The use of BMI as a proxy for adiposity, the true determinant of the obese state, has been criticised, given that body weight is the sum of individual organs and tissues and therefore it includes adipose tissue, skeletal muscle mass and organs mass. On a population level, a strong positive correlation between BMI and overall body fat content has been reported [3]. However, at an individual level, a substantial variation in percentage body fat may be observed for any given BMI value [4]. Therefore, a high BMI may correspond to a low fat-free mass and a substantial fat accumulation in an obese patient, or to a large skeletal muscle mass and normal fat mass in a healthy athlete, in which high BMI simply reflects increased muscle mass, which has nothing to do with obesity and associated diseases. Visual inspection is usually sufficient to discriminate these extremes in body composition, but in some subjects the distinction may be more subtle, and a more precise determination of body composition may be requested. Fat mass and fat-free mass may be reliably distinguished and measured by direct densitometric methods (underwater weighing; total body densitometry). Dual-energy X-ray absorptiometry (DEXA) has a good reproducibility for total body fat mass (coefficient of variation: 2–3 %) and total body fat-free mass or lean soft tissue (1–2 %), and it is sensitive in assessing minimal changes in body composition [5]. Unfortunately, DEXA is not applicable in routine office practice; it is costly and exposes patients to a low dose of radiations. Bioelectrical impedance analysis (BIA) is an indirect method that derives body composition values from electrical data (reactance, resistance, impedance) measured during the passage of a small electrical current through the patient’s body [6]. The method is applicable also in the outpatient setting and does not have virtually potential side effects, and it is relatively not expensive. However, the reliability of BIA in the accurate determination of fat mass and fat-free mass may be questioned. BIA measurements must be standardised in order to obtain reproducible results (reported mean coefficients of variation for within-day measurements: 1–2 %), and overall reproducibility/precision is estimated around 2.7–4.0 %, with prediction errors for FFM ranging from 3 % to 8 % [7]. These large errors may be even larger in obese patients and limit the utility of BMI in clinical evaluation.

A further important limitation for BMI is that this index does not convey any information on fat distribution (e.g. visceral fat accumulation and fatty infiltrations in individual organs) that is considered now an important determinant of metabolic and cardiovascular risk [8]. A clear evidence in favour of the inclusion of fat distribution in the clinical evaluation comes from the observation of normal-weight or slightly overweight subjects with low subcutaneous but increased visceral fat mass. This TOFI (thin-on-the-outside fat-on-the-inside) sub-phenotype has been observed in both male and female subjects and increases an individual’s risk of metabolic disease [4]. The elevated visceral fat found in individuals classified as TOFI is accompanied by increased levels of ectopic fat deposition both in the liver and in the skeletal muscle. Lipid accumulation in non-adipose cells (ectopic fat) may impair the normal function of some tissues through a process known as “lipotoxicity”. Ectopic storage of excess lipids in organs such as the liver, skeletal muscle, and pancreatic beta cells may be the causative link between fat distribution and the metabolic syndrome or cardiovascular diseases [9]. Similar findings have been already reported in obese individuals, where obese subjects with a disproportionate accumulation of visceral fat had increased incidence of metabolic disorders and cardiovascular events [10].

Visceral fat accumulation may be measured precisely with CT and MRI, but it may be difficult to quantify at a clinical level, and surrogate anthropometric indexes have been proposed. In particular, the waist circumference has been selected as a reliable clinical indicator of visceral fat accumulation, and having a large waist is associated to a higher prevalence of metabolic disorders and cardiovascular diseases [11]. Therefore, the measurement of the waist circumference is suggested for the determination of cardiovascular risk of overweight and obese patients, and the integration of BMI and waist values may be used to better stratify their health risk [11] (Table 15.2). Waist circumference should be measured with a plastic stretch-resistant tape on the subject in the standing position, at the end of a gentle expiration, without constricting the abdomen. Different anatomic landmarks have been suggested for waist measurement [12]. According to WHO guidelines, waist circumference should be measured at the approximate midpoint between the lower margin of the last palpable rib and the top of the iliac crest [13]. The US National Institutes of Health (NIH), by applying the same method used for the US National Health and Nutrition Examination Survey (NHANES) III, indicates that waist circumference measurement should be made at the top of the iliac crest [11]. The two methods did not produce the same results, with the WHO method underestimating waist values in respect to the NIH method, particularly in women [12]. It should be emphasised that the cut-off values proposed for “at-risk” waist values (Table 15.2) and utilised for the original ATP-III definition of the metabolic syndrome are those proposed by the NIH. The simple measurement of waist circumference has replaced the use of the waist-to-hip circumference ratio (WHR), originally proposed as a powerful marker of fat distribution. More recently, on the basis of several epidemiological studies showing that having a large hip circumference may confer some BMI-independent protection from metabolic and cardiovascular diseases, particularly in women, a return to the measurement of hip circumference has been proposed [14]. The reliability of waist circumference in assessing visceral fat accumulation may be reduced in obese women, particularly at higher BMI levels [15]. Other anthropometric indexes have been therefore suggested as more effective than waist circumference for the prediction of visceral fat depots, with the sagittal abdominal diameter (SAD) being the more promising one [16]. SAD is determined at the highest point of the abdominal surface with the subject in the supine position and during normal breathing by means of a specifically made instrument. Abdominal ultrasonography is another reliable, repeatable and less expensive method which has been proposed to detect visceral fat deposition without radiation exposure [17]. Peritoneal fat thickness is considered the gold standard echographic index for visceral fat prediction in abdominal ultrasonography, and it corresponds to the distance from the internal face of the recto-abdominal muscle and the anterior wall of the aorta, measured with the echographic probe transversely placed perpendicular to the skin in the midline of the abdomen [17]. The increasing availability of portable low-cost ultrasonographic instruments will probably stimulate the applicability of ultrasonographic measurements of visceral fat accumulation in clinical practice.

The presence of ectopic fat deposition in the relevant organs may be even more difficult to quantify than visceral fat accumulation in clinical practice. However, liver fat infiltration (hepatic steatosis) may be roughly, albeit imprecisely, estimated by ultrasound [18]. An alternative approach to the quantification of ectopic fat accumulation may be represented by the ultrasonographic measurement of epicardial fat, which has been suggested as a further marker of metabolic and cardiovascular risk [19].