Objectives-Outdoor air pollution has consistently been shown to predict mortality. The finding that this association is stronger in infants than in children or adults raises the question whether air pollution could also be related to pregnancy outcomes-such as birthweight and stillbirth. The association between outdoor air pollution and stillbirths and low birthweight in the Czech Republic, where air pollution was high, was examined. Methods-An ecological study was conducted, with routinely collected data on stillbirths and low birthweight (<2500 g), air pollution (total suspended particulates, sulphur dioxide (SO2), and nitrogen oxides (NOx)), and socioeconomic factors (mean income, car ownership, divorce rate, etc). The analyses were restricted to 45 districts on which data on air pollution were available for the period 1986-8. The effects of exposure variables on frequency of pregnancy outcomes were estimated by logistic regression with district-years as the units of analysis. Results-Stillbirth rate (4.2/1000 births in monitored districts) was not significantly associated with any indicator of air pollution, and was weakly related to mean income and proportion of births outside marriage. Crude prevalence of low birthweight (prevalence 5.5%) showed highly significant associations with several socioeconomic factors; after controlling for these, odds ratios (ORs) / 50 mu g/m(3) increase in pollutant were: 1.04 (95% confidence interval (95% CI) 0.96 to 1.12) for total suspended particles, 1.10 (1.02 to 1.17) for SO2, and 1.07 (0.98 to 1.16) for NOx. When all pollutants were included in one model, SO2 remained related to low birthweight (OR 1.10 (1.01 to 1.20), p=0.033). Conclusion-These results are consistent with a recent study in China where birthweight was also associated with total suspended particles and SO2 but ecological studies must be interpreted cautiously. Residual confounding by socioeconomic factors cannot be ruled out. The association between air pollution and birthweight requires further investigation.
