Our study evaluated the effects of long-term replacement with rhGH on a population of adult GHD patients with respect to bone density and body composition, assessed by dual energy X-ray absorptiometry (DEXA).
The DEXA analysis showed that vertebral bone density increased in males and decreased in females in the long term, despite a transient BMD gain in the latter during the first years of treatment. No significant changes were detected in both sexes at femoral level, but females showed a trend towards reduction of femoral neck BMD. BMD increased in males at radius and remained stable in females. Regarding body composition, ten years of rhGH replacement induced a reduction of fat mass and an increase in lean mass. However, the effect on lean tissue was documented in both sexes, while fat mass decreased only in males.
The present results support and extend our previous study [22]. In fact, the long-term observation confirms the existence of a gender specificity in the bone response to rhGH replacement, with males displaying a greater BMD achievement that is sustained over time. The evidence that the bone response to rhGH varies as a function of sex, being more favorable in males, was not consistent in all studies. Some authors [16, 20] reported a larger BMD increase in males following long-term GH replacement, while others [23] showed similar or even greater benefits in response to rhGH in women as compared to men. Moreover, Tritos [24] reported a positive association between male sex and bone outcomes, that however did not persist on multivariate analysis. In this still unresolved scenario, our observations support the higher sensitivity of the male skeleton to rhGH replacement and its persistence in the long term.
The mechanisms underlying this gender specificity remain matter of speculation. Growth hormone exerts its effect on bone both directly and through stimulation of IGF-1 production. Since IGF-1 hepatic generation in response to rhGH is lower in females than in males, due to the inhibitory action of circulating estrogens on liver function [27], it has been hypothesized that women are less responsive to rhGH replacement due to lower circulating IGF-1 [24]. According to this view, our study showed that IGF-1 levels were higher in males than in females. However, the percentage of patients reaching their therapeutic goal was similar in males and females, thus excluding the possibility that the different bone response observed in females was influenced by undertreatment. Furthermore, in our previous study [22] we excluded a significant correlation between circulating IGF-1 and bone outcomes. In this regard, it has to be considered that rhGH exerts its osteoanabolic action mainly through increased bone expression of IGF-1 [28], which is not necessarily reflected by circulating IGF-1 levels.
Another factor that has to be taken into account is the direct effect of GH on bone. Growth hormone modulates the crosstalk between osteoblasts and osteoclasts [29, 30], influencing the balance between bone apposition and resorption. High rhGH doses may excessively stimulate bone resorption and limit its favorable effects on the skeleton, at least during the early years of treatment [31]. Since in our study females were prescribed higher rhGH doses than males, this could explain, at least in part, the different bone response observed between sexes.
Besides the direct effects of GH and IGF-1 on bone cells, we can speculate that several bone regulatory metabolic pathways, with a gender related outcome, might be activated by rhGH replacement. Conditions of relative GH excess in comparison to IGF-1 may inhibit osteoblast activity via increased sclerostin levels. Uncontrolled acromegaly is indeed characterized by a nonlinear relationship between GH and IGF-1 due to saturation of GH receptors [32], and uncontrolled acromegalic females displayed higher sclerostin levels as compared to both male patients and healthy controls [33]. This may also have occurred in our female population, characterized by higher rhGH doses and lower IGF-1 levels as compared to males.
The interaction between bone metabolism and changes in body composition could also contribute to the gender difference in bone response to rhGH. The increase in lean mass may mediate the osteo-anabolic action of GH, since muscles possess both a biomechanical and a biochemical action, through a paracrine regulation of several myokines and growth factors including IGF-1 [34, 35]. In particular, the increase in muscle mass is associated with a reduction of the synthesis of myostatin [35], which inhibits osteoblast action and promotes differentiation of cells of the osteoclastic lineage [35]. Moreover, it has been reported that rhGH replacement increases the circulating levels of irisin [36], a myokine able to increase bone cortical mass by stimulating the osteoblast pathways [37], in close relationship with changes in body composition [36]. It can thus be speculated that the more favorable effect on lean mass that we observed in males after rhGH replacement could also have induced a greater bone response.
