Task-based fMRI and neuropsychological insights into memory and attention in bipolar disorder: A systematic review

Bipolar Disorder (BD) has been found to be associated with impairments in memory and attention, with functional Magnetic Resonance Imaging (fMRI) studies revealing altered neural circuits underlying these deficits. This study aims to investigate the correlation between in-scan and out-of-scan cognitive assessments, a critical step in bridging the gap between brain activity and cognitive performance.

A systematic search on PubMed, Web of Science, and Scopus for fMRI studies that investigated memory and/or attention in BD using both cognitive assessments inside and outside the MRI scanner was performed. Cognitive performance in the included studies was assessed either continuously using test scores or categorically using established cut-offs or cluster analyses. Participants were evaluated across various clinical states (remitted, partially remitted, symptomatic).

Ten studies met our inclusion criteria. Cognitive performance, assessed through neuropsychological tests administered outside the scanner, showed several significant associations with task-related brain activation. Specifically, reduced working memory (WM) performance was correlated with reduced inferior frontal gyrus (IFG) and ventrolateral prefrontal cortex activation. Lower scores in executive function and episodic memory were associated with medial prefrontal cortex deactivation. Moreover, better WM and executive function correlated with higher activation in the posterior parietal cortex, dorsolateral prefrontral cortex (dlPFC), and IFG. Lastly, poorer cognitive performance was associated with reduced default mode network hypoactivity as well as hypoactivity in the middle temporal gyrus, dlPFC, IFG, temporal cortex, and occipital cortex.

Small sample sizes, lack of longitudinal data, uncontrolled medication effects, task variability and variability in cognitive assessment tools can confound results and limit generalizability.

This review demonstrates that altered activation across prefrontal, temporal, and default-mode networks is consistently associated with poorer cognitive performance in BD, indicating that task-based fMRI signatures can reflect real-world cognitive functioning. These findings highlight the potential value of neural activation patterns as markers of cognitive variation, while also underscoring the need for longitudinal and well-controlled studies to clarify their clinical relevance.