Environment/Climate Change
Evaluating the Association Between Ambient Temperature and Small for Gestational Age Outcomes in Pennsylvania Births in 2010 and 2011 Jennifer Ko* Jennifer Ko Jagadeesh Puvvula Kyle Jackson Angela Aherrera Heather Burris Aimin Chen
Heat exposure can trigger adverse health outcomes in pregnant women, increasing dehydration risk and reducing appetite, which may impair fetal blood flow and nutrition. Prior studies often used city, ZIP code, census tract, and 4km/1km resolution temperature data from urban settings. This study used the Pennsylvania state birth registry to examine the association between gestational ambient temperature exposures and small-for-gestational-age (SGA). This study includes 178,428 live births between 2010-2011, with a gestational age of 32-42 weeks, to ensure an adequate sample size within each week. We assigned ambient temperature exposures on a weekly scale from 2 weeks before the mother’s last menstrual period until delivery. The exposure assessment was based on the study participants’ complete geocoded addresses during pregnancy and ambient temperatures at an 800m resolution from the PRISM climate group. We defined SGA as newborns with birthweight below the 10th percentile stratified by sex and week of gestation. Exposure-outcome associations are modeled using logistic regression within a distributed lag non-linear framework. These associations were adjusted for maternal race/ethnicity, education, marital status, birth year, and maternal smoking. We observed a non-linear association between ambient temperature and SGA during the third trimester. Notably, at 31 weeks of gestation, compared to the reference average weekly Tmax of 25°C, colder temperatures (-10 to -3°C and 3 to 13°C) were linked to lower odds of SGA, while 13 to 25°C showed lower non-significant odds and 25-33°C showed higher odds. Temperatures exceeding 33°C were associated with non-significant odds, with a declining trend potentially explained by protective measures such as spending more time indoors. These measures may reflect risk perception and adaptive behaviors in response to higher temperatures. Future analyses will integrate additional years to improve sample size and precision estimation.
