Resting-state functional connectivity correlates of emotional memory control under cognitive load in subclinical...

Volitional memory control supports adaptive cognition by enabling intentional Suppression of goal-irrelevant, interfering memories and Recall of goal-relevant memories. Neural mechanisms of Suppression and Recall have been studied largely in isolation, and their operation under concurrent working memory load in the context of subclinical anxiety remains unclear. We examined the control of emotionally valenced memories in 47 healthy participants with varying levels of subclinical anxiety under dual-task conditions, involving directed Suppression and Recall while concurrently performing a secondary task that imposed a visual working memory load. Cognitive efficiency in controlling dual-task memory-linked interference, measured by the Balanced Integration Score (BIS), showed no differences between Suppression vs. Recall; across emotions or influence from anxiety. Intrinsic functional brain networks measured by seed-to-voxel resting-state functional connectivity (rsFC) revealed dissociable rsFC profiles linked to Cognitive Control across emotional valences, and these relationships were moderated by anxiety. Efficient Suppression of positive memories correlated with reduced connectivity between anterior cingulate cortex and posterior perceptual-midline regions, and diminished hippocampal-frontal pole coupling. Conversely, efficient Suppression of negative memories correlated with increased posterior parietal - lateral occipital regional connectivity. Anxiety particularly moderated associations between Cognitive Control and prefrontal connectivity during Suppression of positive memories and Recall of positive and neutral memories. Direct comparisons further revealed stronger hippocampal-thalamic rsFC during Suppression vs. Recall of positive memories. Together, we delineate the neural correlates of volitional emotional memory control under cognitive load and suggest that subclinical anxiety severity shapes these networks selectively, providing targets for future mechanistic studies.