The association between serum progesterone level and preterm delivery

OBJECTIVE: To investigate the association of serum progesterone level and preterm delivery. METHODS: The present prospective cohort study enrolled women with singleton pregnancies attending their first prenatal visit at the outpatient specialist clinics at KK Women's and Children's Hospital, Singapore, between September 1, 2010, and August 31, 2014. Progesterone levels were measured at four clinical visits (visit 1: 9-14 weeks; visit 2: 18-22 weeks; visit 3: 28-32 weeks; visit 4: >34 weeks) and were compared (after adjusting for potential confounders) between patients who had term and preterm deliveries, and among subgroups of spontaneous preterm and iatrogenic preterm deliveries. RESULTS: There were 708 patients included. Serum progesterone levels at visit 3 were higher in the preterm delivery group than in the term delivery group (P=0.036). The levels did not differ between the two groups at other visits (all P>0.05). In the subgroup analysis, progesterone levels were higher in the iatrogenic preterm delivery subgroup than the term subgroup at visits 1 and 3. A progesterone cut-off level of 304.5 nmol/L demonstrated 81.8% sensitivity, 40.1% specificity, and negative and positive predictive values of 97.5% and 7.2%, respectively, as a predictor of preterm delivery. CONCLUSION: Higher serum progesterone levels at 28-32 weeks of pregnancy were observed in women who had preterm deliveries; it was weakly predictive of preterm delivery.