Maternal serum androgens in human pregnancy: early increases within the cycle of conception.

Previous studies have demonstrated elevations in testosterone and androstenedione initiated within the cycle of conception in pregnant non-human primates, and minimal data in the human support the same picture. In the present study we have investigated a group of patients scheduled for artificial insemination with regular menstrual cycles. For this study all patients provided blood samples at 5 days after the luteinizing hormone (LH) surges and daily through the luteal phase and into early pregnancy (n = 12). Patients who did not become pregnant served as normal controls (n = 9). We have measured 17-hydroxyprogesterone (17-OHP) as a marker of luteal activity not obscured by progesterone within the cycle of conception and testosterone and androstenedione as the major androgens. There were no significant changes in testosterone and androstenedione in the non-pregnant controls, but both testosterone and androstenedione were significantly elevated in the pregnant luteal phase, with the first increases occurring at 15 and 14 days respectively after the LH surge. Three of 12 pregnant patients did not demonstrate a dramatic increase in either testosterone or androstenedione and when examined more carefully a corresponding lack of increase in 17-OHP in those same subjects indicated less than optimal luteal activity, suggesting that these androgens were products of the corpus luteum. In three subjects in which consecutive non-pregnant and pregnant cycles were followed there was a dramatic increase from the non-pregnant luteal phase to the pregnant luteal phase indicating that the more important observation may be the concentrations of androgens in the conceptive luteal phase compared to some baseline, either previous luteal phase or even follicular phase. We have also studied changes in dehydroepiandrosterone sulphate and found that there was no significant contribution to this increase in androgens in early conception. These studies demonstrate a significant increase in both testosterone and androstenedione presumably of ovarian, specifically luteal, origin and that adrenal androgen production is not a factor in these changes.