years) and adolescent (11–21 years) obesity does not exist. Generally, obesity is defined as weight-for-height above the 90th percentile on the growth charts from the National Center for Health Statistics (NCHS) or weight in excess of 120 percent of the median weight for a given height (Leung and Robson, 1990). The 85th percentile of triceps skinfolds is also used (Garn and Clark, 1976; Gortmaker et al., 1987), and skinfolds correlate well with percentage of body fat in children of both sexes (Forbes and Amirhakimi, 1970; Roche et al., 1981). It may be difficult to measure skinfolds, as evidenced by the mere 8 percent of pediatricians and family practitioners who take these measures (Greecher, 1993).

The Quetelet index or body mass index (BMI) is an indirect measure of body fat, but it is easily and reliably measured, correlating well with more precise estimates of subcutaneous and total body fat (Deurenberg et al., 1991; Roche et al., 1981). In 1985, the National Institutes of Health (NIH) Consensus Conference on Obesity recommended that BMI plus relative weight be used in clinical and public health settings (Burton et al., 1985). Although arbitrary, the 85th and 95th percentiles of BMI have been used to define obesity and superobesity, respectively (Garn and Clark, 1975; Harlan, 1993; Must et al., 1991), and these definitions are used in Table C-1. However, use of the 85th percentile may incorrectly label a large group of children obese, and a BMI in the 90th or 95th percentile may be more appropriate (Dietz and Robinson, 1993; Williams et al., 1993).

The Committee on Clinical Guidelines for Overweight in Adolescent Preventive Services recently proposed specific criteria for defining overweight and obesity in adolescents (Himes and Dietz, 1994), recommending routine use of BMI in screening. The committee also recommended