Abstract: Objective To evaluate the associations of GGN repeat polymorphism of androgen receptor (AR) with ovarian reserve and ovarian response in controlled ovarian stimulation (COS). Methods In this genetic association study conducted in a university-affiliated IVF center,a total of 361 women aged ≤40 years with basal FSH≤12 IU/L underwent the GnRH-agonist long protocol for COS. GGN repeat in the AR gene was analyzed with Sanger sequencing. The primary endpoint was the number of antral follicle counts (AFCs), and the secondary endpoints were stimulation days, total dose of gonadotropin (Gn) used, total number of retrieved oocytes, ovarian sensitivity index (OSI), and follicular output rate (FORT). Results The GGN repeat in exon 1 of the AR gene ranged from 13 to 24, and the median repeat length was observed to be 22. When considering the genotypes (S for GGN repeats <22, L for GGN repeats ≥22), all patients were divided into 3 groups: SS, SL, and LL. Generalized regression analysis indicated that the number of AFCs in group SS was significantly lower than that in group SL (adjusted β 1.8; 95% Cl, 0.2-3.4; P＝0.024) or compared with group LL (adjusted β 1.5; 95% Cl 0.2-2.7; P＝0.021). No significant difference in the number of AFCs was observed between group SL and group LL. In addition, generalized regression analysis indicated no significant differences in ovarian stimulation parameters among the 3 groups, either before or after adjusting for confounding factors. Conclusions GGN repeat length on the AR gene was associated with AFC but not with ovarian response in Chinese women, indicating that polymorphisms of the AR gene may affect ovarian reserve.

Key words: androgen receptor, GGN polymorphism, ovarian reserve, ovarian stimulation