Neurophysiological markers in community-dwelling older adults with mild cognitive impairment: an EEG study

Abstract

Background: Neurodegeneration and structural changes in the brain due to amyloid deposition have been observed even in individuals with mild cognitive impairment (MCI). EEG measurement is considered an effective tool because it is noninvasive, has few restrictions on the measurement environment, and is simple and easy to use. In this study, we investigated the neurophysiological characteristics of community-dwelling older adults with MCI using EEG.

Methods: Demographic characteristics, cognitive function, physical function, resting-state MRI and electroencephalogram (rs-EEG), event-related potentials (ERPs) during Simon tasks, and task proportion of correct responses and reaction times (RTs) were obtained from 402 healthy controls (HC) and 47 MCI participants. We introduced exact low-resolution brain electromagnetic tomography-independent component analysis (eLORETA-ICA) to assess the rs-EEG network in community-dwelling older adults with MCI.

Results: A lower proportion of correct responses to the Simon task and slower RTs were observed in the MCI group (p < 0.01). Despite no difference in brain volume between the HC and MCI groups, significant decreases in dorsal attention network (DAN) activity (p < 0.05) and N2 amplitude of ERP (p < 0.001) were observed in the MCI group. Moreover, DAN activity demonstrated a correlation with education (Rs = 0.32, p = 0.027), global cognitive function (Rs = 0.32, p = 0.030), and processing speed (Rs = 0.37, p = 0.010) in the MCI group. The discrimination accuracy for MCI with the addition of the eLORETA-ICA network ranged from 0.7817 to 0.7929, and the area under the curve ranged from 0.8492 to 0.8495.

Conclusions: The eLORETA-ICA approach of rs-EEG using noninvasive and relatively inexpensive EEG demonstrates specific changes in elders with MCI. It may provide a simple and valid assessment method with few restrictions on the measurement environment and may be useful for early detection of MCI in community-dwelling older adults.

Keywords: Accuracy; Community-dwelling older adults; Electroencephalography; Mild cognitive impairment; eLORETA-ICA.

Fig. 1

Overview of the Simon task

Fig. 2

Descriptive statistics for behavioral data. A The proportion of correct responses in the Congruent condition. B The proportion of correct responses in the incongruent condition. C shows the proportion of correct responses in the no response condition. D The RT in the Congruent condition. E The RT in the incongruent condition. Each violin plot contains a boxplot. The black line within the box represents the median; the box in the center represents the interquartile range; the black dot depicts the remaining distribution, except for any data points identified as “outliers” (i.e., those more than 1.5 standard deviations above or below the median). HC, healthy controls; MCI, mild cognitive impairment. *p < 0.01; **p < 0.001; n.s. not significant

Fig. 3

Sample images of 15 independent components (ICs) in their specified frequency bands obtained by applying eLORETA-ICA to EEG data. Twelve ICs were identified as physiological network activities: IC-2, IC-3, IC-5, IC-6, IC-7, IC-8, IC-9, IC-10, IC-11, IC-13, IC-14, and IC-15 and 3 ICs as artifact activities: IC-1, IC-4, and IC-12. In the color-coded map, red and blue represent increase and decrease in power with increasing IC activity, respectively

Fig. 4

Scatterplot of the DAN (IC-14) activity values with demographic characteristics, cognitive function, and behavioral data. A Age. B Education. C Proportion of correct responses in the congruent condition. D Proportion of correct responses in the incongruent condition. E Proportion of correct responses in the no response condition. F RT in the congruent condition. G RT in the incongruent condition. H MMSE. I Word list memory. J TMT-A. K TMT-B. L SDST