Abstract
PURPOSE: To evaluate the effect of basal serum testosterone levels on the ovarian response and the cumulative live birth rate of infertile women undergoing in vitro fertilization (IVF). METHODS: It is a retrospective study in a university-affiliated assisted reproduction center in Hong Kong. Infertile women undergoing the first IVF cycle in the center between December 2012 and November 2016 with archived serum samples and available information on cumulative live birth were included for the analysis. RESULTS: A total of 1122 women were included for analysis. The median basal serum testosterone level was 0.53 (25-75th percentile: 0.40-0.67) nmol/L. Women with higher basal serum testosterone levels required a lower total dosage of gonadotrophin and a shorter duration of stimulation and had more oocytes retrieved. The cumulative live birth rates did not differ among women with serum testosterone levels in the four quartiles. Basal serum testosterone level was not a significant independent predictor of the cumulative live birth after adjusted for the women's age and number of normally fertilized oocytes in a binary logistic regression. The areas under the receiver operative characteristics (ROC) curves in predicting low or high ovarian response and the cumulative live birth were all below 0.6. CONCLUSION: Higher basal serum testosterone levels were associated with a better ovarian response but had no effect on the cumulative live birth rate of infertile women undergoing IVF.