Evaluation of the BOD POD for estimating percent fat in female college athletes

The purpose of this investigation was to examine the accuracy of percent body fat (%BF) estimates obtained by air displacement plethysmography (ADP) using the BOD POD Body Composition System compared with hydrostatic weighing (HW) in a group of female college athletes (n = 80). In addition, %BF estimates by skinfold measures (SF) were also obtained for comparison. A lean subset (n = 39) of the sample was also examined. Mean %BF estimated for the entire sample by ADP (21.2 ± 5.9%) was significantly greater than that determined by HW (19.4 ± 6.4%) and SF (18.8 ± 5.5%). Results from the lean subset also revealed that %BF determined by ADP (17.1 ± 3.7%) was significantly higher than %BF estimates by HW (14.3 ± 2.8%) and SF (15.2 ± 3.2%). The regression equation for the entire sample (%BF HW = 0.937%BF ADP - 0.452, r² = 0.73, standard error of estimates (SEE) = 3.34) did not differ from the line of identity. In contrast, the line of identity differed significantly from the regression equation for the lean subset of female athletes (%BF HW = 0.48%BF ADP + 6.115, r² = 0.41, SEE = 2.18). The results of this investigation indicate that ADP significantly overestimated %BF by 8% in female athletes and by 16% for a leaner subset of the sample compared with HW. It appears that %BF estimates by SF may be more accurate than those obtained by ADP for female college athletes, regardless of body composition. Coaches and trainers evaluating body composition should consider the use of SF before ADP when measuring %BF in female college athletes. Sports scientists should continue to examine the possible gender and body composition bias for ADP.