Association between school bullying victimization and self-harm in a sample of Chinese children and adolescents: The mediating role of perceived social support

Yusan Che, Jin Lü, Die Fang, Hailiang Ran, Sifan Wang, Xuemeng Liang, Hao Sun, Junwei Peng, Lin Chen, Yuanyuan XiaoFrontiers in Public Health 2022 · DOI

TL;DR

Bullying victimization is a potent driver of self-harm among adolescents, but social support can act as a critical buffer. This study of over 4,600 Chinese students reveals that perceived parental support—more than support from teachers or peers—significantly mediates the relationship between bullying and self-harm, accounting for over 20% of the association.

What Problem Does This Paper Address?

Self-harm among children and adolescents is a global public health crisis, with prevalence rates in China reported as high as 27.6%. While school bullying victimization (BV) is a well-established risk factor for self-harm, the psychological "middlemen" that translate the trauma of bullying into self-injurious behavior are less understood.

Specifically, the authors investigate the role of perceived social support. While Western studies often emphasize peer influence, Eastern collectivist cultures often place a higher premium on family dynamics. This paper seeks to determine which sources of support (parents, teachers, classmates, or friends) actually reduce the risk of self-harm in bullied Chinese youth and whether these effects differ across types of bullying (physical, verbal, or relational).

Methods

The researchers conducted a population-based cross-sectional study in Yunnan province, China, using a two-stage simple random cluster sampling method.

  • Participants: 4,627 students aged 10–17.
  • Measurement Tools:
    • Bullying: The Chinese version of the Olweus Bully/Victim Questionnaire (OBVQ).
  • Self-Harm: The Modified Adolescent Self-harm Scale (MASELF-HARMS), measuring lifetime frequency and severity.
  • Social Support: The Child and Adolescent Perceived Social Support Scale (CASSS), which breaks support down into four sources: parents, teachers, classmates, and friends.
  • Controls: The study adjusted for age, sex, ethnicity, grade, and—crucially—clinical symptoms of depression (PHQ-9) and anxiety (GAD-7).
  • Analysis: The team used multivariate logistic regression to find associations and path modeling (via the Lavaan package in R) to quantify the mediating effects.

Datasets

The study relies on a primary dataset collected between October and November 2020 in Kaiyuan city, Yunnan. The final sample included 4,627 students (49.34% boys) with a high effective response rate of 95.24%. No external or pre-existing benchmarks were used, as this was an original observational survey.

Key Insights and Ideas

  • The "Parental Buffer" is Paramount: Among the four sources of social support studied, only parental support emerged as a significant mediator. While support from friends and teachers is valuable, it did not statistically "explain" the link between bullying and self-harm in the same way.
  • Physical vs. Relational Bullying: The mediating power of parental support varied by the type of bullying. For physical bullying, parental support mediated 100% of the explained variance. For verbal and relational (social exclusion) bullying, it mediated 20.13% and 13.02%, respectively.
  • Cultural Specificity: The authors suggest that in Chinese culture, parents exert a more profound influence on an adolescent's internalizing behaviors than in Western contexts. This highlights the need for culturally tailored mental health interventions.
  • Gender Consistency: Interestingly, the protective effect of parental support was comparable for both boys and girls, suggesting a universal intervention point across genders.

Key Results

  • Prevalence: 12.84% of students reported being victims of bullying, while a staggering 38.62% reported at least one instance of self-harm.
  • Risk Quantification: After controlling for covariates (including depression and anxiety), bullied students were 1.90 times more likely to engage in self-harm (95% CI: 1.57, 2.32).
  • Mediation Statistics: Path analysis showed that parental support accounted for 20.73% of the total association between bullying and self-harm.
  • Model Fit: The path models achieved an "excellent fit" (CFI = 1.00, RMSEA = 0.00), lending high statistical confidence to the mediation findings.

Limitations and Future Work

  • Cross-Sectional Nature: Because the data was collected at a single point in time, the authors cannot definitively prove that bullying causes self-harm or that support causes the reduction; they can only show strong associations. Longitudinal studies are needed to establish causality.
  • Self-Reporting Bias: The data relies on students' honesty regarding sensitive topics like bullying and self-harm, which may lead to under-reporting due to social stigma.
  • Traditional Bullying Only: The study focused on physical, verbal, and relational bullying but did not collect data on cyberbullying, which is an increasingly dominant form of victimization.
  • Geographic Scope: The sample was limited to one province in southwestern China, potentially limiting the generalizability of the results to other regions or urban centers like Beijing or Shanghai.

Summary updated

Abstract

Our study results highlighted the promising interventional benefit of parental support in BV-associated self-harm risk for children and adolescents. For victims of bullying, especially physical bullying, promoting parental support might be effective in reducing self-harm risk. Longitudinal studies are warranted to further corroborate these findings.
BibTeX
@inproceedings{che2022association,
  author = {Yusan Che and Jin Lü and Die Fang and Hailiang Ran and Sifan Wang and Xuemeng Liang and Hao Sun and Junwei Peng and Lin Chen and Yuanyuan Xiao},
  title = {{Association between school bullying victimization and self-harm in a sample of Chinese children and adolescents: The mediating role of perceived social support}},
  booktitle = {Frontiers in Public Health},
  year = {2022},
  doi = {10.3389/fpubh.2022.995546},
}
Figure