A mechanistic model of ADHD as resulting from dopamine phasic/tonic imbalance during reinforcement learning

• Published in: Frontiers in computational neuroscience Vol. 16; p. 849323
• Main Authors: Véronneau-Veilleux, Florence, Robaey, Philippe, Ursino, Mauro, Nekka, Fahima
• Format: Journal Article
• Language: English
• Published: Lausanne Frontiers Research Foundation 18.07.2022; Frontiers Media S.A
• Subjects: Attention deficit hyperactivity disorder; Basal ganglia; Dopamine; Dopamine receptors; Feedback; Genomes; Impulsivity; Learning; Neostriatum; neurocomputational model; Neurodevelopmental disorders; Neurons; Neuroscience; Reinforcement; reinforcement learning; tonic and phasic dopamine
• ISSN: 1662-5188; 1662-5188
• DOI: 10.3389/fncom.2022.849323

Attention deficit hyperactivity disorder (ADHD) is the most common neurodevelopmental disorder in children. Although the involvement of dopamine in this disorder seems to be established, the nature of dopaminergic dysfunction remains controversial. The purpose of this study was to test whether the key response characteristics of ADHD could be simulated by a mechanistic model that combines a decrease in tonic dopaminergic activity with an increase in phasic responses in cortical-striatal loops during learning reinforcement. To this end, we combined a dynamic model of dopamine with a neurocomputational model of the basal ganglia with multiple action channels. We also included a dynamic model of tonic and phasic dopamine release and control, and a learning procedure driven by tonic and phasic dopamine levels. In the model, the dopamine imbalance is the result of impaired presynaptic regulation of dopamine at the terminal level. Using this model, virtual individuals from a dopamine imbalance group and a control group were trained to associate four stimuli with four actions with fully informative reinforcement feedback. In a second phase, they were tested without feedback. Subjects in the dopamine imbalance group showed poorer performance with more variable reaction times due to the presence of fast and very slow responses, difficulty in choosing between stimuli even when they were of high intensity, and greater sensitivity to noise. Learning history was also significantly more variable in the dopamine imbalance group, explaining 75% of the variability in reaction time using quadratic regression. The response profile of the virtual subjects varied as a function of the learning history variability index to produce increasingly severe impairment, beginning with an increase in response variability alone, then accumulating a decrease in performance and finally a learning deficit. Although ADHD is certainly a heterogeneous disorder, these results suggest that typical features of ADHD can be explained by a phasic/tonic imbalance in dopaminergic activity alone.