Abstract
Conduct problems are more prevalent in neighbourhoods that have a vulnerable structure (e.g., high neighbourhood-level deprivation) and disarranged interpersonal characteristics (e.g., low social cohesion and informal social control at the neighbourhood level). However, as an indicator of neighbourhood structure, neighbourhood deprivation has typically not been assessed longitudinally and solely based on neighbourhood-level socioeconomic status rather than a wide variety of census-level deprivation indicators. Also, few studies have examined which CD behaviours (e.g., stealing) interplay with neighbourhood risks (e.g., low social cohesion). In this study, latent transitions of neighbourhood-level deprivation patterns, based on census-level information, were estimated between age 12.5 and 15.5 using the Avon Longitudinal Study of Parents and Children (ALSPAC). In network models, we employed multi-informant variables and estimated interplays between mother-reported CD behaviours and child-reported social cohesion, informal social control and deviant peer affiliations within different patterns of the latent neighbourhood-level deprivation transitions. We identified three constant deprivation patterns: deprived, intermediate and low pattern. In the deprived pattern, the CD behaviour "bullying" had the highest interplay with lack of social cohesions, social control, and high deviant peer affiliation. In contrast, non-violent CD behaviours: "lying" and "staying after dark ", showed importance in the intermediate and low patterns, respectively. Regardless of deprivation patterns, social cohesion played a protective role, whereas affiliation with deviant peers involved in property crime was a risk factor for CD behaviours. The identified CD behaviours can serve as a screening tool, and interventions increasing social cohesion might mitigate CD development.
