Serum oestradiol and oestrogen-receptor gene polymorphism are associated with bone mineral density independently of serum testosterone in normal males.

Abstract

The physiological effects of oestrogens on bone in men were largely unanticipated until recently, when oestrogen deficiency in males with aromatase deficiency and oestrogen resistance was found to cause osteoporosis and delayed fusion of epiphyses despite sufficient serum testosterone. This raises the possibility that in normal men oestrogens rather than androgens are of physiological importance in bone maturation. In the present study, we examined the association of serum oestradiol (E2) compared to that of free testosterone (FT) with bone mineral density (BMD) in normal men. The effect of oestrogen receptor (ER) gene polymorphism on BMD in men was also addressed. Eighty-one Thai men aged 20-79 years. All were healthy and did not take medication which may affect calcium and bone metabolism. BMD was assessed by DEXA. Dietary calcium was assessed by a 3-day dietary record. Serum E2 and FT concentrations were measured by radioimmunoassay. Polymorphism at intron 1 of the alpha isoform of ER gene was determined by PCR-RFLP. Small p represents the presence of the restriction site while capital P indicates the absence of the restriction site. Serum FT decreased with increasing age (r = -0.58, P < 0.0001) while E2 did not. However, there was a positive association between E2 and FT (r = 0.28, P < 0.05). Serum FT was related to BMD at femoral neck (r = 0.26, P < 0.05) and Ward's triangle (r = 0.30, P < 0.01) while E2 was related to BMD at anteroposterior (AP) lumbar spine (r = 0.29, P < 0.05), femoral neck (r = 0.23, P < 0.05) and femoral trochanter (r = 0.27, P < 0.05). Besides FT and E2, age, body weight, fat mass and fat-free mass were also correlated to BMD at various skeletal sites. Using stepwise multiple linear regression to control for the confounding effects among these factors, fat-free mass was found to be strongly associated with BMD at most skeletal sites. Serum E2 was related to BMD independently of other factors including FT at AP lumbar spine (r = 0.22, P < 0.05), femoral neck (r = 0.26, P < 0.01), femoral trochanter (r = 0.22, P < 0.05) and Ward's triangle (r = 0.26, P < 0.01) while serum FT was not associated with BMD at any site after controlling for E2 and other related factors. Concerning ER alpha gene polymorphism, 27 (33.3%) of the subjects had pp genotype, while 42 (51.9%) and 12 (14.8%) Pp and PP genotypes, respectively. After controlling for age, body weight, fat mass, fat-free mass, calcium intake, FT and E2, the presence of P allele was associated with higher BMD at AP L2-L4 (P < 0.05). Serum oestradiol is more related to bone mass than free testosterone in normal men. Oestrogen-receptor gene polymorphism is also associated with bone mass in men independently of oestradiol levels. Serum oestradiol together with oestrogen-receptor genotype may partly determine bone mass in males.