The identification of children with adverse risk factor levels by body mass index cutoffs from 2 classification systems: the Bogalusa Heart Study

Background: The cutoffs from the Centers for Disease Control and Prevention (CDC) growth charts and from the Cooper Institute (FitnessGram) are widely used to identify children who have a high body mass index (BMI). Objective: We compared the abilities of these 2 systems to identify children who have adverse lipid concentrations and blood pressure measurements and the reliability (consistency) of each classification system over time (mean follow-up: 7 y). Design: A cross-sectional analysis based on data from 22,896 examinations of 5- to 17-y-olds was conducted. Principal components analyses were used to summarize levels of the 5 risk factors, and likelihood ratios and the κ statistic were used to compare the screening abilities of the 2 systems. Of these children, 3972 were included in longitudinal analyses. Results: There were marked differences in the prevalence of a high FitnessGram BMI by age, with the prevalence among boys increasing from 2.5% to 21% between the ages of 5 and 11 y. The identification of adverse risk factors by the 2 systems was only fair (κ = 0.25), but there was little difference in the abilities of the CDC and FitnessGram cutoffs to identify high-risk children. Longitudinal analyses, however, indicated that the agreement between initial and follow-up FitnessGram classifications was substantially lower than that based on CDC cutoffs (κ = 0.28 compared with 0.49). Conclusions: The FitnessGram and CDC cutoffs have similar abilities to identify high-risk children. However, a high FitnessGram BMI is difficult to interpret because the reliability over time is low, and the prevalence increases markedly with age.