An ALE meta-analytic comparison of verbal working memory tasks

TL;DR

This meta-analysis compares brain activation patterns in different verbal working memory tasks, revealing distinct neural regions involved in the n-back and PASAT tasks, including cognitive and emotional processing areas.

Contribution

First ALE meta-analyses to compare neural activation in n-back and PASAT tasks, highlighting differences in cognitive and emotional brain regions involved.

Findings

PASAT activates regions linked to phonological loop and emotional processing.

3-back shows higher activation in frontal eye fields and SMA.

PASAT exhibits greater activation in SMA, supramarginal gyrus, and parietal lobule.

Abstract

Background: The n-back and Paced Auditory Serial Addition Test (PASAT) are commonly used verbal working memory tasks that have partially overlapping uses in clinical and experimental psychology. We performed three activation likelihood estimation (ALE) meta-analyses, comparing two load levels of the n-back task (2-back, 3-back) to the PASAT and to each-other. These analyses aimed to determine the involvement of cognitive and emotional brain regions for these tasks. Results: We observed higher overall likelihood of activation the frontal eye fields in the 3-back. The PASAT exhibited higher overall activation in the bilateral supplementary motor areas (SMA), left supramarginal gyrus, and left superior parietal lobule. Furthermore, the 3-back exhibited higher activation in the right SMA, and anterior mid-cingulate cortex versus the 2-back, and the PASAT exhibited higher activation in a cluster near the right premotor area versus the 2-back. A laterality effect was observed in the dorsolateral prefrontal cortex between the PASAT (left) and 3-back(right). These data suggest greater activation of regions traditionally associated with the phonological loop during the PASAT, compared to the 2- and 3-back tasks. Furthermore, individual ALE analyses suggest involvement of emotional processing and salience network regions (insula, cingulate) in addition to the well-established verbal working memory regions (Broca's region, bilateral SMA, premotor, posterior parietal cortices) in all 3 tasks. Conclusions: Here we identify regions activated by the PASAT, which has not been meta-analytically reviewed prior to this study. Using ALE meta-analysis, we have also identified meaningful differences in activation associated with specific cognitive and emotional aspects of verbal working memory during these tasks.