First trimester levels of inhibins and activin A in normal and failing pregnancies

OBJECTIVE Miscarriage is the commonest complication of human pregnancy. We undertook this study to assess whether inhibin A, pro-alphaC inhibin and/or activin A, products of the corpus luteum and placenta, might be useful in either the prediction or diagnosis of miscarriage. DESIGN Case-control study. PATIENTS Ninety-eight asymptomatic women at 6-13 weeks gestation who subsequently had a miscarriage and 198 gestation-matched women with a normal singleton pregnancy. MEASUREMENTS Maternal serum levels of inhibin A, pro-alphaC inhibin, activin A and human chorionic gonadotrophin (hCG) were measured. RESULTS Inhibin A, pro-alphaC and hCG, expressed as multiples of the normal median (MoM; +/-95% confidence intervals) in the miscarriage cases were significantly lower than in the viable controls, 0.56 (0.48-0.69) MoM vs. 1.00 (0.98-1.13) MoM, 0.55 (0.51-0.84) MoM vs. 1.0 (0.86-1.22) MoM and 0.34 (0.23-0.36) MoM vs. 1.00 (0.94-1.08) MoM, respectively (P < 0.0001 for all). Of the three analytes, hCG was the most discriminating between cases and controls. Levels of activin A in the miscarriage cases were not significantly different from controls, 0.96 (0.86-1.07) vs. 1.0 (0.95-1.08). CONCLUSIONS These data suggest that inhibin A, pro-alphaC inhibin and activin A will not be useful in either the prediction or diagnosis of early pregnancy miscarriage.