Abstract
New equations were derived to predict the density of the body (DB) by hydrostatic weighing with the head above water (HWHAW). Hydrostatic weighing with the head below water (HWHBW) was the criterion for DB measurement in 90 subjects (44 M, 46 F). Head volume by immersion (HVIMM) was determined by subtracting the mass in water with the head below water (MWHBW) from the mass in water with the head above water (MWHAW), with subjects at residual lung volume. Equations were derived for head volume prediction (HVPRED) from head measurements and used to correct DB by HWHAW. Equations were also derived for HWHAW using direct regression of DB from uncorrected density (with MWHAW in place of MWHBW). Prediction equations were validated in 45 additional subjects (21 M, 24 F). Results were evaluated using equivalence testing, linear regression, Bland−Altman plots, and paired t-tests. Head girth, face girth, and body mass produced the smallest errors for HVPRED. In both M and F validation groups, equivalence (±2% fat by weight) was demonstrated between body fat percent (BF%) by HWHBW and BF% by HWHAW with HVPRED. Variance in computer-averaged samples of MWHAW was significantly less (p < 0.05) than MWHBW. Prediction error was smaller for BF% by HWHAW with HVPRED than for alternative methods. Conclusions: Equivalence between BF% by HWHBW and BF% by HWHAW with HVPRED was demonstrated and differences were not statistically significant. Weight fluctuations were smaller for HWHAW than HWHBW.