The relative body fat and anthropometric prediction of body density of South Australian females aged 17-35 years.

One hundred and thirty-five females were tested in order to: produce some normative percentage body fat (% BF) data on an Australian sample which represented a cross-section of physical activity patterns, cross-validate existing multiple regression equations which predict body density (BD) from anthropometric measurements, and if necessary develop population specific equations. Measurements were taken of 10 girths, 3 widths and 7 skinfolds. Body density was measured by underwater weighing with the residual volume (RV) being determined by helium dilution. The Siri equation was then used to convert BD to % BF. The % BF scores had an overall mean of 23.4 (range 10.8-49.2). The very active group (n = 45) had a significantly lower (p less than 0.05) relative body fat (X = 20.6% BF) than either the active (n = 45; 23.5% BF) or sedentary groups (n = 45; 26.2% BF). Previously published equations were found to have limited applicability to Australian subjects. A stepwise multiple regression was therefore used to develop the following equation (R = 0.893): BD(g X cm-3) = 1.16957-0.06447 (log10 sigma triceps, subscapular, supraspinale, front thigh, abdominal and calf skinfolds in mm)-0.00081 (gluteal girth in cm) + 0.0017 (forearm girth in cm) + 0.00606 (biepicondylar humerus breadth in cm). Only those predictors which resulted in a statistically significant increase in r (p less than or equal to 0.05) were included. The standard error of estimate of 0.00568 g X cm-3 was equivalent to 2.6% BF at the mean.