Association between ethnic minority status at school and bullying and mental health outcomes in early adolescence: cross-sectional analysis of data from the INCLUSIVE study

Abstract

BackgroundEthnic composition of schools has been the focus of extensive policy interest in the USA and the Netherlands. In the USA, a lower proportion of students of a child's own race or ethnicity at school has been associated with a higher likelihood of bullying and poor mental health outcomes. We investigated whether this association was present in a large study of English schools. MethodsData come from INCLUSIVE (ISRCTN10751359), an ongoing study in 40 secondary schools in South East England (recruited in 2014–15) that aims to assess a programme to reduce bullying among students. We assessed bullying victimisation within the past 3 months (gatehouse bullying Scale [GBS]); prevalence of bullying perpetration through mobile phones or the internet within the past 3 months (cyberbullying perpetration [CP]); and the extent of behavioural, emotional, and peer problems (Strengths and Difficulties Questionnaire [SDQ] score). We stratified students by ethnicity and used multilevel regression models to analyse whether a low proportion of same-ethnicity students within a school was associated with poor outcomes. All models were adjusted for age, sex, family affluence, lone-parent status, and school-level factors (size, single-sex or mixed, type of school, deprivation, rating by the Office for Standards in Education, Children's Services and Skills, and value added). This study was approved by the University College London Ethics Committee and the Institute of Education Ethics Committee. FindingsFor our analysis we included students as categorised by the Office for National Statistics: 2612 White British (WB) students, 1645 Asian/British Asian (ABA) students, and 919 Black/Black British (BBB) students aged 11–12 years. ABA pupils reported significantly fewer behavioural, emotional, and peer problems (raw SDQ score 10·09) than did WB (11·20) and BBB (11·32) students. BBB students were significantly more likely than WB and ABA students to report cyberbullying perpetration (11% [99/888], 8% [200/2577], and 7% [120/1623], respectively) (p<0·05 for all models). There were no significant differences in GBS scores by ethnic group. WB children who attended schools where fewer than 25% of pupils were WB reported higher prevalence of bullying and worse mental health outcomes than those attending schools with more than 25% WB children (GBS odds ratio [OR] 1·77, 95% CI 1·12–2·80; CP OR 3·10, 1·30–7·41; SDQ β=1·60, 0·38–2·82). Equivalent analyses among ABA and BBB groups showed no significant association between same-ethnicity proportion of pupils at their school and any outcome (p>0·05 for all models). InterpretationThese findings suggest that an association between the ethnic composition of schools and adolescent bullying and mental health outcomes might be under-recognised, particularly among white British students. This information could help guide more effective targeting of interventions to address bullying and mental health problems in schools. FundingNone.