EXPLORING THE LONGITUDINAL RELATIONSHIPS BETWEEN REGULAR PHYSICAL ACTIVITY, SPORTS, AND WELL-BEING IN YOUTH USING RANDOM FORESTS

Author(s): WANG, L., LIN, W., Institution: COLLEGE OF LIFE AND ENVIRONMENTAL SCIENCES, UNIVERSITY OF EXETER, Country: UNITED KINGDOM, Abstract-ID: 212

ABSTRACT: Well-being is significantly impacting youth developmental trajectory and overall life outcomes which is crucial. Sports, exercise, and physical activity are universally recognized for their beneficial effects on well-being. However, disentangling their individual contributions poses a persistent challenge for researchers. The objective of this study is to employ innovative methods to assess sports and explore the relationship between sports participation, regular physical activity (RPA), and well-being in both cross-sectional and longitudinal studies.
Method: this study utilizes the random forest approach to investigate the intricately interconnected variables of sports and RPA, alongside factors associated with well-being. The assessment of sports, using objective assessment of table tennis skill learning and performance, and subjective assessment of table tennis engagement. The evaluation of RPA, employing a self-reported questionnaire, emphasizing the frequency and duration of moderate to vigorous physical activities (MVPA).
Results: this study encompassed 163 participants, predominantly males (n=133) compared to females (n=30), with an average age of 18.99 years (SD=0.88), The random forest analysis yielded two models with satisfactory fits. The cross-sectional model, which includes baseline sports and regular physical activity, showed a root mean squared error (RMSE) of 14.66 and an R-squared (R²) value of 0.12. The longitudinal model, examining well-being changes over 8-month period, demonstrated an RMSE of 14.76 and an R² of 0.06. Post-adjustment for gender, age, and BMI, the decision tree from Model 1 identified sports and RPA as the most significant factors, accounting for 16.19% of the variance in well-being. Individually, sports accounted for 16.13% of the well-being variance, followed by RPA, which contributed 7.97%. In Model 3, the decision tree similarly highlighted sports and RPA as pivotal, responsible for 16.77% of the well-being variance, with sports alone contributing 13.44% and RPA 7.73%.
Conclusion: this study elucidates that sports have a more substantial correlation with well-being compared to RPA. Moreover, when combining sports with RPA, the correlation with well-being is higher than considering these two factors separately. Future research needs to clarify which aspects of sports participation are more closely associated with well-being. Also, its important to identify which patterns of RPA that most robustly associate with improved well-being.
Discussion: This research is the first to discuss the relationship between sports, RPA, and the well-being of youth using the random forest method. It is also the first study to assess the effects of sports through the evaluation of motor skills, engagement, and performance in sports activities. It suggests that participating in organized sports to develop motor skill, engagement, and performance could be more effective for the well-being of youth than engaging RPA.