Document Type



Criterion values for body composition include N and K measurements and (in epidemiological surveys) Wt/ht2. Methods of estimating body fat (skinfold data and hydrostatic weighing) have been evaluated against these criteria in a sample of 12 older but moderately active female subjects (age 48-58 years, average body fat by hydrostatic weighing 30.1%). In this age group, the hydrostatic estimate is less closely correlated with criterion values than is a skinfold estimate. It is suggested that this reflects difficulties in determining the density of older subjects, and uncertainties concerning change in the density of lean tissue with age. Potassium estimates of body fat are strengthened by simultaneous consideration of N and Ca data, while N estimates are strengthened by consideration of K and Ca figures. Nevertheless, the simple anthropometric criterion (wt/ht2) yields a higher correlation with skinfold data than do estimates derived from N and/or K, in part because of uncertainty regarding age-related changes in body composition. Moreover, wt/ht2 estimates of body fat cannot be strengthened by considering whole body counter estimates of body fat. Within the limitations imposed by this small sample of healthy older women, anthropometric data thus seem preferable to either hydrostatic or whole body counter estimates of body fat.