School bullying in children and adolescents with neurodevelopmental and psychiatric conditions: a systematic review and meta-analysis.

BACKGROUND: Bullying is a common form of violence among children and adolescents. Young people with neurodevelopmental or psychiatric conditions might have an increased risk of bullying victimisation and perpetration. We aimed to assess the odds of bullying involvement and its association with mental health measures in these populations. METHODS: In this systematic review and meta-analysis, we searched PubMed, ERIC, Psychology and Behavioural Sciences Collection, Web of Science Core Collection, PsycArticles, and PsycInfo databases from inception up to Aug 8, 2023, and included articles reporting data on bullying outcomes of current bullying (within the past year) among children and adolescents (aged 4-17 years) with a diagnosis of a neurodevelopmental or psychiatric condtion provided by a health professional. Bullying type was classified as traditional (physical, verbal, or relational) or as cyberbullying (intentional and repeated harm inflicted through electronic devices and social media), and bullying involvement was classified as victimisation, perpetration, and perpetration-victimisation. Mental health measures were collected and the associations with bullying involvement assessed. We used random-effects meta-analyses to estimate prevalence and odds ratios (ORs) for bullying involvement. Heterogeneity was assessed using the I2 statistic, and publication bias was tested with Egger's regression. This study is registered with PROSPERO, CRD42021235043. FINDINGS: We included 212 studies in the meta-analysis. The total sample comprised 126 717 cases (mean age 12·34 years [SD 1·82], 37·6% girls) and 504 806 controls (12·5 years [SD 1·86], 47·6% girls). For traditional bullying, the pooled prevalence was 42·2% (95% CI 39·6-44·9) for victimisation, 24·4% (22·6-26·3) for perpetration, and 14·0% (11·4-17·1) for perpetration-victimisation. For cyberbullying, the prevalence was 21·8% (16·0-28·9) for victimisation, 19·6% (13·4-27·7) for perpetration, and 20·7% (8·4-42·6) for perpetration-victimisation. Compared with controls, young people with neurodevelopmental or psychiatric conditions were more likely to be involved in traditional and cyberbullying as a victim (OR 2·85 [95% CI 2·62-3·09] and 2·07 [1·63-2·61]), perpetrator (2·42 [2·20-2·66] and 1·91 [1·60-2·28]), and perpetrator-victim (3·66 [2·83-4·74] and 1·85 [1·05-3·28]). Bullying involvement was associated with higher scores in mental health measures in young people with neurodevelopmental or psychiatric conditions, particularly internalising symptoms and externalising symptoms. INTERPRETATION: Our study underscores bullying involvement as a prevalent risk factor in young people with neurodevelopmental or psychiatric conditions that might add to their disease burden through its negative effects on mental health. Interventions targeting these vulnerable populations are warranted to improve their mental health and their future social integration. FUNDING: Instituto de Salud Carlos III, Spanish Ministry of Science and Innovation, and Consorcio Centro de Investigación Biomédica en Red.

Genetic and Structural Brain Correlates of Cognitive Subtypes Across Youth at Family Risk for Schizophrenia and Bipolar Disorder.

OBJECTIVE: Cognitive impairment is an important feature of schizophrenia (SZ) and bipolar disorder (BP) with severity across the two disorders characterized by significant heterogeneity. Youth at family risk for SZ and BP were clustered based on cognitive function and examined in terms of the clinical, genetic, and brain imaging correlates of cluster membership. METHOD: One hundred sixty participants, 32 offspring of patients with SZ, 59 offspring of patients with BP and 69 offspring of healthy control parents underwent clinical and cognitive assessments, genotyping and structural MRI. K-means clustering was used to group family risk participants based on cognitive measures. Clusters were compared in terms of cortical and subcortical brain measures as well as polygenic risk scores. RESULTS: Participants were grouped in 3 clusters with intact, intermediate, and impaired cognitive performance. The intermediate and impaired clusters had lower total brain surface area compared with the intact cluster, with prominent localization in frontal and temporal cortices. No between-cluster differences were identified in cortical thickness and subcortical brain volumes. The impaired cluster also had poorer psychosocial functioning and worse PRS-COG compared with the other 2 clusters and with offspring of healthy control parents, while there was no significant between-cluster difference in terms of PRS-SZ and PRS-BP. PRS-COG predicted psychosocial functioning, yet this effect did not appear to be mediated by an effect of PRS-COG on brain area. CONCLUSION: Stratification based on cognition may help to elucidate the biological underpinnings of cognitive heterogeneity across SZ and BP risk.