The impact of mild cognitive impairment on the neural dynamics serving verbal working memory

Abstract

AbstractBackgroundIn healthy adults, neural oscillatory activity within a predominantly left‐lateralized network of brain regions underlies verbal working memory (VWM) performance, but how mild cognitive impairment (MCI) impacts these oscillatory dynamics is not well characterized. The present study utilized the spatiotemporal precision of magnetoencephalographic (MEG) brain imaging to investigate the effects of MCI on the neural oscillations serving specific phases (i.e., encoding, maintenance) of VWM. In congruence with the compensation‐related utilization of neural circuits hypothesis (CRUNCH), we hypothesized that MCI patients would recruit greater neural resources during VWM.MethodSixty‐one adults (13 clinically‐diagnosed MCI, 48 controls; 28 females; M age: 63.5) completed a Sternberg‐type VWM task during MEG. Only correct trials were analyzed, and the average number of trials included did not differ between groups. All MEG data underwent standard preprocessing, were transformed into the time‐frequency domain, and significant neural oscillatory responses relative to baseline were imaged using a beamformer. To determine the effect of group (MCI vs. controls), ANCOVAs were performed on the resulting encoding and maintenance whole‐brain maps with age as a covariate.ResultThroughout encoding and maintenance, decreases in alpha/beta (9‐16 Hz) oscillatory activity were seen in left fronto‐temporal regions, which were significantly larger in MCI patients during the encoding phase (p < .05, corrected). Additionally, significant group differences were detected in right inferior frontal and superior temporal cortices throughout both encoding and maintenance, such that MCI patients exhibited decreased alpha/beta activity within these regions, while these responses were largely absent in controls (p < .05, corrected). Interestingly, the strength of the alpha/beta response within the right inferior frontal cortex significantly predicted behavioral performance on the task, such that individuals who demonstrated stronger decreases in alpha/beta within this region during encoding and maintenance, tended to have faster reaction times (p < .01).ConclusionWhile both groups recruited a similar left‐hemispheric cortical network during VWM encoding and maintenance, individuals with MCI recruited some nodes more strongly and additionally recruited right fronto‐temporal regions. Additionally, stronger recruitment of the right inferior frontal cortex was beneficial for task performance. These results offer supporting evidence for CRUNCH.