Associations of fish consumption during pregnancy with birth outcomes

Data from the 2009-2010 South Carolina Pregnancy Risk Assessment Monitoring System (SC PRAMS) and birth certificate records were used to assess the associations of maternal fish consumption during pregnancy and birth outcomes. In 2009-2010, as a part of National Environmental Public Health Tracking Program, SC PRAMS added state-specific questions to assess the frequency, types, and water sources of fish consumed by mothers during pregnancy. Multiple logistic regression models were used to examine the associations between fish consumption and binary outcomes (low birth weight: <2500g; preterm birth: <37 weeks). Multinomial logistic regression was used for categorical birthweight outcome (small-, average-, or large-for-gestational age). Linear regression models were used for continuous birthweight and duration of gestation. SAS-callable SUDAAN was used. In our sample (unweighted n=791), 6.8% of infants were born at low birth weight and mean birth weight was 3245.4 g (±22.7). The percentage of preterm births was 8.3% and mean gestational age was 272.8 days (±8.3). Women who consumed fish 4 or more times a month during pregnancy had 1.7 times (95% confidence interval (CI): 1.1, 2.7) higher odds of having a low-birthweight infant than women who didn't eat any fish during pregnancy. After adjustment, the association was attenuated (odds ratio: 1.7, 95% CI: 0.9, 3.1). No significant results were found for other measures of birth outcomes. Fish consumption during pregnancy was not associated with increased birth weight and gestational duration. Further investigation is needed to confirm these conclusions, as there are inconsistencies with the current research.