What explains adolescents' physical activity and sports participation during the COVID-19 pandemic? - an interpretable machine learning approach.

Journal: Journal of sports sciences
PMID:

Abstract

Adolescents' physical activity (PA) and sports participation declined due to the COVID-19 pandemic. This study aimed to determine the critical socio-ecological factors for PA and sports participation using a machine learning approach. We did a cross-sectional secondary data analysis utilising the 2021 National Survey of Children's Health (NSCH) dataset (=16,166; 49.0% female). We applied an interpretable machine learning approach (e.g. decision tree-based models) that examined the critical factors associated with PA and sports participation. The factors related to the intrapersonal, interpersonal, organisational, and community levels of the socio-ecological model. Out of the 25 factors examined, our findings unveiled the 11 critical factors associated with PA and the 10 critical factors associated with sports participation. Factors at the intrapersonal levels (e.g. age, screen time, and race) held greater importance to PA than those at the other three levels. While interpersonal factors (e.g. parent participation in children's events/activities, family's highest educational level, and family income level) were most important for sports participation. This study identified that the common critical factors of physical activity and sports participation during the COVID-19 pandemic mainly relied on intrapersonal and interpersonal levels. Unique factors were discussed.

Authors

  • Lingyi Fu
    Department of Health & Kinesiology, University of Utah, Salt Lake City, USA.
  • Ryan D Burns
    Department of Health & Kinesiology, University of Utah, Salt Lake City, USA.
  • Shandian Zhe
    Department of Computer Science, University of Utah, Salt Lake City, 84112 Utah, USA.
  • Yang Bai
    Key Laboratory of Digital Medical Engineering of Hebei Province, College of Electronic and Information Engineering, Hebei University, Baoding 071000, Hebei, China.