Abstract

This study examined a method of predicting body density based on hydrostatic weighing without head submersion (HWwithoutHS). Donnelly and Sintek (1984) developed a method to predict body density based on hydrostatic weight without head submersion. This method predicts the difference (D) between HWwithoutHS and hydrostatic weight with head submersion (HWwithHS) from anthropometric variables (head length and head width), and then calculates body density using D as a correction factor. We developed several prediction equations to estimate D based on head anthropometry and differences between the sexes, and compared their prediction accuracy with Donnelly and Sintek’s equation. Thirty-two males and 32 females aged 17 – 26 years participated in the study. Multiple linear regression analysis was performed to obtain the prediction equations, and the systematic errors of their predictions were assessed by Bland-Altman plots. The best prediction equations obtained were: Males: D(g)¼7164.12X17125.81X27111.03X3+ 100.66X4 + 6488.63, where X1¼head length (cm), X2¼head circumference (cm), X3¼head breadth (cm), X4¼head thickness (cm) (R¼0.858, R2¼0.737, adjusted R2¼0.687, standard error of the estimate¼224.1); Females: D(g)¼7156.03X1714.03X27 38.45X378.87X4+ 7852.45, where X1¼head circumference (cm), X2¼body mass (g), X3¼head length (cm), X4¼height (cm) (R¼0.913, R2¼0.833, adjusted R2¼0.808, standard error of the estimate¼137.7). The effective predictors in these prediction equations differed from those of Donnelly and Sintek’s equation, and head circumference and head length were included in both equations. The prediction accuracy was improved by statistically selecting effective predictors. Since we did not assess cross-validity, the equations cannot be used to generalize to other populations, and further investigation is required. Verf.-Referat