Longitudinal Profiles of Thyroid Hormone Parameters in Pregnancy and Associations with Preterm Birth

INTRODUCTION: Overt thyroid disease in pregnancy is associated with numerous maternal and neonatal complications including preterm birth. Less is known about the contribution of trimester-specific subclinical alterations in individual thyroid hormones, especially in late gestation, on the risk of preterm birth. Herein, we examined the associations between subclinical changes in maternal thyroid hormone concentrations (TSH, total T3, free and total T4), measured at multiple time points in pregnancy, and the odds of preterm birth in pregnant women without clinical thyroid disease. PARTICIPANTS AND METHODS: Data were obtained from pregnant women participating in a nested case-control study of preterm birth within on ongoing birth cohort study at Brigham and Women’s Hospital in Boston, MA (N = 439; 116 cases and 323 controls). We measured thyroid hormones in plasma collected at up to four time points in pregnancy (median = 10, 18, 26, and 35 weeks). We used multivariate logistic regression models stratified by study visit of sample collection to examine associations. To reveal potential biological pathways, we also explored these relationships by obstetric presentation of preterm birth (e.g., spontaneous preterm delivery) that have been previously hypothesized to share common underlying mechanisms. RESULTS: In samples collected at median 10 and 26 weeks of gestation, we found inverse associations between FT4 and the odds of overall preterm birth (odds ratio [OR] = 0.57, 95% confidence interval (CI) = 0.33, 1.00; and OR = 0.53, 95% CI = 0.34, 0.84, respectively). Positive associations were detected for total T3 at these same time points (OR = 2.52, 95% CI = 1.20, 5.31; and OR = 3.40, 95% CI = 1.56, 7.40, respectively). These effect estimates were stronger for spontaneous preterm birth. CONCLUSIONS: Our results suggest that subclinical alterations in individual maternal thyroid hormones may influence the risk of preterm birth, and the strength of these associations vary by gestational age.