Predicting Bullying Victimization In Chinese Adolescents: A Multivariate Analysis With Socio-ecological Factors And Nomogram Model

PURPOSE: Bullying victimization during adolescence can have long-term psychological effects and is a significant public health concern. This study aims to explore the factors contributing to bullying victimization among Chinese adolescents and develop a predictive model to identify students at high risk. METHODS: Data from the 2021-2023 Youth Health Survey included a sample of 18,397 adolescents aged 12-18 years (mean age: 15.23 ± 1.236 years), consisting of 9,586 males (52.1%). Univariate analysis and RF modeling were employed to identify key predictors within the socio-ecological model. To assess the impact of these variables, a multivariable logistic regression model was constructed, with Lasso regression applied to address multicollinearity issues. A nomogram was then developed to estimate the probability of bullying victimization. RESULTS: The result yielded that eight variables of socio-ecological model were identified as independent predictors of bullying: grade (OR = 1.57, [1.23-2.00], p < .01), parental conflict (OR = 1.62, [1.38-1.89], p < .01), physical activity (OR = 1.78, [1.49-2.13], p < .01), smoking (OR = 3.54, [3.04 - 4.12], p < .01), drinking (OR = 3.92, [3.41 -4.50], p < . 01), sedentary behavior (OR = 1.52, [1.32-1.76], p < .01), social competence (OR = 0.98, [0.98-0.99], p < .01), and antisocial behavior (OR = 1.00, [1.00-1.01], p < .01). The regression model indicated that older students and those frequently experiencing parental conflict had a higher likelihood of being bullying victims. Additionally, increased levels of physical activity were associated with a higher risk of bullying, while sedentary behavior was also linked to greater victimization risk. Using the results from the multivariable logistic regression, a nomogram was developed to estimate the likelihood of bullying victimization. The model demonstrated strong predictive performance, with an AUC of 0.83 in the training set and AUC values of 0.82 in external validation sets. CONCLUSION: The findings highlight the critical role of socio-ecological factors in bullying victimization. This study offers a solid empirical foundation for creating targeted intervention strategies and presents an effective risk assessment tool for educators and parents to identify potential bullying victims.