It has to be taken into account that the existence of a gender specificity in the responsiveness to rhGH replacement with regard to body composition is still a matter of debate [11, 13, 38, 39]; our results are in line with those of Franco et al [11] and Mukherjee et al [13], suggesting that males may experience greater changes in fat mass than females in response to GH. The reason for this difference still needs to be elucidated, but some evidences suggest that androgens may play a major role. Testosterone increases tissue responsiveness to GH through enhanced GHR expression [40], thus potentiating its biological action; moreover, androgens and GH exert additive effects in stimulating protein synthesis, increasing resting energy expenditure and promoting fat oxidation [41], and androgen and GH replacement are both necessary to achieve the full anabolic effect in men with hypopituitarism [41]. On the other hand, the impact of estrogens in this setting seems to be less relevant, as suggested by Franco et al [11], reporting a greater improvement in body composition in response to rhGH in man than in post-menopausal women, characterized by very low circulating estrogens.
Our previous hypothesis [22] that in women the presence of estrogen and the GH-mediated reduction of fat mass could influence the balance between peripheral and central effects of leptin in favor of the latter, enhancing its catabolic action on bone, is not confirmed by the present study; indeed, rhGH replacement reduced fat mass in males but not in females, and the bone anabolic action of leptin is mainly mediated by its circulating levels [42], directly correlating with fat mass.
We did not observe significant variations of femoral bone area following rhGH treatment, in line with our previous work [22]. However, we conducted a preliminary evaluation of the geometrical features of proximal femur with HSA (Hip Structural Analysis), which showed a slight but significant increase in the sub-periosteal diameter of the diaphyseal portion of the femur (p = 0.03) after two years of treatment, not evident in the other femoral sections (data not shown). Unlike femoral neck and intertrochanteric area, femoral diaphysis is composed only by cortical bone [43], and GH exerts its action with a more pronounced effect on cortical bone [28, 44, 45]. It is therefore possible that rhGH replacement induces dimensional changes by activating bone modeling only in specific bone sites and with a limited extent.
Bone area was similar in males and females at all measured sites, both at baseline and after rhGH replacement, thus excluding the hypothesis that the different bone response observed in the two genders was attributable to differences in the activation of bone modeling at periosteal surface.
Since GH acts more on cortical than on trabecular bone [28, 44, 45], it would be expected a greater BMD improvement at femur and proximal radius, predominantly constituted of cortical bone, than at lumbar spine, composed mainly of trabecular bone. However, data from our male population show the opposite, in agreement with several previous studies [16, 20]. Trabecular bone in the lumbar spine is sensitive to sex steroids, while BMD at femur decrease with age in both sexes [20]. It is therefore possible that the different response to rhGH observed at various skeleton sites reflects the interaction between GH and sex steroids.
In our study rhGH therapy did not determine any significant change in BMI in both sexes, in contrast to previous studies [21, 46] reporting an increase of BMI after at least ten years of rhGH treatment. This discrepancy may be related, at least in part, to the lower mean age of the population included in our study as compared to that of previous reports.
Limitations of our work relate to its retrospective nature and to the lack of a control group. In this regard, it has to be considered that a long-term placebo-controlled study would be deemed unethical, given the beneficial effect of rhGH replacement in patients affected by GHD. Some GHD patients referring to our institution spontaneously refused rhGH treatment, but their number was too small to represent an adequate control group. Another factor that could have hampered our evaluation is the limited female sample size, that prevented us from stratifying female patients based on gonadal status. On the other hand, strong points of the study lie in the long observational time and in the fact that all patients were followed in the same institution with a standardized management protocol, samples were analyzed in the same laboratory with the same assays, and DXA scans were performed by the same operator with a standardized procedure.
In conclusion, our findings support the existence of a gender difference in the response to long-term rhGH replacement, with males displaying greater benefits with regard to both body composition and bone mass and retaining them over time. Females showed only a transient and site-specific effect on bone, and a limited positive effect on lean but not fat mass. Potential paracrine and endocrine crosstalks with effect on bone, activated by rhGH replacement and regulated by the concerted action of GH and gonadal steroids, may explain this gender specificity and deserve further investigations.