5.11 Polygenic Risk for ADHD Indexes Behavior Ratings of Executive Functions in Child Psychiatry Outpatients

The first significant genome-wide association study (GWAS) of ADHD has made it possible to quantify common genetic variation underlying the condition in individuals. Determining the clinical utility of such polygenic risk scores (PRS) awaits a better understanding of the phenotypes that these scores index in clinical samples. Parent rating scales of executive function (EF) associate with functional difficulties and capture behaviors beyond the DSM ADHD criteria. Thus, the extent to which EF rating scales relate to ADHD PRS has implications for clinical translation. Subjects were 365 youth (11.4 ± 3.1 years; 35.3% girls) referred for neuropsychiatric evaluation, enrolled in the Longitudinal Study of Genetic Influences on Cognition and genotyped with the Illumina PsychChip. ADHD PRS was derived from the largest published ADHD GWAS. Parents rated EF using the Behavior Rating Inventory of Executive Function (BRIEF), which reflects behavioral and emotional regulation (BER) and metacognition (MC). Linear regression models related ADHD PRS to these constructs alone, after accounting for polygenic variation (PV) for cognitive ability (COG), and after controlling for ADHD diagnoses. Finally, using structural equation modeling (SEM), we examined the extent to which ADHD PRS and BRIEF associations were mediated by ADHD diagnoses. Analyses controlled for age, sex, medication, and the first 5 principal ancestry components. Benjamini-Hochberg False Discovery Rate (FDR) 0.05 was used to correct for multiple testing. ADHD PRS was significantly associated with BER and MC. Controlling for COG PV resulted in a smaller but significant association of ADHD PRS with BER (ΔR2 = 2.40%; t = 3.22; p = 0.0014), but the association with MC was not significant. Controlling for ADHD only yielded a significant association with BER (ΔR2 = 2.41%; t = 3.27; p = 0.0012). SEM indicated a partial mediation effect for the model for BER, with a significant indirect effect of ADHD PRS via ADHD (b = 0.02; z = 2.13; p = 0.0335) and a direct effect (b = 0.16; z = 3.32; p = 0.0009). In contrast, the association between ADHD PRS and MC was fully mediated by ADHD status. ADHD PRS influence behavioral and emotional regulation indirectly via ADHD as well as directly, suggesting that the risk score captures variation relevant to functional outcome beyond the traditional ADHD diagnosis.