Multilevel analyses of school and children's characteristics associated with physical activity

Thayse Natacha Gomes, Fernanda K. dos Santos, Weimo Zhu, Joey Eisenmann, José A.R. Maia

Research output: Contribution to journalArticlepeer-review


BACKGROUND: Children spend most of their awake time at school, and it is important to identify individual and school-level correlates of their physical activity (PA) levels. This study aimed to identify the between-school variability in Portuguese children PA and to investigate student and school PA correlates using multilevel modeling. METHODS: The sample included 1075 Portuguese children of both sexes, aged 6-10 years, from 24 schools. Height and weight were measured and body mass index (BMI) was estimated. Physical activity was estimated using the Godin and Shephard questionnaire (total PA score was used); cardiorespiratory fitness was estimated with the 1-mile run/walk test. A structured inventory was used to access information about the school environment. A multilevel analysis (level-1: student-level; level-2: school-level) was used. RESULTS: Student-level variables (age, sex, 1-mile run/walk test) explained 7% of the 64% variance fraction of the individual-level PA; however, school context explained approximately 36% of the total PA variance. Variables included in the model (school size, school setting, playground area, frequency and duration of physical education class, and qualification of physical education teacher) are responsible for 80% of the context variance. CONCLUSIONS: School environment is an important correlate of PA among children, enhancing children's opportunities for being active and healthy.

Original languageEnglish (US)
Pages (from-to)668-676
Number of pages9
JournalJournal of School Health
Issue number10
StatePublished - Oct 2014


  • Correlates
  • Exercise
  • Hierarchical linear modeling
  • Youth

ASJC Scopus subject areas

  • Education
  • Philosophy
  • Public Health, Environmental and Occupational Health


Dive into the research topics of 'Multilevel analyses of school and children's characteristics associated with physical activity'. Together they form a unique fingerprint.

Cite this