Validity and reliability of body composition assessed by the sulphur hexafluoride dilution method

The aims of this study were to assess the validity and reliability of body volume and percent body fat determined by sulphur hexafluoride dilution, using underwater weighing as the criterion method, and to determine the influence of the magnitude of body volume. Thirty-one healthy Japanese individuals aged 18-27 years (16 males: height 1.70 +/- 0.06 m, mass 64.8 +/- 7.7 kg; 15 females: height 1.60 +/- 0.05 m, mass 55.2 +/- 6.2 kg; mean +/- s) participated in the study. Sulphur hexafluoride dilution measures the concentration of sulphur hexafluoride gas in the chamber (BSF-200, Shimazu Corp.). Underwater weighing was performed five times using a weight scale (AD-6204, A&D) after residual volume had been determined (System9, Minato Medical Corp.). There were no significant differences in the mean between two trials for body volume, body density or percent body fat determined by sulphur hexafluoride dilution. The intra-class correlation coefficient of these variables ranged from 0.985 to 0.999. The results suggest that sulphur hexafluoride dilution is a reliable method for assessing body composition. There was no significant difference in body volume or percent body fat between sulphur hexafluoride dilution (males: 61.3 +/- 7.6 litres, 18.4 +/- 6.7%; females: 52.8 +/- 6.9 litres, 21.0 +/- 8.9%) and underwater weighing (males: 60.6 +/- 7.0 litres, 15.6 +/- 3.5%; females: 53.0 +/- 6.5 litres, 23.7 +/- 6.1%) and there was a high correlation between the two (r = 0.997, P < 0.05). A Bland-Altman plot of the difference between percent body fat estimated by underwater weighing and sulphur hexafluoride dilution versus average percent body fat by the two methods showed no systematic difference (mean difference = -0.12 +/- 6.6 kg). The upper and lower limits of agreement were 13.2% and -13.4%, respectively. Determination by sulphur hexafluoride dilution resulted in both over- and underestimations in body volume and the difference between the two body volumes (determined by underwater weighing and by sulphur hexafluoride dilution) was inversely proportional to the mean body volume by the two methods. This suggests that improvements need to be made to the device or to the technique to maintain a constant volume of sulphur hexafluoride in the chamber.