Percentage fat in overweight and obese children: comparison of DXA and air displacement plethysmography.

OBJECTIVE: To compare percentage body fat (percentage fat) estimates from DXA and air displacement plethysmography (ADP) in overweight and obese children. RESEARCH METHODS AND PROCEDURES: Sixty-nine children (49 boys and 20 girls) 14.0+/-1.65 years of age, with a BMI of 31.3+/-5.6 kg/m2 and a percentage fat (DXA) of 42.5+/-8.4%, participated in the study. ADP body fat content was estimated from body density (Db) using equations devised by Siri (ADP(Siri)) and Lohman (ADP(Loh)).
RESULTS: ADP estimates of percentage fat were highly correlated with those of DXA in both male and female subjects (r=0.90 to 0.93, all p<0.001; standard error of estimate=2.50% to 3.39%). Compared with DXA estimates, ADP(Siri) and ADP(Loh) produced significantly (p<0.01) lower estimates of mean body fat content in boys (-2.85% and -4.64%, respectively) and girls (-2.95% and -5.15%, respectively). Agreement between ADP and DXA methods was further examined using the total error and methods of Bland and Altman. Total error ranged from 4.46% to 6.38% in both male and female subjects. The 95% limits of agreement were relatively similar for all percentage fat estimates, ranging from +/-6.73% to +/-7.94%. DISCUSSION: In this study, conversion of Db using the Siri equation led to mean percentage fat estimates that agreed better with those determined by DXA compared with the Lohman equations. However, relatively high limits of agreement using either equation resulted in percentage fat estimates that were not interchangeable with percentage fat determined by DXA.