38-study meta-analysis links school bullying to PTSD symptoms at r=0.24, with a prediction interval that includes no effect

Serap Keles and colleagues report an updated systematic review and meta-analysis in Educational Research Review on school bullying and post-traumatic stress symptoms and diagnosis among children and youth. A search of seven databases yielded 38 eligible studies under criteria set in advance. The review narrows earlier work that pooled school and workplace bullying. The authors treat as open whether peer victimization counts as a traumatic event that can support a PTSD diagnosis, or whether it mainly tracks broader psychological distress.

What they pooled was association, not incidence. The proportion scoring above PTSD symptom cutoffs ranged from 12.1 percent to 65.8 percent, too wide to quote as one prevalence. Using a correlated-hierarchical effects model with robust variance estimation, the pooled correlation was r = 0.237 (95 percent CI 0.173 to 0.299). A correlation of that size is small to moderate. It is not a relative risk and supplies no base rate for how many additional children receive a diagnosis. Results were heterogeneous. The 95 percent prediction interval, from -0.155 to 0.563, includes zero.

Coverage that stops at the pooled correlation misses a wider uncertainty. The authors report a significant average association and real moderators—mean age, and whether youth were victims, bullies, or both—while also saying other diagnoses cannot be excluded. A score above a symptom cutoff is not a clinician diagnosis. DSM-5-TR Criterion A requires actual or threatened death, serious injury, or sexual violence; much school bullying is harmful without meeting that gate. Nielsen and colleagues’ 2015 meta-analysis, mixing school and work bullying, found a similar association and the same gap between symptoms and established causation.

This design can estimate whether bullying and PTSD-like scores co-occur across published samples. It cannot show that bullying causes PTSD, give a classroom base rate, or separate bullying from earlier adversity, family violence, or pre-existing illness. The news summary does not state funding or conflicts. Absolute risk, not another correlation, is what schools and clinicians need. The next study should be prospective, define the bullying exposure, code Criterion A explicitly, use a clinician interview, and report adjusted risk differences. Until then, the association is real on average and too unstable, and too confounded, to treat bullying as diagnostic of PTSD.