Anne Fleur Kortekaas-Rijlaarsdam

CHAPTER 5 96 ABS T RAC T Objective This study investigated the effects of medication on reinforcement learning in children with ADHD. We extended previous work by investigating acquisition of stimulus- reward associations, generalization of this knowledge to novel stimulus-pairs, and reversal learning, in the context of a double-blind placebo-controlled cross-over design. Children with ADHD were hypothesized to show impairments in reinforcement learning and MPH was hypothesized to ameliorate these effects. Materials and methods Sixty-three children diagnosed with ADHD (age 8-13, IQ>70) were randomly allocated to a 7 day-methylphenidate or -placebo treatment and were compared to 67 typically developing controls. In the reinforcement learning task children had to learn to select the stimulus with the largest reward rate from a stimulus pair. There were three reward probability conditions (70/30, 85/15, and 100/0%). Data were analyzed using mixed- model analysis. Results Compared to controls, children with ADHD showed intact acquisition of stimulus- reward associations ( p =.727), intact reversal learning ( p =.316), but were impaired in generalization of acquired knowledge ( p =.023). Methylphenidate improved acquisition of stimulus-reward associations ( p =.015) and there was a trend effect for generalization of knowledge ( p =.086). Conclusion Children with ADHD have intact instrumental learning but face difficulties in applying this knowledge to novel contexts. Medication, when considered in relation to reinforcement learning in ADHD, possibly has some potential benefit as a way of improving generalization of stimulus-reward associations.