Oestrogen, bones and men: when testosterone just isn't enough

It is well established that oestrogen plays a major role in regulating the female skeleton. Thus, cancellous bone mass (assessed at the spine) decreases markedly following oophorectomy or natural menopause in women, and this can be prevented by oestrogen replacement (Riggs et al., 1998). Cortical bone mass (assessed at the radius) also decreases following oophorectomy, although to a much lesser extent. The changes in bone mass following oestrogen deficiency are due to a marked stimulation of bone resorption and a coupled, but smaller increase in bone formation (Riggs et al., 1998). Thus, there is a net remodelling imbalance and hence, bone loss. Although much less studied, sex steroid deficiency in men appears to have a similar effect on the male skeleton. In a study of 12 Czech men imprisoned for sexual delinquency, Stepan et al. (1989) showed that orchidectomy was associated with marked decreases in bone density and a stimulation of bone resorption similar to that observed in women following oophorectomy. Based on these types of data, the traditional view of sex steroid action on the skeleton was that since oestrogen and testosterone were the main sex steroids in women and men, respectively, oestrogen was the major regulator of bone metabolism in women and testosterone played the analogous role in men. As discussed in a previous commentary in this journal by Balestrieri and colleagues (Balestrieri et al., 2001), this paradigm was literally shattered by the description in 1994 by Smith and colleagues of a unique experiment of nature ± the oestrogen receptor (ER) negative male (Smith et al., 1994). This individual had homozygous mutations in the ER gene and presented with unfused epiphyses at the age of 23 years, was tall with a eunuchoid body habitus, and had a markedly delayed bone age of 15 years. Moreover, despite having normal testosterone levels, he had marked osteopenia and elevated indices of bone resorption, presumably since he was unable to respond to oestrogen. Soon thereafter, two additional males with complete deficiency of the aromatase gene were described (Morishima et al., 1995; Carani et al., 1997). These individuals had undetectable oestrogen levels, but again, despite normal testosterone levels, had skeletal phenotypes virtually identical to that of the ER negative male. In addition, oestrogen treatment of the aromatase deficient males led to marked increases in bone density as well as closure of the epiphyses (Carani et al., 1997; Bilezikian et al., 1998). These unique experiments of nature clearly forced a reevaluation of the traditional view of sex steroid action on the skeleton. The skeletal findings in these individuals taught us that, even in boys, oestrogen was necessary for epiphyseal closure, that it had major effects on skeletal modelling during growth, and that it was necessary for the optimal acquisition of bone mass during puberty in boys. While the ER negative and aromatase deficient males led to a shift in our thinking about the role of oestrogen in the male skeleton, they left several major issues unresolved. First, while they left no doubt about the role of oestrogen in the growth and development of the male skeleton, particularly at the time of puberty, it remained unclear what role, if any, oestrogen played in regulating bone remodeling in adult men with mature skeletons. Second, was oestrogen necessary for the maintenance of bone mass in adult men? Finally, what, if any, was the role of oestrogen (or possible oestrogen deficiency) in mediating age-related bone loss in men? To try to answer some of these questions, a number of crosssectional observational studies related bone mineral density in adult and elderly men to sex steroid levels (Greendal et al., 1997; Slemenda et al., 1997; Khosla et al., 1998; Ongphiphadhanakul et al., 1998; Center et al., 1999). In general, these studies did find stronger associations between oestrogen levels and bone mineral density than between testosterone levels and bone density. Moreover, in a recent longitudinal study, our group also found that circulating oestrogen levels correlated better than testosterone levels both with the acquisition of bone mass in young adult men and with bone loss in elderly men (Khosla et al., 2001). Since bioavailable (or nonsex hormone binding globulin bound) oestrogen levels decline markedly with age in men (Khosla et al., 1998), these data also suggested that these decreases may contribute substantially to `agerelated' bone loss in men. Correlation, however, does not prove causality, and the relative contributions of oestrogen vs. testosterone on bone metabolism in men needed to be assessed in a direct interventional study. Thus, we studied 59 elderly men (mean age 68 years) in whom endogenous testosterone and oestrogen production was eliminated using a combination of a GnRH agonist and an aromatase inhibitor (Falahati-Nini et al. 2000).