Testosterone: More Is Not Always Better

Abstract

Having conducted population-based studies on the relationship of sex steroid levels to bone mass in Rochester, Minnesota for many years, I have been struck by just how often male members of the medical staff at our institution who participate in these studies inquire about their testosterone level. Discovering that they happen to be in the top 10th percentile is often a cause for celebration; conversely, being told that they happen to be “average” or, most distressingly, “below average,” generally leads to at least temporary anguish. Not surprisingly, despite the numerous publications from our group and others regarding the key role for estrogen in bone metabolism in men (summarized in Ref. 1), none of my colleagues inquire or seem to care one way or another about their estradiol level. At least based on this anecdotal experience, it seems that even in the minds of well-informed male medical professionals, more testosterone is clearly “better,” whereas estrogen is largely irrelevant. It is in this context that men with modest testosterone levels should perhaps have their chance to celebrate based on the results of the study by Burnett-Bowie et al. (2) published in this issue of JCEM. This carefully conducted clinical trial clearly demonstrates that at least for bone, more testosterone is not necessarily better, especially if it comes at the price of lower estrogen production. The trial is based on the premise that because male aging is associated with declining testosterone levels and with increased body fat, decreased muscle mass, strength, and bone mineral density (BMD) (1), replacing testosterone in elderly men should be beneficial. However, rather than administering exogenous testosterone, which requires topical or parenteral dosing and significantly increases estrogen levels (via peripheral aromatization) and which may, in turn, result in side effects such as gynecomastia, the authors chose to test the effects of increasing endogenous testosterone production by blocking aromatization of testosterone to estradiol using a potent orally administered aromatase inhibitor, anastrozole (2). By removing negative feedback signals, the anastrozole-induced decrease in estradiol increases GnRH and gonadotropin production at the hypothalamus and pituitary, respectively, thereby increasing testosterone production (3). The Burnett-Bowie et al. (2) paper reports on 69 men, aged 60 and older, who completed the 1-yr, double-blind, randomized, placebo-controlled trial. Anastrozole had the desired effects on serum testosterone levels, with mean values increasing from 319 ng/dl (the lower end of the normal range) at baseline to 524 ng/dl at 3 months and a slight decline to 474 ng/dl at 1 yr, an overall increase of approximately 50%. As expected, estradiol levels decreased (by 20%), from 15 to 12 pg/ml at 3 months, and remained stable at this level thereafter. However, despite the substantial increase in serum testosterone levels, posterior-anterior (PA) spine BMD decreased by 1.7% in the anastrozole group while increasing by 0.8% in the placebo group (P 0.0014 for the difference between groups), with similar trends for differences between the groups for femoral neck, total hip, and total body BMD. The inescapable conclusion from these data is that the 50% increase in testosterone levels was trumped by the much more modest 20% reduction in estradiol levels, leading to negative skeletal effects of anastrozole therapy in aging men. To some extent, these findings are not surprising and are perhaps predictable, based on the now extensive body of evidence for a critical role for estrogen in regulating the male skeleton (1). Although previous studies in men had shown that serum estradiol levels were related to bone density (4) and that estrogen regulated bone turnover (5,