Physical activity has beneficial effects on overall academic performance in children. However, there is a lack of evidence regarding how the individual characteristics of physical activity interact with other confounding variables of academic competence. Leisure-time physical activity with potential confounders-such as developmental, behavioral, family, and school factors, predicting overall, mathematical, and reading academic competence-was studied in a random sample of 1520 Slovenian primary school students in grades 4-6 (51.9% female; mean age = 10.4 years; SD = 0.93). A structured self-reported questionnaire was used to gather data on the children's leisure-time physical activity and social-demographic variables, while academic competence was measured by teachers using the SSRS Academic Competence Evaluation Scale. The findings showed that children engage in physical activity most days a week, with moderate-intensity and unorganized activities. It was predicted that engaging in physical activity would lead to an increase in academic performance by 4.2% in males (p = 0.002) and 3.2% in females (p = 0.024), but after fully adjusting the model for controlling confounding variables, the prediction increased to 81.1% in females and 84.1% in males (p < 0.001). The frequency and intensity of physical activity, the absence of digital games, and attending sports clubs seem to have the most beneficial effects in terms of academic competence in school children, among other relevant confounders mediating in this complex relationship.
Keywords: academic competence; late childhood; leisure time; multiple regression analysis; physical activity; sports.