1 School of Criminology, University of Montreal, Montreal, Canada.
2 Research Center of the Montreal Mental Health University Institute, Montreal, Canada.
3 School of Psychology, Laval University, Quebec City, Canada.
4 University of Groningen, Groningen, Netherlands.
5 Department of Psychology, University of Montreal, Montreal, Canada.
6 Department of Psychology, Pennsylvania State University, University Park, PA, USA.
7 Sainte-Justine Hospital Research Center, Montreal, Canada.
8 Department of Pediatrics and Psychology, University of Montreal, Montreal, Canada.
9 School of Public Health, Physiotherapy and Population Science, University College Dublin, Dublin, Ireland.
10 Bordeaux Population Health, University of Bordeaux, Bordeaux, France.
11 Department of Social and Preventive Medicine, School of Public Health, University of Montreal, Montreal, Canada.
12 Department of Psychology, Concordia University, Montreal, Canada.
13 Department of Neurology, Klinikum rechts der Isar, School of Medicine, Technical University of Munich, Munich, Germany.
14 School of Criminology, University of Montreal, Montreal, Canada. isabelle.ouellet-morin@umontreal.ca.
15 Research Center of the Montreal Mental Health University Institute, Montreal, Canada. isabelle.ouellet-morin@umontreal.ca.

Individual differences in physical aggression (PA) are largely heritable. However, it remains unclear how genetic propensity for aggression manifests early in life and relates to school-age PA. Preschool externalizing behavior problems (EXT) and adverse experiences are known predictors of school-age PA and also partly heritable. We investigated whether early childhood EXT and adverse experiences mediated the association between polygenic scores and school-age PA. Genotype data collected from 721 participants in the Quebec Longitudinal Study of Child Development (45% males) were used to derive polygenic scores for aggression (PGS_AGG) and attention deficit/hyperactivity disorder (PGS_ADHD). PA was reported annually or biennially on six occasions between ages 6 and 13 by distinct teachers. Mothers reported EXT (PA, hyperactivity, opposition) and adverse experiences (peer difficulties and harsh parenting) on three occasions between ages 3½ and 5. Both PGSs proved significant predictors of school-age PA. However, no early behaviors or adverse experiences mediated the association between PGS_AGG and school-age PA. Contrastingly, hyperactivity and opposition each partially mediated the association between PGS_ADHD and school-age PA. Notably, when both mediators were included in the same model, only hyperactivity remained significant. This study contributes to identifying early manifestations of genetic risk for PA, which may serve as early targets for intervention.