PURPOSE: To compare measurements of body density (D(b)) obtained from air displacement plethysmography (AP) and hydrostatic weighing (HW) and to determine the accuracy of substituting D(b) via AP (D(b)-AP) for D(b) via HW (D(b)-HW) in estimating body fatness (%Fat(4C)) and the composition and density of the fat-free mass (Dffm) from a four-component model (fat, mineral, water, and protein). METHODS: D(b) was measured in 50 young adults using AP and HW. Total body water via deuterium dilution, bone mineral content via dual-energy x-ray absorptiometry, and D(b) were used to estimate %Fat(4C). RESULTS: D(b)-AP and D(b)-HW were highly correlated (r = 0.89, SEE = 0.008 g x mL(-1)), but D(b)-AP (1.065 +/- 0.003 g x mL(-1)) was significantly higher (P < 0.05) than D(b)-HW (1.058 +/- 0.003 g x mL(-1)), resulting in a mean difference of 2.8%fat. Differences between %Fat(4C-AP) (17.8 +/- 1.2%) and %Fat(4C-HW) (19.3 +/- 1.2%) were significant (P < 0.05), but the SD of the differences (2.3%) was low. When D(b)-AP was used in a four-component model in place of D(b)-HW, the calculated Dffm was significantly higher (1.109 +/- 0.002 vs 1.105 +/- 0.002 g x mL(-1)) based on a higher (P < 0.05) protein fraction (22.0 +/- 0.4% vs 20.6 +/- 0.4%) and lower (P < 0.05) water (71.1 +/- 0.4% vs 72.4 +/- 0.4%) and mineral fractions (7.0 +/- 0.1% vs 7.1 +/- 0.1%). CONCLUSIONS: AP yields a higher D(b) than HW and may not be a valid method for measuring D(b) or estimating %fat using densitometry. However, due to relatively small bias and low individual error, D(b)-AP is an acceptable substitute for D(b)-HW when estimating %fat with a four-component model in young adults.
PURPOSE: To compare measurements of body density (D(b)) obtained from air displacement plethysmography (AP) and hydrostatic weighing (HW) and to determine the accuracy of substituting D(b) via AP (D(b)-AP) for D(b) via HW (D(b)-HW) in estimating body fatness (%Fat(4C)) and the composition and density of the fat-free mass (Dffm) from a four-component model (fat, mineral, water, and protein). METHODS: D(b) was measured in 50 young adults using AP and HW. Total body water via deuterium dilution, bone mineral content via dual-energy x-ray absorptiometry, and D(b) were used to estimate %Fat(4C). RESULTS: D(b)-AP and D(b)-HW were highly correlated (r = 0.89, SEE = 0.008 g x mL(-1)), but D(b)-AP (1.065 +/- 0.003 g x mL(-1)) was significantly higher (P < 0.05) than D(b)-HW (1.058 +/- 0.003 g x mL(-1)), resulting in a mean difference of 2.8%fat. Differences between %Fat(4C-AP) (17.8 +/- 1.2%) and %Fat(4C-HW) (19.3 +/- 1.2%) were significant (P < 0.05), but the SD of the differences (2.3%) was low. When D(b)-AP was used in a four-component model in place of D(b)-HW, the calculated Dffm was significantly higher (1.109 +/- 0.002 vs 1.105 +/- 0.002 g x mL(-1)) based on a higher (P < 0.05) protein fraction (22.0 +/- 0.4% vs 20.6 +/- 0.4%) and lower (P < 0.05) water (71.1 +/- 0.4% vs 72.4 +/- 0.4%) and mineral fractions (7.0 +/- 0.1% vs 7.1 +/- 0.1%). CONCLUSIONS: AP yields a higher D(b) than HW and may not be a valid method for measuring D(b) or estimating %fat using densitometry. However, due to relatively small bias and low individual error, D(b)-AP is an acceptable substitute for D(b)-HW when estimating %fat with a four-component model in young adults.