The study objectives were (1) to determine the variation in prevalence of overweight between school communities, (2) to evaluate the relationship between cardiorespiratory fitness and the probability of being overweight among different school communities, and (3) to test whether this relationship varies between school communities. Using a repeated cross-sectional design, data from 31,424 (15,298 girls, 16,126 boys) Australian school children who had objective assessments of body composition and physical performance were used. Ninety-one schools located across 5 states and territories were included. Independent samples were taken across 12 school years (2000-2011). Analysis used generalised linear mixed models in R with a two-level hierarchical structure-children, nested within school communities. Predictor variables considered were: level 1-gender, age, cardiorespiratory fitness and year of measurement; level 2-school community. A total of 24.6% of the children were overweight and 69% were of low fitness. Variation in the prevalence of overweight between school communities was significant, ranging from 19% to 34%. The probability of being overweight was negatively associated with increasing cardiorespiratory fitness. The relationship was steepest at low fitness and varied markedly between school communities. Children of low fitness had probabilities of being overweight ranging between 26% and 75% depending on school community, whereas those of high fitness had probabilities of <2%. Our findings suggest that most might be gained from a public health perspective by focusing intervention on the least fit children in the worst-performing communities.
Keywords: Childhood obesity; Multi-level analysis; Physical activity; Prevention; Social environment; Sports participation.
Copyright © 2015 The Authors. Published by Elsevier Inc. All rights reserved.