Evaluating narrow windows of maternal exposure to ozone and preterm birth in a large urban area in Southeast Texas.

The association between O3 exposure and preterm birth (PTB) remains unclear. We evaluated associations for three categories of PTB and O3 in Harris County, Texas, during narrow periods of gestation. We computed two sets of exposure metrics during every 4 weeks of pregnancy for 152,214 mothers who delivered singleton, live-born infants in 2005-2007, accounting first for temporal variability and then for temporal and spatial sources of variability in ambient O3 levels. Associations were assessed using multiple logistic regression. We also examined the potential for a fixed cohort bias. In the bias-corrected cohort where associations were somewhat stronger, elevated odds ratios (ORs) per 10 parts per billion increase in O3 exposure (county-level metric) were detected for the fifth (OR=1.08, 95% confidence interval (CI): 1.04-1.12), sixth (OR=1.05, 95% CI=1.01-1.09), and seventh (OR=1.07, 95% CI=1.03-1.10) 4-week periods of pregnancy for late PTB (33-36 completed weeks gestation), the fifth (OR=1.13, 95% CI=1.02-1.25) and seventh (OR=1.15, 95% CI=1.04-1.27) 4-week periods of pregnancy for moderate PTB (29-32 completed weeks gestation), and the fifth (OR=1.21, 95% CI=1.08-1.36) 4-week period of pregnancy for severe PTB (20-28 completed weeks gestation). Conversely, decreased odds were found in the first 4-week period of pregnancy for severe PTB (OR=0.83, 95% CI=0.74-0.94). Associations were slightly attenuated using the spatially interpolated (kriged) metrics, and for women who did not work outside of the home. Our analyses confirm reports in other parts of the United States and elsewhere with findings that suggest that maternal exposure to ambient levels of O3 is associated with PTB.