JMIR Publications

ABSTRACT

Background:

Adolescent violence remains a major public health concern with long-term consequences. While rapid digital and social changes may have altered violence patterns in contemporary society, long-term structural shifts and the dynamic evolution of underlying risk factors remain insufficiently understood.

Objective:

We aimed to identify objective trend changes, statistical breakpoints, and shifting risk structures in South Korean adolescent violence over a 12-year period using nationwide population-based data.

Methods:

We analyzed nationally representative data from 801,050 adolescents (aged 13–18 years) from the Korea Youth Risk Behavior Web-based Survey collected between 2012 and 2024. The primary endpoint was violence victimization requiring medical treatment. Long-term trends were examined using data-driven segmented regression to identify significant temporal breakpoints without prespecifying time points. Period-stratified multivariable logistic regression assessed time-varying associations between demographic, behavioral, and socioeconomic factors and violence victimization. All analyses accounted for the complex survey design and sampling weights.

Results:

A significant structural breakpoint was identified around 2020. Violence prevalence declined steadily until 2019 but increased sharply and remained elevated thereafter. While sadness and smoking remained the strongest predictors (adjusted odds ratios 2.2–2.8), their relative primacy shifted: emotional distress led before 2020, but smoking emerged as the strongest predictor during 2020–2022. Crucially, while overall smoking prevalence declined, its association with violence strengthened, suggesting a 'concentration of risk' within a shrinking but increasingly vulnerable subgroup. Conversely, the impact of academic performance weakened, indicating that traditional pressures were eclipsed by pandemic-related environmental shocks.

Conclusions:

Adolescent violence in South Korea underwent a profound structural shift around 2020. The emergence of smoking as a primary risk indicator—despite declining overall prevalence—signals a fundamental change in the risk landscape, where health-risk behaviors now identify highly marginalized and vulnerable subgroups. Public health and digital health preventions must move beyond traditional academic-focused interventions to integrate mental health surveillance and substance use prevention with systemic efforts to rebuild protective social and institutional environments.