Estimation of computerized tomography derived abdominal fat distribution

Abstract

Measurements of intra-abdominal fat (IAF) may be important since it is associated with numerous metabolic disorders. The relationship between computerized tomography (CT) measured fat distribution and densitometry measures was investigated in a sample of 61 male Caucasian subjects, aged 18 to 30 years with varying adiposity. Regression models were developed for estimating CT-derived fat of 40 men to estimate IAF. Two equations were developed to estimate IAF. The first used only anthropometric measures. Waist circumference and log chest ratio entered the equation and accounted for 67% of the variance. The second model included densitometry-measured percentage fat with the centred product of waist and hip circumferences, accounting for 73% of the variance. Regression equations were also developed to estimate subcutaneous fat area so that the ratio of IAF to subcutaneous fat might be estimated. Although subcutaneous fat could be estimated, the ratio between IAF and subcutaneous fat could not be estimated accurately. A validation of all regression equations developed for male subjects who also completed using a separate validation sample (n = 21). Only the studies with sample characteristics similar to those found in the validation sample validated satisfactorily. Results indicate that anthropometric and densitometry measures cannot be used to estimate CT-derived abdominal fat with precision, however they may be of value in health risk screening of individuals with high levels of IAF. Proper selection procedures with regard to age, adiposity, and morbidity must be used.