Identifying Key Factors in University Students’ Physical and Mental Health: An Integrated Regression and Machine Learning Approach

This study sought to identify key factors in university students’ physical and mental health by using a combination of methods, including classical statistical analysis and machine learning techniques. Physical and mental health test data were collected from undergraduates of the 2020 to 2023 cohorts at a university. A self-designed questionnaire on factors associated with physical and mental health was also sent to randomly selected undergraduate students from the same university. The study data were analyzed by one-way analysis of variance, Pearson correlation analysis, hierarchical regression analysis, and a machine learning model. The results revealed that participation in school sports clubs (β = −0.111, p < 0.001) and amount of exercise (β = 0.182, p < 0.001) were significant predictors of physical health status. Difficult family economic situation (β = 0.162, p < 0.001), major satisfaction (β = −0.092, p = 0.02), the quality of a romantic relationship (β = −0.121, p = 0.003), the quality of interpersonal relationships (β = −0.157, p < 0.001), and an overprotective family parenting style (β = 0.109, p = 0.011) were significant predictors of mental health status. The results of regression analysis and application of the machine learning model identified that the amount of exercise, quality of interpersonal relationships, and family parenting style had consistent effects on both the physical and mental health of university students.