Dissociating striatal and hippocampal function developmentally with a stimulus-response compatibility task

Abstract

The current study examined the development of cognitive and neural systems involved in overriding a learned action in favor of a new one using a stimulus-response compatibility task and functional magnetic resonance imaging. Eight right-handed adults (mean age, 22-30 years), and eight children (7-11 years) were scanned while they performed a task. Both children and adults were less accurate for incompatible stimulus-response mappings than compatible ones; the children's performance was significantly worse. The comparison of the incompatible and compatible conditions showed large volumes of activity in the ventral prefrontal cortex, ventral caudate nucleus, thalamus, and hippocampus. Striatal activity correlated with the percentage of errors in overriding the old stimulus-response association. The hippocampal activity correlated with the reaction time to make a response to a new stimulus-response mapping that required the reversal of a prior association between a stimulus and a response location. Developmental differences were observed in the volume of striatal/pallidal and hippocampal/parahippocampal activity in that these regions were larger and extended more ventrally in children relative to adults. These results suggest that with maturation and learning, projections to and from these regions may become more refined and focal. Moreover, these findings are consistent with the role of ventral frontostriatal circuitry in overriding habitual and well learned actions and hippocampal systems in learning and reversing associations between a given stimulus and spatial location.

Fig. 1.

Loalization of regions of significant activity during the incompatible mapping condition relative to the compatible mapping condition across all subjects is presented in the top half. The graphs from left toright show (1) percent change in MR signal intensity in the inferior frontal cortex/insula cortex as a function of percent change in accuracy for incompatible relative to compatible trials, (2) percent change in MR signal intensity in the caudate nucleus as a function of accuracy, (3) percent change in MR signal intensity in the thalamus for incompatible trials as a function of the child's age, (4) percent change in MR signal intensity in the hippocampus as a function of percent change in reaction time for incompatible relative to compatible trials. Closed circles are data from children, and open circles are data from adults.

Fig. 2.

Localization of brain regions showing a robust MR signal change for the interaction of group (children, adults) × condition (incompatible, compatible) interaction. R,Right; L, left.