Accuracy of residual volume prediction--effects on body composition estimation in pulmonary dysfunction

The accuracy of predicting residual volume (RV) and the effect of employing such predictions in the assessment of body composition by hydrostatic weighing was examined. Prediction equations derived by Wilmore (Med. Sci. Sports, 1969) and Boren et al. (Am. J. Med., 1966) were cross-validated on 400 male subjects with varying degrees of pulmonary dysfunction. Statistical analysis included comparison of predicted vs measured RV and regression analysis of measured on predicted values. Neither equation met acceptable statistical criteria (p greater than .05). Underwater weights were calculated based upon the measured RV and an assumed standard body weight, body density, and percent body fat (i.e., 70 kg, 1.0647 g/cc and 15.0%). Predicted RV was then substituted in the calculation and the effect on standardized values examined, between the calculated and standardized percent fat value with both equations. Results suggest in the present subjects, RV cannot be accurately predicted and the use of such prediction equations introduces sizable errors in body composition estimation.