点击显示 收起
Institute for Prevention Research, University of Southern California, Los Angeles, CA 90033
Dear Sir:
In a letter in the March issue of the Journal regarding body mass index and the Quetelet index in children, Weigley (1) asked whether the body mass index expressed as wt2/ht5 rather than as wt/ht2 might be a more accurate index of body fat in children during growth and development. To examine this issue, we analyzed 491 measurements of body fat in children based on dual-energy X-ray absorptiometry conducted in our laboratory over the past 6 y. The mean (±SD) age of the children was 9.7 ± 2.2 (range: 4.214.5 y) and they varied greatly in body weight (42.7 ± 16.8 kg; range: 14.1108.6 kg) and percentage body fat (27.2 ± 11.5%; range: 3.953.9%). The correlation between percentage body fat and wt/ht2 was r = 0.81, and this was not improved by using the alternative index (r = 0.75). As we showed previously (2), the correlations between actual fat mass (kg) and the various indexes were significantly higher, but even so, the alternative index did not improve the fit of the data (r = 0.92 for wt/ht2 and r = 0.83 for the alternative index). In fact, the wt/ht2 index only marginally improved the fit of the data over body weight alone, which had a correlation of r = 0.91 with total fat mass. In a simple multiple regression model, weight and height, when entered separately, explained 90% of the variance in total body fat mass. Thus, as we concluded previously (2), weight, height, and factors such as sex and race may be used to generate simple prediction equations for adiposity in children. These biological models may avoid the use of complex exponential mathematical models, which are difficult to justify from the perspective of sound physiologic models.
REFERENCES