A multi-cohort analysis of interacting prenatal and genomic risks for childhood ADHD symptoms

Importance: Attention-deficit/hyperactivity disorder (ADHD) is highly prevalent in children and has a multifactorial etiology. Both genetic and environmental factors, such as prenatal stress, contribute to disease susceptibility in a sex-specific manner. However, the complex interplay of these factors in ADHD pathophysiology remains poorly understood. Objective: i) assess whether polygenic susceptibility for ADHD modifies the association between prenatal maternal stress and child ADHD symptoms, and how this differs based on ancestry; ii) examine how child sex modifies the link between prenatal stress and child ADHD symptoms. Design, Setting, Participants: Data included four longitudinal birth cohorts from the DREAM BIG consortium (ALSPAC: N=14,009; GUSTO: N=1,176; MAVAN: N=590, GenR: N = 9,778). Polygenic susceptibility for ADHD was computed using the 2023 genome-wide association study results from the Psychiatric Genomics Consortium. Prenatal maternal stress was assessed through a combination of a latent maternal affective symptoms factor and a latent prenatal sociodemographic adversity factor. Repeated assessments of ADHD symptoms between ages 4-7 years were combined into a composite score in each cohort. Cohort-specific analyses used alternating optimization via Latent Environmental and Genetic InTeraction. Main and interaction effects were meta-analyzed across cohorts using inverse variance weighting. Exposure: Prenatal maternal stress was measured using two factors: Prenatal Social-Environmental Adversity (A-Factor) and Prenatal Maternal Affective Psychopathology (M-Factor). Main Outcomes and Measures: Teacher-reported ADHD symptoms in children aged 4-7 years. Results: Meta-regression analysis indicate that genetic susceptibility, prenatal maternal adversity, and child sex are significantly related to ADHD symptoms. However, when the sample is divided by genetic ancestry, these associations are only observed for children with European genetic ancestry. Furthermore, the contribution of our latent environmental factor (prenatal maternal affective symptoms or perinatal socio-environmental adversity) depended on genetic ancestry or child sex. Conclusion: Results suggest that ADHD is influenced by both genetic and environmental factors in complex ways which may differ by type of perinatal adversity, genetic ancestry, and child sex. Our findings suggest that current methods for assessing polygenic risk are potentially insufficient for use in non-European populations.