Maternal plasma corticotropin-releasing hormone trajectories vary depending on the cause of preterm delivery

Abstract

Objective: Although second trimester maternal plasma corticotropin-releasing hormone concentrations are elevated in women who are destined to deliver preterm, the sensitivity and positive predictive value of an individual test of corticotropin-releasing hormone concentration is low. This poor screening performance may be due in part to the observation that the causes of preterm delivery are heterogenous, with potentially different effects on corticotropin-releasing hormone production. We have reanalyzed data from a previous cohort of women with multiple samplings to determine whether the trajectory of increase of placental corticotropin-releasing hormone provided more information than did a single sample. Study Design: In this cohort, where 2 to 4 samples that had been assayed for corticotropin-releasing hormone were available on 305 women, the general form of the exponential equation y = aebt was fitted to the corticotropin-releasing hormone data of each individual by the Gauss-Newton nonlinear least squares method, which generated the parameters, y-intercept and rate of rise, in corticotropin-releasing hormone concentration for each woman. Nonparametric statistical techniques, including bootstrapping, were applied to compare the results for the group of women who delivered at term with the group of women who delivered spontaneously preterm and the group of women who were delivered preterm by induction or cesarean delivery because of obstetric indication. Results: The 3 clinically defined groups have significantly different parameters, y-intercept and rate of rise, which describe the corticotropin-releasing hormone trajectory. The group that delivered preterm because of obstetric indications had a similar mean y-intercept but significantly greater mean rate of rise in corticotropin-releasing hormone concentration than the term group (0.3491 vs 0.1788; 95% CI, 0.2331-0.4615 vs 0.1394-0.2330). The group that delivered spontaneously preterm had a significantly greater mean y-intercept than the group that delivered preterm because of obstetric indication (17.08 vs 1.83; 95% CI, 5.89-28.43 vs 0.03-5.19) but a similar mean rate of rise to the group that delivered at term. Conclusion: These data suggest that spontaneous preterm delivery is associated with an abnormal setting of the production of corticotropin-releasing hormone that is present from very early in pregnancy, although women who experience an induced preterm delivery are characterized by rapidly rising placental corticotropin-releasing hormone concentrations. These data further suggest that clinical abnormalities that are associated with preterm delivery by induction or cesarean delivery are associated with abnormalities that lead to a rapidly increasing corticotropin-releasing hormone concentrations. Trajectories for corticotropin-releasing hormone provide information beyond that obtained from a single sample. (Am J Obstet Gynecol 2002;186:257-60.)