Abstract
BackgroundConduct problems in adolescence are a critical public health concern. Putative social risks include exposure to racial discrimination, childhood adversity, association with ‘deviant peers’, and low socio-economic status. In the UK, evidence suggests black Caribbean, and Mixed ethnic groups may be exposed to more of these social risks than white British peers.
Given social risks may cluster by ethnicity, same-ethnicity friendship selection, combined with peer influence of conduct problems, may amplify any differences between ethnic groups in trajectories of conduct problems. To prevent development of conduct problems, and to mitigate ethnic inequality, we need to understand developmental pathways to conduct problems, and to identify and understand any disparities between ethnic groups in diverse settings.
The aims of this thesis are to estimate the prevalence and examine development of conduct problems in diverse inner-city schools, and to explore putative social risks together with peer friendship mechanisms, which may underpin any observed ethnic differences in the development of conduct problems.
The aims of this thesis are to estimate the prevalence and examine development of conduct problems in diverse inner-city schools, and to explore putative social risks together with peer friendship mechanisms, which may underpin any observed ethnic differences in the development of conduct problems.
Method
To meet these aims, I use data from a new accelerated school-based cohort study, Resilience Ethnicity and AdolseCent mental Health (REACH), based in ethnically and economically diverse London boroughs, Southwark, and Lambeth. This sample (~4350) is comprised of three cohorts of secondary-school pupils aged 11-12 (school year 7), 12-13 (school year 8) and 13-14 years (school year 9) which we have followed up over three years. We collected questionnaire data at each time point on conduct problems and putative risk factors for conduct problems.
To estimate the baseline prevalence of conduct problems in diverse inner-city schools, I used intercept-only multilevel logistic regression, then I examined associations with ethnic group, gender, and putative risk factors. To examine conduct problem development, I used latent class analysis to determine developmental trajectories of conduct problems and used multilevel multinomial regressions to examine associations with ethnic group, gender, and putative risk factors. Finally, to explore peer network mechanisms which may underpin aspects of conduct problem development, I used longitudinal social network analysis to test processes of both same-ethnic friend selection and influence of friends on conduct problems.
Results
The prevalence of conduct problems in inner-city school was around three times higher than reported in national studies and there were distinct developmental trajectories of conduct problems (low-persistent, high-persistent, decreasing and increasing) through secondary school. Conduct problem prevalence and trajectories varied by ethnic group, and by exposure to putative risk factors (racism, socioeconomic status, troublesome friends, parent-child relationships, adverse life events) for conduct problems. Broadly, white British pupils were both exposed to fewer putative social risk factors for conduct problems and at decreased risks of developing conduct problems than the overall sample. Among black Caribbean and Mixed ethnic groups, putative social risk factors for conduct problems were more common, and these groups were at increased risks of developing conduct problems through secondary school than the overall sample. There is a tendency for pupils to form friendships with people from the same ethnic group and evidence that conduct problems are socially influenced: students tend to converge towards the average conduct-problem level of their friends.
Conclusions
Together the results of this thesis identify major health inequalities by location and ethnic group in the UK. The findings also elucidate peer selection and influence effects as mechanisms which may amplify existing inequalities in exposure to social risks for conduct problem development. Implications for addressing inequalities in exposure to risk factors for conduct problems, harnessing peer influence effects as a protective intervention, and behaviour management strategies are discussed.
| Date of Award | 1 Jan 2022 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Craig Morgan (Supervisor) & Gemma Knowles (Supervisor) |