Working memory and reasoning tasks are associated with different modes of large-scale brain dynamics in healthy older adults

Abstract Researchers have proposed that solving complex reasoning problems, a key indicator of fluid intelligence, involves the same cognitive processes as solving working memory tasks. This proposal is supported by an overlap of the functional brain activations associated with the two types of tasks and by high correlations between inter-individual differences in performance. We replicated these findings in fifty-three older subjects but also showed that solving reasoning and working memory problems favors different modes of large-scale neurodynamics and that this dissimilarity increases with higher difficulty load. Specifically, superior performance in a typical working memory paradigm (n-back) was associated with up-regulation of modularity (increased between-network segregation), whereas performance in the reasoning task was associated with effective down-regulation of modularity. We also showed that working memory training promotes task-invariant increases in modularity. Since superior reasoning performance is associated with down-regulation of modular dynamics, training may thus have fostered an inefficient way of solving the reasoning tasks. This could help explain why working memory training does little to promote complex reasoning performance. The study concludes that complex reasoning abilities cannot be reduced to working memory and suggests the need to reconsider the feasibility of using working memory training interventions to attempt to achieve effects that transfer to broader cognition. Significance statementReasoning ability (fluid intelligence) is important to many real-life outcomes. Understanding the cognitive and neural processes involved in reasoning is therefore of great interest. Some have suggested that reasoning is structurally similar to working memory (WM). As expected, we found that performance in reasoning and WM tasks was highly correlated and brain activation patterns overlapped. However, complex reasoning performance was better when brain networks were interacting in a more globally integrated way, whereas the opposite was true of WM. WM training increased the level of between-network segregation in ways that did not vary by task, which could be one reason why WM training did little to promote complex reasoning performance. Complex reasoning cannot be reduced to WM.