RESUMEN
As both socioeconomic deprivation and the prevalence of childhood mental health difficulties continue to increase, exploring the relationship between them is important to guide policy. We aimed to replicate the finding of a mental health gap that widened with age between those living in the most and least deprived areas among primary school pupils. We used data from 2075 children aged 4-9 years in the South West of England recruited to the STARS (Supporting Teachers and childRen in Schools) trial, which collected teacher- and parent-reported Strength and Difficulties Questionnaire (SDQ) at baseline, 18-month and 30-month follow-up. We fitted multilevel regression models to explore the relationship between Index of Multiple Deprivation (IMD) quintile and SDQ total difficulties score and an algorithm-generated "probable disorder" variable that combined SDQ data from teachers and parents. Teacher- and parent-reported SDQ total difficulties scores indicated worse mental health in children living in more deprived neighbourhoods, which was attenuated by controlling for special educational needs and disabilities but remained significant by parent report, and there was no interaction year group status (age) at baseline. We did not detect an association between probable disorder and IMD although an interaction with time was evident (p = 0.003). Analysis by study wave revealed associations at baseline (odds ratio 1.94, 95% confidence interval 0.97-3.89) and 18 months (1.96, 1.07-3.59) but not 30 months (0.94, 0.54-1.57). These findings augment the existing, highly compelling evidence demonstrating worse mental health in children exposed to socioeconomic deprivation.