FITNESSGRAM® Body Composition Assessment in South African Children

Abstract

BACKGROUND: Childhood obesity prevalence in the US for youth ages 6-19 is ≥17%, nearly a three-fold increase in 30 years. International childhood obesity is increasing in industrialized nations. FITNESSGRAM®, a health-related assessment program, is used in ≥20 countries. FITNESSGRAM® body composition (BC) options include BMI and percent fat (%fat) from skinfolds. It is desirable that students be consistently classified into a "Healthy Fitness Zone" (HFZ) regardless of the BC measure used. PURPOSE: To determine the relation between categorization into HFZ from BMI and %fat measurements in South African children and if the relations are a function of gender and age. METHODS: Triceps and calf skinfolds and height and weight were taken according to FITNESSGRAM® specifications. HFZ achievement was determined based on age- and gender-specific criteria and defined as any BC measurement within the HFZ and below. Contingency tables were developed to determine measurement method agreement across both genders and boys and girls separately. Percent agreement (P), Chi Square, phi coefficient, and kappa were used to determine concordance between BC classification methods. Logistic regression was used to determine the odds for agreement between BMI and %fat classification as a function of gender and age. RESULTS: 1969 (M = 1051; F = 918) South African school-age children (5-13; M = 9.5 ±2.5 yr) completed FITNESSGRAM® testing. Generally, BMI and %fat measures resulted in similar HFZ classifications across and by gender (all P ≥91%; all chi squares p <.001; all phis ≥.60; all kappas ≥.58). Logistic regression indicated no relation between gender and odds of agreement between BMI and %fat classification. Odds of measurement protocol agreement decreased by 19% (95%CI =.76 to.88) for each year. CONCLUSION: Generally, BMI and %fat HFZ agreement exists. Similarity of classification may be dependent on student age. Different classifications can be individually important. Misclassification can lead to mixed messages for parents, students, administrators, and health care providers. Students tested with both BC methods and found to have different classifications should be further tested to determine the exact nature of BC classification.