Behavioral facilitation and increased brain responses from a high interference working memory context

Abstract

Many real-life situations require flexible behavior in changing environments. Evidence suggests that anticipation of conflict or task difficulty results in behavioral and neural allocation of task-relevant resources. Here we used a high- and low-interference version of an item-recognition task to examine the neurobehavioral underpinnings of context-sensitive adjustment in working memory (WM). We hypothesized that task environments that included high-interference trials would require participants to allocate neurocognitive resources to adjust to the more demanding task context. The results of two independent behavioral experiments showed enhanced WM performance in the high-interference context, which indicated that a high-interference context improves performance on non-interference trials. A third behavioral experiment showed that when WM load was increased, this effect was no longer significant. Neuroimaging results further showed greater engagement of inferior frontal gyrus, striatum, parietal cortex, hippocampus, and midbrain in participants performing the task in the high- than in the low-interference context. This effect could arise from an active or dormant mode of anticipation that seems to engage fronto-striatal and midbrain regions to flexibly adjust resources to task demands. Our results extend the model of conflict adaptation beyond trial-to-trial adjustments by showing that a high interference context affects both behavioral and biological aspects of cognition.

Figure 1

Task performance (reaction time) on non-recent probes (A) across the four experiments. (B) The combined normalized results (z-scores) of the four experiments show that participants who completed the high-interference (interference) version of the task performed better than those who completed the low-interference (control) version. Error bars indicate standard means across participants. Asterisks indicate significant main effects (*P < 0.05; ***P < 0.005).

Figure 2

Task performance (accuracy) on non-recent probes (A) across the four experiments. (B) The combined normalized results (z-scores) of the four experiments show that participants who completed the high-interference (interference) version of the task performed better than those who completed the low-interference (control) version. Error bars indicate standard means across participants. Asterisks indicate significant main effects (**P < 0.01; ***P < 0.005).

Figure 3

fMRI results. Differences in BOLD activity in individuals in the interference group (high-interference context; dark gray bars) and the control group (low-interference context; light gray bars). Between-group whole-brain results are based on first-level contrast of non-recent negative probes and positive probes (for display purposes, thresholded at P < 0.001 uncorrected). BOLD activation (% signal change) extracted from functional ROIs around the local cluster maxima. Error bars indicate standard means across participants.