Fig. 1: Association between PRS(C + T) for mental disorder and social behaviour in ALSPAC.

ΔMcFadden’s-R2 is shown for the prediction of low-prosociality and peer-problem scores by ADHD-PRS (a, b), ASD-PRS (c, d), BP-PRS (e, f), MD-PRS (g, h), SCZ-PRS (i, j). Mental disorder genome-wide summary statistics (ADHD-PGC/iPSYCH, ASD-PGC/iPSYCH, BP-PGC, MD-PGC/UKBB, and SCZ-PGC) were used to construct Z-standardised PRS(C + T) in ALSPAC (ADHD-PRS, ASD-PRS, BP-PRS, MD-PRS, and SCZ-PRS) at multiple P-value thresholds. Association analyses with social behaviour (low-prosociality and peer-problem scores) were conducted using negative binomial regression (non-adjusted for cross-disorder PRS effects; multiple-testing corrected P-value: *P ≤ 0.001). ADHD Attention-deficit/hyperactivity disorder, ALSPAC Avon Longitudinal study of Parents and Children, ASD Autism spectrum disorders, BP Bipolar disorder, C + T clumping and thresholding, iPSYCH Lundbeck Foundation Initiative for Integrative Psychiatric Research, MD Major depression, PGC Psychiatric Genomics consortium, PRS Polygenic risk scores, PT PRS P-value threshold, SCZ Schizophrenia. Low-prosociality and peer-problem scores were assessed using the Strengths-and-Difficulties questionnaire.