To gain insights into the relationship between impulsivity and CREM as relevant to addiction vulnerability, we examined three independent, longitudinal ADHD populations, each with age-matched controls. First, in preschool-aged children both with and without diagnosed ADHD27, rs12765063 associated with impulsive behavioral traits, namely distractibility (β=1.14, t127=2.54, p=0.01), engaging in dangerous activities (β=1.20, t127=2.41, p=0.02), and failing to listen to instruction (β=1.13, t127=2.32, p=0.02, Fig. 4a). In the second population composed of adolescents (~18 years-old) with ADHD studied since ~9 years-old28, rs12765063 strongly associated with measures of hyperactivity, namely ankle movement during neuropsychological testing (β=53.41, t127=2.59, p=0.011, Fig. 4b) and self-reported hyperactivity (β=4.26, t138=2.29, p=0.023, Fig. 4c). In this adolescent population, use of alcohol, cigarettes, marijuana and non-prescribed psychostimulants were measured revealing a significant ADHD × rs12765063 interaction (F1,138=5.35, p=0.02, Supplementary Fig. 8a). In more detail, drug use was considerably greater in ADHD subjects with the A-allele compared to controls with the same allele (t138=2.15, p=0.03). Since ADHD severity may change over time, subjects were stratified based on the presence or absence of ADHD symptoms at follow-up which again revealed a
