Conversely, in men, lower total T, free T, and weak androgen levels have been associated with increased FM and especially VAT [14, 15]. This has consistently been described in population studies: obese men when compared to non-obese subjects of similar age show lower androgen levels; among overweight and obese with similar BMI (mean 30 kg/m2), men with higher percentage of FM have lower androgen levels [19, 20]. Moreover, healthy eugonadal men undergoing T suppression through GnRH analogues showed an increase in FM and a decrease in FFM, which were partially reversed when T was administered [16, 21–25]; similar results were found among overweight adults undergoing pharmacological castration with GnRH antagonist (acycline) [26].
It is interesting to note that the effects of T on adipose tissue and muscle mass depend on its absolute serum levels in both sexes, rather than the relative levels compared to healthy people. In adults, total T levels between 100 and 300 ng/dL facilitates hypertrophy and hyperplasia of VAT, leading to inflammation and insulin resistance: in females this is associated with PCOS, in males with hypogonadism [11]. Similarly, in male adolescents, the absolute serum levels of T play a crucial role in modulating the body composition. As a matter of fact, the progressive rise in T serum levels (from 100 to >300 ng/dL) drives the bone maturation and raise both the mineral density and the FFM. In obese adolescent, however, serum T is lower (between 50 and 300 ng/dL), while oestrogens are up to twice when compared to non-obese subjects [27, 28].
Preclinical Studies
Tissue mass depends both on number and on dimension of cells; an increase can be therefore due to hyperplasia or hypertrophy. Adipocyte and myocyte share the pluripotent mesenchymal cell as common precursor. Some experiments in rats and mice and cell cultures from men and women showed that their differentiation is influenced by androgens, with T and DHT inhibiting the adipogenic lineage through peroxisome proliferator-activated receptor gamma and CAAT/enhancer-binding protein alpha, while promoting the myogenic lineage through Wnt signaling [29–31]. As to what concerns size, T increases hormone-sensitive lipase expression while reducing lipoprotein lipase in preadipocytes and adipocytes, thus enhancing catecholamine-stimulated lipolysis and limiting fatty acids uptake and esterification to triglycerides [19, 25, 32, 33]. In myocyte, it stimulates myonuclei and protein synthesis, thus leading to hypertrophy and both type 1 and 2 muscle fibres [34].
It is worth noting that a large variability of androgens effect has been reported both in preclinical and clinical studies. Epididymal and retroperitoneal adipose tissue of rats are commonly analyzed as model for human VAT, while the inguinal fat is considered a model for human SAT [35]. In cell cultures of preadipocytes of male rats, T and DHT decrease both epididymal and subcutaneous differentiation, the former being more responsive [29, 30, 36]. In male rats, castration has been associated both to inhibition of epididymal preadipocytes differentiation and increase of perirenal ones [30, 37]. In female ovariectomized rats, T reduced the number of adipocyte in inguinal fat depots more than in retroperitoneal, while reducing the percentage of preadipocytes only in the inguinal region [38]. Although these differences have been partially attributable to sex and type of adipose tissue, a clear inhibitory effect of androgens on lipolysis and adipogenesis in female rats is still to be demonstrated [30].
In vitro studies conducted in adipocytes obtained from transsexuals, before and after cross-sex hormonal therapy owing to the gender reassignment, confirmed results found in animal studies: visceral and subcutaneous adipocytes of male-to-female transsexuals (MtoF) after combined antiandrogen and oestrogens showed a significant reduction in basal lipolytic activity along with an increase in cells size, whereas diametrically opposite results were found in the adipocytes coming from female-to-male transsexuals (FtoM) after T supplementation [39].
The Role of the AR
Given that T acts through the AR, its expression and function could also play a role. In female and male rats, AR and oestrogen receptor (ERα and ERβ) densities vary according to the type of adipose tissue rather than the sex, with VAT showing higher AR density than SAT. Moreover, in vitro studies in mouse-derived cell models (3T3-L1), oestrogens seem to raise both AR and ERα, highlighting the fact that oestrogens can positively modulate both androgens and oestrogens action on adipose tissue; glucocorticoids stimulate the AR expression, even though they inhibit the androgen effects by means of inducing the inactivation of DHT to 5α-androstane-3α,17β-diol (AD) especially in VAT in both sexes [40, 41].
The first exon of AR is involved in the transactivation of the receptor and is characterized by a polymorphism: a higher number of CAG repeats is associated with lower transcriptional activity [42]. In adolescent males, low and intermediate CAG repeats is associated with higher VAT; in adult men, it negatively correlates with muscle mass but not with VAT [43]. Unfortunately, results on this topic are still conflicting, as other studies found no correlation between AR polymorphism and body composition [44–46].
Androgen Metabolism in Adipose Tissue
In adipose tissue, one of the main steroid-converting enzymes is aldo-keto reductase 1C (AKR1C) family which encompass 3 members (AKR1C1, AKR1C2, and AKR1C3) showing different substrate specificity. In men, AKR1C2 (3α-hydroxysteroid dehydrogenase), in particular, converted