ones), including dehydroepiandrosterone, dehydroepiandrosterone-sulphate and androstenedione (A), which begin to be active as far as converted into T and DHT. Recently, it has been pointed out that the major weak androgen is 11β-hydroxyandrostenedione, which is synthetized by the adrenal glands and can be converted both into 11β-hydroxytestosterone and, thereafter, into 11β-hydroxydihydrotestosterone by the 5α-reductase in target tissues; these last 2 androgens bind the AR as well, although being less active than T and DHT [1].
The synthesis of androgens in the gonads is regulated by the luteinizing hormone (LH), secreted by the pituitary gland upon the hypothalamic action of the gonadotropin-releasing hormone (GnRH). Indeed, LH acts on Leydig cells in males and theca cells in females via the steroidogenic acute regulatory protein, promoting the transfer of cholesterol to the inner mitochondrial membrane. Despite common pathways, androgens show sex-dependent features. In males, testes synthesize T, which is converted into DHT in target tissues, such as prostate and seminal vesicles. In females, ovaries synthesize about the half of the circulating T and smaller quantities of dehydroepiandrosterone and A; the remaining T is produced from the conversion of circulating weak androgens (mainly from A). The secretion of adrenal androgens is commonly regarded as being LH-independent and is predominantly characterized by the synthesis of weak androgens in both sexes.
Given that AR is present, hormones can exert their effect in the vicinity of the secreting gland or enter the blood stream. In both sexes, A and T can be converted into oestrogens by aromatase in different tissues, including visceral adipose tissue (VAT). The principal end-metabolites of androgens is androsterone glucuronide (ADTG) and 5α-androstane-3α,17β-diol glucuronide (ADG), produced mainly in splanchnic tissues (gut and liver) rather than in peripheral ones (skin and male internal genitalia). Extra-gonadal metabolic pathways are not influenced by LH but are modulated either by the characteristics of local enzymatic pathways and their co-regulators (i.e., IGF-1) or by the type and quantity of androgens locally present [2, 3].
In males, androgens are responsible for the development of primary and secondary sexual characteristics, whereas in both sexes, they influence sexual behaviour, glycaemic control, lipid profile, bone metabolism and erythropoiesis [4–6]. The present paper aims at reviewing androgen metabolism and its influence on body composition in both sexes.
Methods for the Determination of Body Composition: Benefits and Limits
Body mass can be divided into the following components: (i) in a 2-component model, fat mass (FM) and fat-free mass (FFM); (ii) in a 3-component model, FM, bone and lean tissue mass (LTM); (iii) in a 4-component model, FM, bone, protein and total body water. Several approaches have been proposed in order to estimate body composition: the anthropometric methods include body mass index (BMI), waist circumference and skinfolds measurement; the imaging methods consist of dual energy X-ray absorptiometry (DXA), bioelectric impedance, ultrasonography (US), computed tomography (CT), and magnetic resonance imaging (MRI).
The ratio between the body weight (kilograms) and the height (meters square) defines BMI. As recommended by the World Health Organization, a subject can be defined as being underweight whenever the BMI is less than 18.5 kg/m2, as being normal if the BMI is between 18.5–24.9 kg/m2, as overweight if 25–29.9 kg/m2 and as obese if ≥30 kg/m2. It is commonly used as a measure of adiposity; however, it cannot discriminate neither between FM and FFM, nor between subcutaneous adipose tissue (SAT) or VAT. Its utility is confirmed in public health care setting, but when considering the individual subjects, one seems to be limited by the low sensibility: in a study carried out on 1,393 adults, BMI misclassified 39% as non-obese when compared to DXA [7–9].
Waist circumference can be measured according to the following methods: at the superior border of the iliac crest with cut-off of 102 cm in men and 88 cm in women, according to National Cholesterol Education Program Third Adult Treatment Panel; midway between the lowest ribs and the iliac crest, with cut-off of 94 cm in men and 80 cm in women, according to International Diabetes Federation. It is generally regarded as an accurate measure to estimate FM. Its limitations include the impossibility of distinguishing SAT from VAT and the evaluation of only one parameter, not accounting, for example, for height and therefore overestimating FM in taller people [10]. Notwithstanding these demerits, BMI and waist circumference are the most accurate surrogate markers of visceral adiposity in young and middle-aged male and female in clinical practice, both being valuable parameters of insulin resistance, hepatic steatosis and serum T [11, 12].
The skinfold method is based on the measurement of thickness at triceps, midaxilla, supraspinale levels in men, while at levels of midaxilla, biceps, medial calf and abdominal in women; the obtained data is input in an equation depending on sex in order to estimate FM. Although it is a low-cost method, it is time consuming and has low reproducibility when performed by untrained personnel: the SAT can be compressed up to 40% and measuring 1 cm apart from the standardized point may induce significant error. Other anthropometric methods include body adiposity index, waist-to-hip ratio and waist-to-height ratio.
DXA is based on 2 X-ray beams absorbed by tissues. It is characterized by a high accuracy in estimating FM, SAT, VAT, and LTM when compared to CT; it is fast, easy to use and widely available. Limitations include body mass limit (200 kg), height and width limits (197 × 66 cm), radiation exposure limiting its use in paediatric, pregnancy and several measurements in a short time.
Bioelectric impedance is based on the measuring of resistance to a small electrical current when passing through the body tissues. It can distinguish between FM and FFM, although a tendency to underestimate FM was reported both in lean and obese subjects. Some authors report a high accuracy for appendicular composition, lower values for the trunk. It has the advantages of portability and non-invasiveness; limitations include high variability depending on hydration status, fasting, physical exercise, coffee and alcohol.
B-mode US is based on ultrasonic waves emitted by a probe and reflected by tissues. It can be used to measure subcutaneous and visceral thickness, without compression limiting the skinfold method. It has been shown to be comparable to DXA, but accuracy depends on training [7].
Finally, CT and MRI are the most accurate indirect methods to estimate body composition. Their use is strongly limited by radiation exposure and costs respectively. Furthermore, none of them can normally be used in grade III obese subjects (BMI ≥40) due to the limitation of scanner width [8, 13].
In conclusion, the method should be defined with a patient-centred approach, depending on the outcomes to be evaluated and patient’s characteristics.
Fig. 1. Body composition across life in healthy men and levels of sexual hormones. Sexual hormones were indirectly evaluated either by RIA involving chromatography steps (before the 2000s) or by mass spectrometry [20,
