BMI (kg/m2) is a useful clinical/epidemiological index of fatness, based on the ready availability of height and weight. In practice however, an individual with high fat free mass relative to stature, for example, will be misclassified (MC) as overweight. In pediatrics this is compounded by the complexity of a definition of obesity. BMI% for black and white children by age and sex (Must 1991) can be used to divide subjects into fatness categories(Himes 1994): Normal <85%, Overweight Risk 85-95%, Overweight >95%. To estimate the rate of MC, body fat per cent by dual energy x-ray absorptiometry(BF%) was compared to fatness category by BMI% in 213 children, ages 6-18 years (mean 11.3), 111 white, 102 black. We classified those with BF% at or below the mean for a lower fatness category as “MC high”, those with BF% at or above the mean for a higher fatness category as “MC low”. Table

Table 1
Full size table

19/213 (9%) are “MC low”, of these 6 are likely missed overweight; and 10/213 (5%) are “MC high”, of these 6 are probably incorrectly labeled overweight. BMI% identifies groups with significantly different mean BF% which is sufficient for screening and epidemiological purposes. Our study suggests that this stratification of fatness is not sufficient for individual clinical management and research studies where more precise measurement of body fat is often required.

(supported by NIDDK 37352)