Genetic architecture and risk of childhood maltreatment across five psychiatric diagnoses

Importance: Childhood Maltreatment (CM) is associated with psychiatric disorders. Underlying mechanisms are complex and involve genetics. Objective: To investigate the polygenic architecture of CM exposed individuals across psychiatric conditions, and if genetics modulates absolute CM risk in the presence of high-impact risk factors such as parental psychiatric diagnoses. Design and participants: The population-based case-cohort iPSYCH was used to analyse 13 polygenic scores (PGS) in CM-exposed individuals across five psychiatric ICD-10 diagnoses benchmarked against controls. Individuals were stratified into PGS quantiles, and absolute CM risk was calculated using Cox regression. Sex-specific analyses were also performed. Main outcome and measures: PGSs were generated using summary statistics from genome-wide association studies of phenotypes representing psychiatric disorders, CM, educational attainment, and substance use, and tested for their association with CM across psychiatric disorders. 3 Results: This study included 102,856 individuals (47,938 females, 54,918 males) 8-35 years old. We analysed 2,179 CM-exposed individuals across individuals with attention deficit hyperactivity disorder (ADHD; N=22,674), autism (N=18,941), schizophrenia (N=6,103), bipolar disorder (N=3,061), depression (N=28,896) and controls (N=34,689). PGSs for ADHD and educational attainment were associated with CM across all psychiatric diagnoses. The absolute CM risk was increased in the highest PGS groups, e.g., for ADHD the absolute CM risk was 5.6% in the highest ADHD-PGS quartile while only 3,3% in the lowest ADHD-PGS quartile (hazard rate ratio quantile-4 vs quantile-1 = 1.81; CI = 1.47 – 2.22). CM risk was more than twice as high for children with psychiatric diagnosed parents (5.7%) than for children with non-diagnosed parents (2.5%), but even in the presence of this risk factor individuals could still be stratified into risk groups based on their genetics. No genetic differences between CM exposed males and females were observed, but striking sex-differences in absolute CM risk, which reached 5.6% for females in the highest ADHD-PGS quartile while it was 2.0% for males. Conclusions and relevance: Individuals with high a ADHD-PRS and/or low educational attainment-PRS are at elevated risk of CM. Extra attention should be given to CM-high-risk individuals across all five psychiatric diagnoses i.e. females with a high ADHD-PGS and/or a parent diagnosed with a psychiatric disorder.