Serum levels of inhibin in maternal and umbilical blood during pregnancy.

Inhibin levels were measured by a double antibody heterologous radioimmunoassay in the peripheral serum of 75 pregnant women throughout gestation and in serum from the umbilical vein and artery, which was obtained at the time of delivery. For reference, samples were obtained from 20 nonpregnant women in the early (days 0 to 3), mid (days 4 to 8), and late (days 9 to 14) luteal or follicular phase. Maternal serum levels of inhibin (mean +/- SEM) in early (6 to 12 weeks) gestation (36.4 +/- 2.6 U/ml, n = 36) were significantly (p less than 0.01) higher than those in serum from nonpregnant women in the mid (23.9 +/- 2.5 U/ml, n = 19) or late (11.3 +/- 0.6 U/ml, n = 19) luteal phase. Inhibin levels in maternal serum fell to 15.9 +/- 1.4 U/ml (n = 24) in mid (14 to 20 weeks) gestation and then gradually increased during late (21 to 40 weeks) gestation to peak levels of 49.4 +/- 5.1 U/ml (n = 9) at 36 to 37 weeks. Inhibin levels declined in parallel with human chorionic gonadotropin concentrations during the first trimester (r = 0.587 at p less than 0.01). Significant positive correlations (p less than 0.001) were observed between serum levels of inhibin and 17 beta-estradiol (r = 0.560), progesterone (r = 0.648), and human placental lactogen (r = 0.715) during mid and late (20 to 40 weeks) gestation. Inhibin levels in umbilical vein serum (38.5 +/- 1.3 U/ml, n = 5) were not different from those in umbilical artery serum (39.4 +/- 3.6 U/ml) but were significantly (p less than 0.01) lower than those in maternal serum (50.9 +/- 5.3 U/ml), which was obtained at the time of delivery. By day 5 of puerperium, serum levels of inhibin in the maternal vein were extremely low (2.3 +/- 0.1 U/ml, n = 7); these levels were nearly one fifth lower than follicular phase levels of 10.9 +/- 3.4 U/ml (n = 38). We propose that maternal inhibin in early gestation is secreted from the corpus luteum of pregnancy but that increasing inhibin levels during mid and late gestation result from inhibin that is produced by the placenta. The lack of an umbilical arterial-venous gradient for inhibin and the higher levels of inhibin in maternal serum argue against a fetal source of inhibin in the maternal circulation. The physiologic function of inhibin that is produced by the corpus luteum and by the placenta remains to be determined.