Explaining ethnic variations in adolescent mental health: a secondary analysis of the Millennium Cohort Study

PURPOSE: The relationship between ethnicity and adolescent mental health was investigated using cross-sectional data from the nationally representative UK Millennium Cohort Study. METHODS: Parental Strengths and Difficulties Questionnaire reports identified mental health problems in 10,357 young people aged 14 (n = 2042 from ethnic minority backgrounds: Mixed n = 492, Indian n = 275, Pakistani n = 496, Bangladeshi n = 221, Black Caribbean n = 102, Black African n = 187, Other Ethnic Group n = 269). Univariable logistic regression models investigated associations between each factor and outcome; a bivariable model investigated whether household income explained differences by ethnicity, and a multivariable model additionally adjusted for factors of social support (self-assessed support, parental relationship), participation (socialising, organised activities, religious attendance), and adversity (bullying, victimisation, substance use). Results were stratified by sex as evidence of a sex/ethnicity interaction was found (P = 0.0002). RESULTS: There were lower unadjusted odds for mental health problems in boys from Black African (OR 0.15, 95% CI 0.04–0.61) and Indian backgrounds (OR 0.42, 95% CI 0.21–0.86) compared to White peers. After adjustment for income, odds were lower in boys from Black African (OR 0.10, 95% CI 0.02–0.38), Indian (OR 0.40, 95% CI 0.21–0.77), and Pakistani (OR 0.49, 95% CI 0.27–0.89) backgrounds, and girls from Bangladeshi (OR 0.18, 95% CI 0.05–0.65) and Pakistani (OR 0.63, 95% CI 0.41–0.99) backgrounds. After further adjustment for social support, participation, and adversity factors, only boys from a Black African background had lower odds (OR 0.16, 95% CI 0.03–0.71) of mental health problems. CONCLUSIONS: Household income confounded lower prevalence of mental health problems in some young people from Pakistani and Bangladeshi backgrounds; findings suggest ethnic differences are partly but not fully accounted for by income, social support, participation, and adversity. Addressing income inequalities and socially focused interventions may protect against mental health problems irrespective of ethnicity. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s00127-021-02167-w.