An energy-efficient intrinsic functional organization of human working memory: A resting-state functional connectivity study

• The intrinsic functional organization of the human working memory remains unknown, we hypothesize that it is an energy-efficient system.
• We tested this hypothesis by analyzing associations between WM performance at different task difficulties and the FCD and FCS in 282 young adults.
• These findings suggest that the intrinsic working memory network is an energy-efficient and hierarchical system.

Working memory (WM) is the active maintenance of currently relevant information that was just experienced or retrieved from long-term memory but no longer exists in the external environment; however, the intrinsic functional organization of the brain underlying human WM performance remains largely unknown. We hypothesize that the intrinsic functional organization of human WM is an energy-efficient system. We tested this hypothesis by analyzing associations between WM performance (reaction times of correct responses) at different task difficulties (2-back and 3-back tasks) and the resting-state functional connectivity density (FCD) and strength (FCS) in 282 healthy young adults. Voxel-based FCD analysis showed that the reaction times were negatively correlated with the FCD values of several brain regions known to be engaged in WM performance: the right inferior parietal lobule and inferior frontal gyrus for both the 2-back and the 3-back tasks and the right superior parietal lobule, supramarginal gyrus, left inferior parietal lobule and bilateral middle occipital gyrus for the 3-back task. Further analyses showed that the FCS values of these regions with several frontal, parietal and occipital regions were also negatively correlated with the reaction times; the 3-back task was associated with much more functional connections than the 2-back task. These findings suggest that the intrinsic working memory network is an energy-efficient and hierarchical system. A simple working memory task is controlled only by the core subsystem; however, a complex working memory task is associated with more nodes and connections of the system.