Prevalence and XGBoost-Based Prediction of Smartphone Addiction Among Chinese Medical Students

Abstract

Smartphone addiction is becoming an increasingly important behavioral problem among university students. However, evidence among medical students and using sophisticated analytical approaches is still scarce. This study estimated the prevalence of smartphone addiction among Chinese medical students and identified its predictors based on an XGBoost classification model. A cross-sectional anonymous online survey was conducted among medical students from a medical university in Anhui Province, China (N = 557 valid questionnaires). The survey used standardized instruments to measure smartphone addiction, perceived stress and academic burnout, eHealth literacy as well as sociodemographic and study related characteristics. Descriptive statistics and conventional regression analyses were conducted, and an XGBoost classification model was further applied to analyze the questionnaire data and model the prevalence of smartphone addiction. The prevalence of smartphone addiction was 58.7% among males and 58.4% among females. Female gender as well as poorer adaptation to online learning, higher perceived study pressure and psychological stress, and greater academic burnout were significantly associated with smartphone addiction, whereas higher eHealth literacy presented a protective association. In addition, the XGBoost classification model indicated that longer daily non-learning smartphone use duration and higher academic burnout were the two most important risk predictors, followed by perceived stress, whereas higher eHealth literacy served as an important protective factor. Our findings suggest that smartphone addiction is prevalent among Chinese medical students and is closely associated with academic and psychological burdens. Conducting student health monitoring with machine learning–based prediction models, such as XGBoost, may facilitate early identification of high-risk individuals and guide targeted interventions for education and mental health in order to reduce smartphone addiction and enhance healthy technology use.