Objective: The objective of the study was to analyse placental hormone profiles in twin pregnancies to determine if they could be used to predict preterm birth.
Study design: Progesterone, estradiol, estriol and corticotropin-releasing hormone were measured using competitive immunoassay and radioimmunoassay in serum and saliva samples of 98 women with twin pregnancies,at 3 or more gestational timepoints. Hormone profiles throughout gestation were compared between very preterm (<34 weeks; n = 8), preterm (<37 weeks; n = 40) and term (37+ weeks; n = 50) deliveries.
Results: No significant differences were found between preterm and term deliveries in either absolute hormone concentrations or ratios. Estimated hormone concentrations and ratios at 26 weeks did not appear to predict preterm delivery. Salivary and serum hormone concentrations were generally poorly correlated.
Conclusion: Our results suggest that serial progesterone, estradiol, estriol and corticotropin-releasing hormone measurements in saliva and serum are not robust biomarkers for preterm birth in twin pregnancies.