Association of caffeine metabolites in umbilical cord blood with IUGR and preterm delivery: A prospective cohort study of 1609 pregnancies

Purpose To examine the association between IUGR and preterm delivery (PTD) and fetal caffeine exposure when estimated by serum caffeine and paraxanthine in umbilical cord. Methods Prospective cohort study conducted in Connecticut and Massachusetts specifically testing the relation of caffeine and its metabolites, measured in cord blood (ng/ml) to IUGR (N = 1606) and PTD (N = 1609). Important covariates were controlled for in logistic regression models. Ridge regression for generalized linear models was used to control for collinearity among metabolites. All metabolite analyses were done blind to consumption data and reproductive outcomes. Results We observed a 24% increase in risk for IUGR per paraxanthine quartile change (Ridge adjusted OR 1.24; 95% CI 1.10, 1.39), while caffeine was protective for IUGR: Ridge adjusted OR 0.87 (95% CI 0.78, 0.97) per quartile change. When the ratio of paraxanthine:caffeine was modeled, there was a 6-fold increase in risk for IUGR for every one-unit increase in the ratio (MLE adjusted OR 6.03; 95% CI 2.14, 17.03). There was an 18% increase in risk for preterm delivery per paraxanthine quartile change (Ridge adjusted OR 1.18; 95% CI 1.05, 1.33). Caffeine was not associated with PTD nor was the paraxanthine:caffeine ratio. Maternal self-reported caffeine consumption during the third trimester was not associated with either outcome. Conclusions Paraxanthine, rather than caffeine, may increase risk for IUGR and PTD. These findings may explain the equivocal literature evaluating reported consumption only. More research to replicate and extend these findings which have relevance to other health outcomes is warranted.