The genetic risk of mental health disorders in children from diverse population-based cohorts is modulated by poverty

Attention-deficit/hyperactivity disorder (ADHD) and depression share substantial genetic liability, yet the extent to which socioeconomic disadvantage shapes the expression of this genetic risk remains unclear. We investigated whether poverty moderates and mediates associations between polygenic risk scores (PGS) for ADHD and major depressive disorder (MDD) and corresponding symptoms in early adolescence. Data were drawn from two population-based cohorts: the 2004 Pelotas Birth Cohort (Brazil; N = 3,470) and the Adolescent Brain Cognitive Development (ABCD) Study (United States; N = 10,218). ADHD and depressive symptoms were assessed using caregiver reports (Strengths and Difficulties Questionnaire in Pelotas; Child Behavior Checklist in ABCD). Household income was harmonized and categorized into low (bottom 30%), middle (40%), and high (top 30%) income. PGS were derived using SBayesRC and PLINK2. Linear regression models tested main genetic effects and gene–environment interactions, and mediation analyses quantified indirect effects via poverty, adjusting for age, sex, and ancestry principal components. In Pelotas, ADHD-PGS was associated with ADHD symptoms (β = 0.86, SE = 0.23, p < 0.0001) and emotional symptoms (β = 0.69, SE = 0.24, p = 0.003), with a significant interaction with poverty for ADHD symptoms (pinteraction = 0.024). In ABCD, both ADHD- and MDD-PGS were associated with ADHD and depressive symptoms (all p < 0.001; ΔR² = 0.21–0.48), and gene–environment interactions were observed for all associations (pinteraction < 0.01), indicating attenuated PGS effects under greater socioeconomic disadvantage. Mediation analyses showed that poverty accounted for 7.9–10.2% of the total PGS effect on ADHD symptoms in Pelotas and 4.7–5.3% in ABCD; indirect effects for depressive symptoms were significant only in ABCD (5.3–10.1%). Socioeconomic disadvantage both modifies and partially mediates genetic liability to adolescent psychopathology, suggesting that polygenic risk reflects context-dependent vulnerability shaped by structural conditions.