William Dietz, MD, PhD
Director, Division of Nutrition and Physical Activity
Centers for Disease Control and Prevention

The Body Mass Index (or BMI) is used to identify overweight and at risk of overweight in children and adolescents. The criteria we use are that a BMI between the 85^th and 95^th percentiles for age and gender identifies children at risk of overweight, and a BMI greater than or equal to the 95^th percentile identifies children or teenagers who are overweight. The rationale for the use of these criteria is that a BMI at the 95^th percentile in the U.S. population corresponds to a BMI of 30 in a young adult. In a young adult, a BMI greater than or equal to 30 is considered obesity.

In addition, a BMI at the 85^th percentile in a young adult identifies a young adult with a BMI of 25, which is considered the top end of normal weight.

Therefore, the criteria used in children and teenagers do correspond (roughly) to the criteria used in adults for the identification of obesity and overweight.

There are several potential limitations to Body Mass Index which require that it not be used as the exclusive standard by which to judge a child at risk of overweight or overweight. The first limitation is that, as you know, BMI is weight divided by height squared.

Weight and height do not directly measure body fatness, so that an additional criteria that should be used for determining whether somebody who is overweight (as indicated by BMI greater than the 95^th percentile) is also overfat, is a measure of a skinfold thickness -- like the tricep skinfold thickness. This will help differentiate children and teenagers who are both overweight and overfat from those who are overweight because of increased muscle mass or bone mass.

One of the questions is how sensitive and specific the BMI is for the identification of children with increased fatness. We compared body fatness in children and teenagers measured by DEXA (which is a definitive measure of body composition) with children with a BMI over the 95^th percentile. The overwhelming majority of those children -- 95 percent of those children with a BMI greater than the 95^th percentile -- also had increases in percent body fat compared to the general population.

The second consideration is how valid Body Mass Index is as a predictor of risk.

We know, from studies of 5 to 10 year-old children whose BMI is greater than the 95^th percentile, that 60 percent have at least one additional cardiovascular disease risk factor -- like elevated systolic or diastolic blood pressure, elevated cholesterol or triglycerides, or elevated insulin levels. In addition about 15 percent of those same children (with a BMI greater than the 95^th percentile) have two or more risk factors.

Therefore, BMI not only identifies children who have increased body fatness, but in addition BMI also indicates that those children who are overweight have associated risk factors.