Objective: To compare methods for the assessment of visceral fat with computed tomography (CT) and establish cutoffs to define visceral obesity based on such alternative methods.
Research methods and procedures: One hundred women (50.4 +/- 7.7 years; BMI 39.2 +/- 5.4 kg/m2 underwent anthropometric evaluation, bioelectrical impedance, DXA, abdominal ultrasonography (US), and CT scan.
Results: Waist circumference, waist-to-hip ratio (WHR), and US-determined visceral fat values showed the best correlation coefficients with visceral fat determined by CT (r = 0.55, 0.54, and 0.71, respectively; p < 0.01). Fat mass determined by DXA was inversely correlated with visceral-to-subcutaneous-fat ratio (r = -0.47, p < 0.01). Bioimpedance-determined fat mass and skinfolds were correlated with only subcutaneous abdominal fat quantified by CT. Linear regression indicated US visceral-fat distance and WHR as the main predictors of CT-determined visceral fat (adjusted r2 = 0.51, p < 0.01). A waist measurement of 107 cm (82.7% specificity, 60.6% sensitivity) and WHR of 0.97 (78.8% specificity, 63.8% sensitivity) were chosen as discriminator values corresponding with visceral obesity diagnosed by CT. A value of 6.90 cm for visceral fat US-determined diagnosed visceral obesity with a specificity of 82.8%, a sensitivity of 69.2%, and a diagnostic concordance of 74% with CT.
Discussion: US seemed to be the best alternative method for the assessment of intra-abdominal fat in obese women. Its diagnostic value could be optimized by an anthropometric measurement. Prospective studies are needed to establish CT and US cutoffs for defining visceral-fat levels related to elevated cardiovascular risk.