Can response to ADHD medication be predicted?

Predictors for the pharmacological effect of ADHD medication in children and adolescents are lacking. This study examined clinically relevant factors in a large (N = 638) prospective cohort reflecting real-world evidence. Children and adolescents aged 6-17 diagnosed with ADHD were evaluated at baseline and three months following ADHD medication initiation. The outcome was measured as a reduction in total SNAP-IV (Swanson Nolan and Pelham teacher and parent rating scale) score at three months compared to baseline. Outcome groups were defined as Responders (≥ 40% reduction), Intermediate Responders (≥ 20 to < 40% reduction), and Non-responders (< 20% reduction). Included independent variables were the Autism Spectrum Questionnaire (ASSQ), the Spence Children's Anxiety Scale (SCAS), the Pediatric Side Effect Checklist (P-SEC), anthropometrics measures, geographical region, relative age, Children´s Global Assessment Scale (CGAS), Intelligence quotient (IQ), pharmacological treatment initiation month, ADHD symptom severity, ADHD presentation, and psychotic-like experiences. Multinomial logistic regression suggested that ADHD symptom severity, region, relative age, and stating ADHD medication at three-month follow-up were associated with the response outcome group. However, when validating the data with Bootstrap Forest, none of the variables were significant. Thus, in our large naturalistic cohort, we could not identify any clinically relevant factors that reliably predict pharmacological treatment outcomes.Trial registration: NCT02136147.