Despite numerous reports that abdominal obesity is related to disease risk, the study of body-fat distribution remains a largely empirical science. Elucidation of the pathophysiologic linkage between abdominal obesity and disease would be better served by generating hypotheses and testing them--the process of deductive science. Physiologists have proposed that adipose tissue of the intra-abdominal compartment (that drained by the portal vein) contributes strongly to atherosclerosis. If this is so, then the volume of the intra-abdominal fat depot might be better correlated with disease states than a less specific anthropometric index such as the waist-to-hip girth ratio. Alternative abdominal-obesity indices (e.g. sagittal abdominal diameter divided by thigh girth) could be tested in epidemiologic studies to improve our pathophysiologic understanding of how body-fat distribution is related to atherosclerosis and other diseases.