School-based physical activity interventions and physical activity enjoyment: A meta-analysis

Abstract

The purpose of this meta-analysis was to examine the effectiveness of school-based physical activity interventions on increasing students' physical activity enjoyment. An internet search with several databases using the keywords "Adolescents", "Children", "Enjoyment", "Physical Activity", and "Schools" was performed yielding over 200 published studies. Studies were eliminated based on the lack of experimental manipulation (i.e., non-intervention studies), no assessment of physical activity enjoyment as an outcome variable, a lack of a control or comparison group, and no reporting of the effect estimate's variability (i.e., standard deviation, standard error, etc.). This procedure resulted in 10 studies being examined in the meta-analysis. Data were analyzed in the state of Utah, USA in 2017. The Hartung-Knapp-Sidak-Jonkman method for a random effects meta-analysis was employed with studies being weighted by inverse variance. The pooled Standardized Mean Difference=0.38 (95% C.I. [0.10-0.65], p=0.011). Cochran's Q test showed statistical significance (p<0.001) and the I²=76.6%, suggesting large heterogeneity across the 10 studies. Egger's regression model yielded an intercept coefficient that was statistically significant (bias=3.28, 95% C.I. [0.21-6.36], p=0.039), indicating the presence of small-study effects. This meta-analysis provides evidence that school-based physical activity interventions can be effective in increasing physical activity enjoyment in children and adolescents. However, the magnitude of the pooled effect was small-to-moderate and there was evidence for publication bias and large between-study heterogeneity.

Keywords: Adolescent; Child; Exercise; Meta-analysis; Schools.