Background/Purpose: Most tracking studies use BMI as the indicator of body fat for children, without knowing BMI-for-age percentile stability. This study was intended to examine changes in middle school children's BMI-for-age percentile in three years.
Method: Height, weight, and corresponding BMI-for-age percentile were obtained from 214 sixth graders (115 girls) at the beginning of their 6th grade (baseline) and the end of 8th grade. Based on the baseline data, participants were classified as overweight (n = 54), at risk of overweight (46), or healthy weight (114). The participants were regrouped based on their 8th-grade BMI-for-age percentiles. Spearman rank correlation, percentage agreements, and Kappa statistics were used to examine the three-year stability of BMI-for-age percentiles.
Analysis/Results: Spearman rank correlation was .88 for the entire cohort, .67 for the overweight group, .55 for the at risk of overweight group, and .60 for the healthy weight group; all at p < .001 level. Percentage of participants remaining in the same group across the three-year period was 81.48% for the overweight group, 60.86% for the at risk of overweight group, and 89.47% for the healthy weight group. The Kappa statistic (.633) indicated a good likelihood (p < .001) that a particular child would stay in the same group across the three years.
Conclusions: The results indicate that children tend to stay in their original BMI-for-age percentile groups in middle schools. It is especially alarming that more than 80% of overweight children remain overweight in three years. More interventions are needed in middle schools for decreasing childhood overweight prevalence.