Prevalence of Psychological Distress among the University Students: Logistic Regression Analysis with Machine Learning Algorithm

The significance of mental health is paramount for the future achievements of students. The prolonged closure of educational institutions as a result of COVID-19 has notably impacted the mental health status of students. This study aims to explore the prevalence of psychological distress among university students and classify them using a machine learning algorithm. A cross-sectional study incorporating a convenience sample (n=435) was carried out among university students. The DASS-21 scale, along with a self-administered questionnaire, is utilized to quantify the intensity of psychological distress. The collected data was analyzed by univariate, bivariate, multivariate analysis, and machine learning (ML) approach. The prevalence of moderate to extremely severe depression, anxiety, and stress was 43.2%, 72.7%, and 32.7%, respectively. Logistics regression shows rurality, lower academic performance, fear of academic delay, not performing physical activities, having suicidal thoughts, and being in a relationship contribute to the risk of developing a higher level of psychological distress. The performance of correctly predicting the models for depression, anxiety, and stress was 75.36%, 84.18%, and 71.33%, respectively. The COVID-19 pandemic has the potential to affect students' mental well-being significantly. The findings suggest the importance of taking steps to recognize students who may be experiencing psychological distress and ensuring they receive appropriate support. The results of this study may provide valuable insights for future research and the development of psychological distress intervention programs.