A multi-level analysis of individual- and school-level correlates of physical fitness in children

Abstract

Background: Children’s physical fitness is potentially influenced by biological, behavioural and environmental factors. Aim: To investigate the importance of individual-level and school-level characteristics in explaining variation in children’s physical fitness. Subjects and methods: The sample comprised 354 Portuguese children, aged 5–10 years. Physical fitness [(handgrip (HG) strength, standing long jump (SLJ), shuttle-run (SR), total physical fitness score (PFz)], gross motor coordination (GMC) and body mass index (BMI) were assessed. Moderate-to-vigorous physical activity (MVPA) was objectively monitored; birth weight and socio-economic status (SES) were obtained by questionnaire. School contexts were assessed via an objective audit. Multilevel models using the Bayesian approach were used. Results: Age was positively associated with HG, SLJ and PFz (p < 0.05); BMI was positively associated with HG (p < 0.05) and GMC with all physical fitness components (p < 0.05); MVPA was only associated with PFz (p < 0.05); children with level A SES outperformed those with level C SES in HG (p < 0.05), but those with level B SES outperformed their level A SES peers in SR (p < 0.05). Within the school context, no equipment for physical education was negatively associated with HG (p < 0.05). Conclusion: In the primary school years, individual-level characteristics were more important than the school context in explaining variations in physical fitness.

Keywords

• children
• physical fitness
• school context
• multi-level modelling