Background: Recent reviews conclude that environmental tobacco smoke (ETS) leads to diminished birth weight. However, the threshold and magnitude of that effect is uncertain. We aimed to determine the magnitude and shape of the relations between ETS and various adverse pregnancy outcomes using a highly sensitive biochemical assay.
Methods: Maternal serum specimens were collected from more than 3000 women enrolled in California's prenatal screening program in 1992 and analyzed for cotinine. Information on pregnancy outcomes was obtained from live birth/fetal death records and hospital questionnaires. We conducted analyses on 2777 woman-live birth pairs and 19 woman-fetal death pairs in which the mother was presumed to be a nonsmoker (midtrimester cotinine levels < or =10 ng/mL).
Results: In multiple logistic regression analyses, the odds ratios of fetal death, preterm delivery, and term-low birth weight were 3.4, 1.8, and 1.8, respectively, in the highest cotinine quintile (0.236-10 ng/mL), compared with the lowest quintile (<0.026 ng/mL). In adjusted linear models, there was a linear dose-dependent effect of log cotinine on mean birth weight (-109 g) and mean infant length (-0.84 cm) over the range of cotinine values. Linear relations were not found with respect to infant head circumference or the ratio of brain weight to body weight. Infant's body mass index declined with exposures above approximately 0.5 ng/mL cotinine. We estimated that ETS levels at or above 0.05 ng/mL (experienced by 62% of the study population) accounted for 12% of all adverse outcomes.
Conclusions: ETS exposure in pregnant women adversely affects pregnancy by increasing fetal mortality and preterm delivery at higher exposure levels and slowing fetal growth across all levels of ETS exposure.