RESUMEN
BACKGROUND AND OBJECTIVE: Preterm birth is a major determinant of infant mortality and morbidity. Air pollution has been suggested as a risk factor for preterm delivery; however, the scientific evidence on this impact remains inconsistent. We assessed the association between residential exposure to air pollution during pregnancy and preterm birth (gestational age at delivery <37 weeks) in Spain. METHODS: This study was based on 2409 pregnant women participating in the INMA birth cohorts in Asturias, Gipuzkoa, Sabadell and Valencia. Ambient levels of nitrogen dioxide (NO2) and benzene were estimated for each woman's residence for each trimester and for the whole pregnancy, using temporally adjusted land-use regression models. The association between air pollution exposure and preterm birth was assessed for each cohort separately by means of logistic regression models controlling for potential confounders, under single- and two-pollutant models, for all the women in the study and for those spending more than 15h/day at home. Combined estimates of the association across cohorts were obtained through meta-analysis. RESULTS: Throughout the whole sample, suggestive but no statistically significant associations were found between exposure and preterm birth. For pregnant women spending more time at home significant associations were found for both pollutants, under single- and two-pollutant models. Under the last ones, NO2 exposure during the second trimester and the whole pregnancy was associated with a higher risk of preterm delivery (OR=1.58, (95%CI: 1.04-2.42) per 10µg/m(3) increase). Benzene exposure during the third trimester was also associated with preterm birth in that subsample (OR=1.45, (95%CI: 1.00-2.09) per 1µg/m(3) increase). CONCLUSION: We found suggestive associations between NO2 and benzene exposure during pregnancy and preterm birth. Estimates of the association were higher among women who spent more time at home, probably reflecting a better exposure assessment in this group.