Abstract
BACKGROUND: Extreme temperature is associated with adverse birth outcomes but it is unclear whether it increases early delivery risk. OBJECTIVES: We aimed to determine the association between ambient temperature and early delivery. METHODS: Medical records from 223,375 singleton deliveries from 12 U.S. sites were linked to local ambient temperature. Exposure to hot (> 90th percentile) or cold (< 10th percentile) using site-specific and window-specific temperature distributions were defined for 3-months preconception, 7-week periods during the first two trimesters, 1 week preceding delivery, and whole pregnancy. Poisson regression with generalized estimating equations calculated the relative risk (RR) and 95% confidence interval for early deliveries associated with hot/cold exposures, adjusting for conception month, humidity, site, sex, maternal demographics, parity, insurance, prepregnancy body mass index, pregnancy complications, and smoking or drinking during pregnancy. Acute temperature associations were estimated separately for warm (May-September) and cold season (October-April) in a case-crossover analysis using conditional logistic regression. RESULTS: Compared with mild temperature (10-90th percentile), exposure to hot or cold during weeks 1-7 increased risk for early preterm (< 34 weeks) [RR(hot): 1.11 (95% CI: 1.01, 1.21); RR(cold): 1.20 (95% CI: 1.11, 1.30)], late preterm (34-36 weeks) [RR(cold): 1.09 (95% CI: 1.04, 1.15)], and early term (37-38 weeks) [RR(hot): 1.04 (95% CI: 1.02, 1.07); RR(cold): 1.03 (95% CI: 1.00, 1.05)] delivery. Findings were similar for hot exposures during weeks 15-21. Examining deliveries at each week from 23 through 38, whole-pregnancy hot exposures increased delivery risk by 6-21% at weeks 34 and 36-38. In the case-crossover analysis, a 5°F increase during the week preceding delivery was associated with 12-16% higher and 4-5% lower early delivery risk during warm and cold season, respectively. CONCLUSIONS: Both acute and chronic ambient temperature extremes may affect early delivery risk. Citation: Ha S, Liu D, Zhu Y, Kim SS, Sherman S, Mendola P. 2017. Ambient temperature and early delivery of singleton pregnancies. Environ Health Perspect 125:453-459; http://dx.doi.org/10.1289/EHP97.