Relation of androgen receptor gene polymorphism to bone mineral density and fracture risk in early postmenopausal women during a 5-year randomized hormone replacement therapy trial

In women, the influence of androgens on bone health is not clear. It has been suggested that the androgen receptor (AR) genotype is associated with bone mineral density and serum androgen levels in pre- and perimenopausal women, but the association between AR genotype, bone mineral density, and fracture risk has not been studied in postmenopausal women. Therefore, we studied whether AR polymorphism affects bone mineral density, bone mineral density change, or fracture risk in a 5-year randomized hormone replacement therapy (HRT) trial on 331 early postmenopausal women (mean baseline age, 52.7 +/- 2.3 years). The participants consisted of two treatment groups: the HRT group (n = 151) received a sequential combination of 2 mg estradiol valerate and 1 mg cyproterone acetate with or without vitamin D-3, 100-300 IU + 93 mg calcium as lactate/day, and the non-HRT group (n = 180) received 93 mg calcium alone or in combination with vitamin D-3, 100-300 IU/day for 5 years. Bone mineral density was measured from lumbar spine and proximal femur (DXA) before and after the 5-year trial. All new symptomatic, radiographically defined fractures were recorded during the follow-up. The length of CAG repeat in exon 1 of AR gene was evaluated after polymerase chain reaction (PCR) amplification. The subjects were divided into three repeat groups according to AR alleles. None of the baseline characteristics were associated with AR gene polymorphism and HRT treatment. The polymorphism did not influence the calculated annual changes of lumbar or femoral neck bone mineral density during the 5-year follow-up in the HRT (p = 0.926 and 0.146, respectively) or non-HRT (p = 0.818 and 0.917, respectively) groups. In all, 28 women sustained 33 fractures during the follow-up. Thus, the numbers of fractures were limited. The AR repeat length variation was not significantly associated with fracture risk in the HRT or non-HRT groups (p = 0.632 and 0.459, respectively; Cox proportional hazards model). In conclusion, AR gene polymorphism was not associated with baseline bone mineral density, 5-year bone mineral density change, or fracture risk in early postmenopausal Finnish women.