Physical growth, school environment, and birth information in the physical activity of children: a multilevel analysis

Introduction: There is increasing academic and policy interest in promoting children and young people's health by ensuring that the school environment supports healthy behaviors. Objective: The study aim to describe and interpret the variability of children's PA based on individual characteristics and school context using the multilevel modeling approach. Methodology: A total of 453 (236 boys and 217 girls) children aged 5 to 7.5 years old were assessed. Anthropometry, Birth information, organized regular physical activity (sports), physical activity (PA), Socioeconomic status and School environment context were assessed. A multilevel modeling approach was used to identify hierarchical effects (child and school levels). Results: In model (M0) (β=60.54; p<0.01), only 1.4% of the total variance in children's MVPA level was explained by differences in the school context, while the remaining 98.6% is explained by individual predictors. After the inclusion of child-level predictors (M1), only the nutritional status variable was significantly associated with time in MVPA (β= -5.13; p=0.04). In Model 2, adding the school context information did not change the order of any factor’s importance. However, the nutritional status variable remains significantly associated with time in MVPA (β= -5.08; p=0.04). Conclusions: In conclusion, only nutritional status was associated with variation in MVPA. Although individual characteristics exert greater influence on MVPA than characteristics of the school context, policies to promote physical activity and control overweight and obesity at school should be developed with the aim of encouraging the adoption of more active and healthy lifestyles.