Association between serum progesterone concentration in early pregnancy and duration of pregnancy: a cohort study

Objective: To examine the association between progesterone concentration in early pregnancy and duration of pregnancy and risk of preterm delivery. Methods: Women enrolled in the Born in Guangzhou Cohort Study from 2013–2014, with a singleton pregnancy, who had serum progesterone measured at least one time between 4 and 10 weeks of gestation were included. The association between progesterone concentration both continuous and as categorical variable (quartile) and the risk of preterm delivery was assessed with Cox proportional hazards regression. Differences of length of gestation in four progesterone concentration quartiles were assessed using the Log-rank test. Results: We studied 1860 mother-newborn pairs. The mean overall progesterone concentration was 65.7 ± 21.3 nmol/L, with mean progesterone concentrations in the four quartiles of 42.4 ± 6.2 nmol/L (n = 463), 56.2 ± 3.3 nmol/L (n = 462), 68.9 ± 4.5 nmol/L (n = 470), and 95.1 ± 15.3 nmol/L (n = 465). There was no significantly difference in duration of gestation in four progesterone concentration groups (p=.511). There was no relation between progesterone level and preterm delivery (adjusted hazard ratio (HR) per 10 nmol/l progesterone level 1.00 (95% confidence interval (CI) 0.90, 1.11)). After adjusting for potential confounders, the HR of any preterm delivery for quartiles 1, 2 and 3 versus the highest quartile of progesterone level (> 77.3 nmol/L) was 1.04 (95% CI 0.52, 2.07), 1.17 (95% CI 0.60, 2.28), and 1.46 (95% CI 0.76, 2.78), respectively. When analysis was done for spontaneous preterm delivery only, also no association with first trimester progesterone was found. Conclusion: Lower first trimester serum progesterone concentration is not associated with reduction of length of gestation or increased risk of preterm delivery.