Quantification of the relative contribution of estrogen to bone mineral density in men and women

That estrogen plays a critical role in bone health in women is a well-known fact. Deficiency of estrogen leads to increased bone loss, lower bone mineral density (BMD), and increased fracture risk [1]. In women with reduced bone density and/or an existing fracture, treatment with estrogen could reduce bone loss, and lower fracture risk [2]. Thus, the BMD-mediated causal relationship between estrogen and fracture in women has been well established.

However, recently emerging evidence has suggested that estrogen also has important effects on bone health in men. Almost 20 years ago, Smith and colleagues reported a 28-year old man with a bone age of 15 years, who had estrogen resistance due to disruptive mutation of the estrogen receptor gene [3]. The man had normal level of testosterone and elevated estradiol, but had severe osteopenia associated with increased values of bone turnover markers. Subsequently, two young men with undetectable estradiol levels due to mutation in the CYP19A1 gene (this gene is responsible for converting androgens to estrogens) were identified [4, 5]. The two men also had low bone density as initially reported by Smith et al.; however, when the men were treated with estrogen, their BMD increased. These cases have demonstrated that estrogen does play an important role in the skeletal maturation and mineralization in men, and changes the traditional view that estrogen is a “female hormone”.

A series of subsequent case reports have indicated that estrogen is essential for the normal growth and maturation of the skeleton in men with aromatase deficiency and low serum estradiol (E2) levels [4‐7]. Some cross-sectional studies have further demonstrated a significant positive relationship between serum E levels and BMD in men [8‐14]. Most of these studies were based on elderly populations of Caucasian background, and as a result, it is difficult to assess the relative contribution of estrogen to the inter-individual variation in BMD in the general population.

Although these studies collectively suggest that estrogen might be an important factor for bone growth and bone maintenance in men, most studies have not considered the simultaneous effects of estrogen and testosterone on BMD in men. The question of interest is therefore: what is the proportion of variation in BMD among men and women of Asian background that can be attributed to the variations in estrogen and testosterone. The present study was designed to address that research question by analyzing the relative contributions of serum levels of estradiol and testosterone to the variation in BMD in Asian men and women.

The study involved 205 men and 432 women, aged between 18 and 87 years (Table 1). The average age among men was 44, which was slightly lower than that among women (48 years). Fifty-one percent of men and 0.7% of women reported being current smokers. Approximately 20% men and 13% women were obese (body mass index greater than 25 kg/m²). On average, men had significantly lower E2 and higher TT concentrations than women. The mean value of TT was 487 pg/mL in men and 20 pg/mL in women; and that of E2 was 28 pg/mL and 80 pg/mL in men and women, respectively.

The traditional view of sex hormones and bone health is that estrogens are the main sex steroids affecting bone maturation in women, and that androgens are the corresponding sex steroids in men. However, this view has been challenged by an initial case report [3] and subsequent reports [4, 5] which showed that the “female hormone” is actually important for bone mineralization and bone growth in men. However, such a relationship has not been well documented in Asian men, and it is not clear of the magnitude of effect of estrogen in bone mass in men. The present study on Asian men shows that estrogen, not testosterone, was an independent determinant of bone density. However, the contribution of estrogen to the variance of bone density in men is likely to be modest.

Our finding of positive association between estradiol and bone mass in men is largely consistent with previous studies in Caucasian men. Indeed, several observational studies in older Caucasian men have shown that circulating free estrogen or total estradiol, not testosterone, was associated with bone density [18‐20] and bone loss [14, 21‐23]. Our finding is also in line with a recent observation that in Chinese men, low levels of bioavailable E2 were associated with low bone density and greater bone loss [24]. Our results further show that the magnitude of association between estradiol and BMD in men was comparable with or greater than that in women. Collectively, the present data and evidence to date support the conjecture that estradiol is probably more important than testosterone as a determinant of bone growth and bone maintenance in men and women alike.

The mechanism of effect of estrogen on bone is well known. Deficiency of estrogen leads to an increased osteoclastic formation, expanding the remodeling space, increased cortical porosity, and enlarged the resorption area on trabecular surface [25, 26]. The net result of these effects is bone loss. Indeed, estrogen deficiency is associated with an increased in bone resorption marker in men [14] and women [27]. Moreover, treatment of elderly men with an aromatase inhibitor resulted in increased bone resorption and decreased bone formation [28]. Treating osteoporotic men with selective estrogen receptor modulator increased femoral neck BMD [29]. Taken together, these studies suggest that estrogen is causally linked to bone mass in men, and that estrogen might be more important for bone health than are declining levels of testosterone.

However, it can not rule out the role of androgens in maintaining skeletal integrity and bone homeostasis in men. Testosterone contributes to the initiation of bone resorption and maintenance of bone formation. Moreover, the majority of estrogens in elderly men is derived from androgens by peripheral conversion [30]. The gene involved in the conversion is the CYP19A1 gene. Men with TTTA repeats in the gene had higher estradiol levels and reduced bone loss than those with lower number of such repeats [31].

Despite the established relationship between estrogen and bone density in men, the magnitude of this association has not been quantified. In this study, by using a sophisticated statistical technique, we were able to estimate that estradiol explained only 1.2 to 2.5% of between-individual variation in BMD. The upper estimate of attributable proportion was ~8%. This relative attribution was much lower than body weight which accounted for 20% of variation in BMD. Most of the association between body weight and BMD is due to muscle mass [32, 33]. The dominant contribution of body weight and modest attribution of estrogen imply that in men physical fitness may be more important than sex hormones as a factor for maintaining bone mass in the old age.

In our study, serum levels of testosterone in men remained fairly stable while there was a linear decline of estradiol levels with advancing age in men with modest magnitude of decrease. We found perhaps not surprising that the age-related decrease in estradiol levels in women was greater than that in men, such that before the age of 50, estradiol concentrations in women was higher than that in men, but after the age of 50 years, estradiol levels in women was lower than in men. This finding is consistent with previous studies [21, 34‐37].

The present results should be interpreted within a number of potential strengths and weaknesses. The study participants were randomly drawn from the general population which should enhance the findings’ external validity. The DXA measurement of BMD is considered gold accurate and reliable, which ensure the study’s internal validity. Moreover, the technique of measurement of estradiol and testosterone was based on the novel Elecsys automated assay, which has been shown to be a precise method for measuring sex hormones over a wide reportable range in serum. Indeed, recent studies have shown that measurement of estradiol and testosterone by this method was highly concordant with the liquid chromatography tandem mass spectrometry methods [38‐40]. Nevertheless, we did not measure SHBG, free estradiol and free testosterone in the study. The study was designed as a cross-sectional investigation, and as a result, no causal inferences could be made for the observed relationships between estradiol and bone density. In this study we did not ascertain the phase of menstrual cycle which could affect the measured value of estradiol within an individual. However, the intra-invididual variability was statistically considered as a component of the random error in the multiple linear regression model.

In summary, these data suggest that estradiol, but not testosterone, was a significant determinant of bone mineral density in Asian men and women. However, the relative contributions of estradiol to the between-individual variation in bone density is likely to be modest for men. This finding is consistent with the hypothesis that estrogen is a key hormone in the skeletal maturation, bone mineralization and maintenance of bone mass in men and women. The finding also provides further supportive evidence for the “unitary model” of pathogenesis of osteoporosis [22].