admissions due to gestational complications. Hence, surveillance of environmental effects on maternal health should be intensified, and plans and strategies to minimise these should be drawn up.