Performance of BMI-for-Age As A Screening Tool

"The validity of selected cutoff points to identify adolescents with the highest percentage of body fat has been investigated. In general, common cutoff points for BMI and relative weight have low sensitivities but high specificities. For example, BMIs > 85^th percentile has a sensitivity of 29% and 23% for identifying adolescent males and females, respectively, who are above the 90^th percentile for percentage body fat; corresponding specificities are 99% and 100% (Himes and Bouchard, 1989). In screening for adolescent overweight, specificity may be more important than sensitivity. Maximizing specificity minimizes the proportion of adolescents who will be incorrectly considered overweight by the screen" (Himes and Dietz, 1994).

Recently it has been shown that cardiovascular risk factors are associated with the established BMI-for-age cutoffs. Freedman et al., (1999) found that approximately 60% of 5 to 10 year-old children with BMI-for-age values > the 95^th percentile had at least one biochemical or clinical risk factor for cardiovascular disease such as hypertension, elevated insulin levels, and hyperlipidemia. Twenty percent of children had two or more risk factors.