“Instead of trying to produce a programme to simulate the adult mind, why not rather try to produce one which simulates the child’s?” – Alan Turing (Computing Machinery, p456)
One of the defining features of childhood cognition is “behaving without thinking.” Not surprisingly, developmental cognitive psychology has latched onto the idea of impulse control – and other processes putatively requiring inhibition – as a central explanatory construct, playing a role in attention deficit disorder, post-traumatic stress disorder and everything in between (including developmental trends in normal children).
To determine how these processes are implemented in the brain, a recent article by Bunge et al investigated the neural correlates of two types of inhibition across developmental time: the ability to inhibit a prepotent manual response (response inhibition) and the ability to ignore irrelevant stimuli in the environment. To do so, 16 adults and 16 8-12 year old children completed the following task while being scanned with functional magnetic resonance neuroimaging.
Casual readers may wish to skip to the conclusions from this study, which follow the metholodological details in italics below:
First, subjects viewed an array of five objects and were to press a “right” button if the central object (always an arrow) pointed to the right, or a “left” button if the central arrow pointed to the left. The surrounding objects were usually irrelevant for the task – but could nonetheless be arrows pointing in the same direction as the central arrow (“congruent” trials), arrows pointing in the opposite direction of the central arrow (“incongruent” trials), or diamonds (“neutral” trials). However, on a minority of trials, the surrounding arrows were X’s – on these trials subjects should withhold all responses (“no-go” trials).
The behavioral results showed, as expected, that children made more errors and were slower than adults in general, and in particular were less accurate and slower on the incongruent and no-go trials. Somewhat surprisingly, there were no age-related differences in performance among children.
In adults, incongruent trials were associated with greater insula, inferior parietal, putamen, and right ventrolateral prefrontal (vlPFC) activity relative to neutral trials. Right vlPFC is often thought to be involved in response inhibition or response override processing. In contrast, children more strongly activated insula, inferior parietal and left vlPFC regions on incongruent relative to congruent trials; interestingly, left vlPFC activity among children was similar in magnitude and variability to right vlPFC activity in adults.
The efficacy with which adults ignored the irrelevant but incongruent surrounding arrows (as indexed by behavioral slowing on incongruent relative to neutral trials) was related to the extent of activity in right prefrontal regions, including anterior insula, dorsolateral (dl) and vl PFC. Conversely, children more strongly activated left lateralized regions (including left vlPFC, insula, caudate nucleus, and pulvinar) when they were more efficient on incongruent relative to neutral trials. Children, but not adults, with more efficient incongruent trials also tended to have higher scores on a test of verbal reasoning ability.
On “no-go” trials, adults more strongly activated bilateral vlPFC, dlPFC, precuneus, right cerebellum & temporal as well as right superior & inferior parietal regions relative to neutral trials, whereas children only seemed to activate right inferior frontal gyrus and medial frontal gyrus to “no-go” trials relative to neutral trials. An even larger slew of regions was correlated with no-go performance, including right middle frontal gyrus in children and right vlPFC in in adults. Some regions showed activation during both types of “inhibition” trials – in adults, this included right inferior frontal gyrus (rIFG) whereas it included left iFG in children.
There are many complications to running a study like this: as Bunge et al note there can be multiple reasons for different neural activations across age groups, only some of which are interesting. Developmental neuroimaging thus faces the challenge of separating theoretically important differences in neural activity (such as those that may reflect things like neural myelination and cognitive strategy) and those that are theoretically uninteresting (as diverse as baseline differences in glucose metabolism and boredom/frustration). The results of this study are therefore difficult to interpret – a wide variety of regions showed different patterns of activity, and it’s hard to know which are worthy of attention.
Forunately, a few important trends in the data were remarkably consistent. Children tended to recruit left hemispheric regions for both types of inhibition tasks, whereas adults tended to recruit right hemispheric regions. Because left lateralized regions are thought to be important for linguistic processing, Bunge et al. suggest this difference may reflect a verbal strategy among children in which they verbally encode the intended response, in order to minimize interference from irrelevant items in the display. Further support for this claim comes from the fact that task performance was related to verbal reasoning ability, but only in children.
In adults, inhibitory control has been linked to prefrontal regions – yet better-performing children in the current study more strongly activated posterior rather than frontal brain regions. It actually seems that prefrontal activity was stronger for children that performed worse at the task, perhaps suggesting that they were exerting reactive control (i.e., reacting to the need for inhibition rather than actively preparing throughout the task, as adults seem to do).
An alternative possibility for the fact that children and adults recruited different hemispheres for the same tasks has to do with the production/monitoring hemispheric asymmetry hypothesis. According to this hypothesis, left prefrontal regions are important for maintaining online the information that will bias more posterior regions, either for the purposes of memory encoding, selection from memory, or the maintenance of task goals. In contrast, the theory suggests that right prefrontal regions may be more responsible for monitoring and reorienting processes.
Due to their lower memory and attentional abilities, children may have had difficulty remembering the rules of the task (indeed, with four possible task constantly switching, this could be a challenging task for many adults). If this were the case, the selection/monitoring theory suggests that left prefrontal regions should be more active for children than adults, precisely as was found. Conversely, adults may be more actively monitoring for the presence of a task-switch, a no-go trial, or for the presence of interference in the display (as in the incongruent trials) and thus show more right lateralized activity, which was also found.