Document Type



Studies of subcutaneous fat site choice have focused on the site which has the greatest correlation with other subcutaneous areas or with body weight or percent body fat. While it is important to obtain a good predictor of overall fatness, this procedure may miss a related aspect of human fatness: the anatomical distribution of fat. It is a determinant of bodily configuration of possible importance in the predisposition of overweight persons to chronic disease. Leg fat (below the knee) has tended to be ignored in previous studies of subcutaneous fat site selection, because of its low association with other indicators of fatness. Using principal components analysis, we show that leg fat may be the most important indicator of individual differences in anatomical distribution of fat. Principal components analysis is a useful analytic tool for discovering the pat­tern of associations inherent in a series of related measurements, but is cumber­some in practice and implies the necessity of using a large number of mea­surements. We show that a two-skinfold index is a sufficient predictor of the major principal component of anatomical fat distribution (trunk-extremity). The index requires that leg and trunk fat be contrasted. The two most currently used sites- triceps and subscapular—may be appropriate as measures of fatness, but do not discriminate well among individual differences in anatomical fat distribution.