TRANSDIAGNOSTIC GENETIC RESEARCH IN CHILDREN AND ADOLESCENTS

Overall Abstract It is well known that mental health problems including neurodevelopmental issues often start in childhood and adolescence and can continue into adulthood. Previous research has shown that genetic variants influencing adult disorders are also associated with various childhood problems and that genetic correlations between the various mental disorders are high, underlining the need for transdiagnostic research across the lifespan. This symposium will discuss current research further disentangling how genetic variants play a role in childhood symptoms and their persistence over time. The first two presentations will focus on findings in clinical cohorts of children with neurodevelopmental problems. Christel Middeldorp will present findings on a clinical cohort of families with children with neurodevelopmental problems taking a transdiagnostic approach by not focusing on disorders but on symptoms of ADHD, ASD, and depression in children and parents. Exploring how polygenic scores are related to severity in a clinical sample can aid in identifying children at high risk for poor outcome. It is also known that parental mental health can be a risk factor for a less beneficial course in children. Hence, the association between their symptoms and polygenic scores are also explored. The role of biomarkers in childhood mental health is largely unclear. Harry McIntosh will present the results of a study using three cohorts: a clinical study on children with neurodevelopmental problems, one with children diagnosed with Autism Spectrum Disorder and one population based sample. He will explore the genetic basis of urinary metabolite concentrations, and their links with mental health problems. We will continue with two presentations focusing on improving prediction. Swathi Gangaraju will show to what extent polygenic scores add to the prediction of depression and adhd at age 14 using machine learning. She will compare four predictor models: models including only environmental predictors measrued at age 11, models including single disorder PRSs, models including multi-PRSs that include scores from genetically correlated traits selected based on published genetic correlation data, and combined models that include both environmental and multi-PRSs features. Finally, Bochao Lin used genomic structural equation modelling (gSEM) to decompose polygenic risk scores for psychiatric disorders into shared and disorder-specific genetic components. They found that shared genetic liability predicts persistent difficulties across multiple domains. Notably, MDD-specific genetic risk emerged as a broad early marker across symptom trajectories, while schizophrenia-specific risk showed no associations. These findings highlight how genetically informed trajectory analyses can help disentangle heterogeneous developmental pathways and may inform early intervention strategies. Overall, this symposium will show how taking a transdiagnostic approach in research into the genetic risk for developing childhood mental health symptoms provides valuable insights.