The Neural Basis of Habit Formation Measured in Goal-Directed Response Switching

To override ongoing habitual responses requires switching well-learned actions with new goal-directed processing. However, the neural circuits responsible for these processes remain unclear. This study infers habit strength by introducing a novel task capturing the increased cost associated with switching a habitual response. We employed neuroimaging and brain stimulation to examine the dynamic interactions between human brain regions involved in habits and their interference with ongoing incompatible goal-directed behavior. Training S-R links in overtrained stimuli (compared to less trained ones, termed standard-trained stimuli) increased RT switch costs, explained by drift diffusion computations governing both the training and outcome devaluation phases. Training engaged sensorimotor areas and the posterior putamen, whereas standard trained behaviors, recruited the posterior caudate, insula, and prefrontal regions. A cortical network orchestrated habit expression (right S1 with the left anterior insula/prefrontal areas) while also implicating basal ganglia when overriding habits (left premotor with the putamen). Importantly, stimulation of the left premotor played a causal role in habit control, enhancing performance across both the training and devaluation phases. Our findings reveal an interaction between habitual and goal-directed brain regions, highlighting shared neural dynamics when overriding habitual behaviors.