A signature of attention-elicited electrocortical activity distinguishes response from non-response to the non-stimulant atomoxetine in children and adolescents with ADHD

Objective: Atomoxetine has several characteristics that make it an attractive alternative to stimulants for treating ADHD, but there are currently no tests identifying individuals for whom the medication should be a first-line option. Method: Within the ADHD Controlled Trial Investigation Of a Non-stimulant (ACTION) study, we examined neuro-cortical activity in 52 youth with ADHD. Baseline event-related potentials (ERP) were compared between those who subsequently responded to 6 weeks of atomoxetine versus those who did not. Results: Responders were distinguished by significantly lower auditory oddball N2 amplitudes than both non-responders and typically developing controls, particularly in the right frontocentral region (p =.002, Cohen's d = 1.1). Leave-one-out cross validation determined that N2 amplitude in this region was able to accurately predict non-responders with a specificity of 80.8%. There were no P3 differences between responders and non-responders. Conclusion: The N2 amplitude is a biomarker that may have utility in predicting response to atomoxetine for youth with ADHD.