Contributions of CAG Repeat Length in the Androgen Receptor Gene and Androgen Profiles to Premature Pubarche in Korean Girls

Abstract

Introduction: Because not all premature pubarche (PP) children exhibit increased androgen levels, the sensitivity of the pilosebaceous unit to the androgen receptor (AR) might influence the development of the condition. There are ethnic differences in the CAG repeat length in exon 1 of the AR gene, which exhibits an inverse relationship to AR sensitivity

however, no Asian studies on PP have been reported, including Korea. Therefore, the aims of the present study were to examine the clinical and hormonal characteristics and AR CAG repeat length, and then assess their contributions to PP in Korean girls. Methods: Forty-eight subjects were enrolled and classified into three groups as follows: PP (16 girls), normal pubarche (NP, 16 girls), and normal controls (NC, 8 girls and 8 boys). Anthropometric measurements were collected, and the levels of the hormones, including dehydroepiandrosterone (DHEA), dehydroepiandrosterone sulfate (DHEAS), 17-hydroxyprogesterone (17-OHP), androstenedione (AD), and total testosterone (TT), were measured. Free testosterone (FT) levels and free androgen index (FAI) values were calculated using the TT and sex hormone-binding globulin data. X-chromosome inactivation (XCI) analysis was used to calculate methylation-weighted (MW) average CAG repeat lengths. Results: The median ages at diagnosis of pubarche were 7.4 years in the PP group and 8.9 years in the NP group. Anthropometric data and fasting glucose and insulin levels were similar among all three groups. The levels of 17-OHP, DHEAS, AD, FT, and FAI were similar in the PP and NP groups. The PP group exhibited a higher DHEAS:DHEA ratio than the NP group (P=0.014). The medians and interquartile ranges of the MW average CAG repeat length of the AR gene were 22.4 (21.3-23.3) for all subjects and did not differ among the PP (median, 22.3), NP (median, 22.4), and NC (median, 22.2) groups. The MW average CAG repeat lengths in the PP and NP groups did not correlate with DHEAS (r=-0.060, P=0.745) or FT (r=0.223, P=0.221) levels. The XCI proportion tended to be higher in the PP group (71.9%) than in the NP (65.6%) or NC (60.4%) groups, although it did not reach statistical significance (PP vs. NP, P=0.130

PP vs. NC, P=0.490). Conclusions: The CAG repeat length of the AR gene was not involved in the development of PP in Korean girls. However, excessive adrenal androgen levels, particularly those caused by increased sulfotransferase activity, might be important in the pathogenesis of PP, although further work is required.