Connectivity changes in two-channel prefrontal ERP associated with early cognitive decline in the elderly population: beta band responses to the auditory oddball stimuli

Abstract

Background: This study utilized recent advancements in electroencephalography (EEG) technology that enable the measurement of prefrontal event-related potentials (ERPs) to facilitate the early detection of mild cognitive impairment (MCI). We investigated two-channel prefrontal ERP signals obtained from a large cohort of elderly participants and compare among cognitively normal (CN), subjective cognitive decline (SCD), amnestic MCI (aMCI), and nonamnestic MCI (naMCI) groups.

Methods: Signal processing and ERP component analyses, specifically adapted for two-channel prefrontal ERP signals evoked by the auditory oddball task, were performed on a total of 1,754 elderly participants. Connectivity analyses were conducted to assess brain synchronization, especially in the beta band involving the phase locking value (PLV) and coherence (COH). Time-frequency, time-trial, grand average, and further statistical analyses of the standard and target epochs were also conducted to explore differences among the cognition groups.

Results: The MCI group's response to target stimuli was characterized by greater response time variability (p < 0.001) and greater variability in the P300 latency (p < 0.05), leading to less consistent responses than those of the healthy control (HC) group (CN+SCD subgroups). In the connectivity analyses of PLV and COH waveforms, significant differences were observed, indicating a loss of synchronization in the beta band in response to standard stimuli in the MCI group. In addition, the absence of event-related desynchronization (ERD) indicated that information processing related to readiness and task performance in the beta band was not efficient in the MCI group. Furthermore, the observed decline in the P200 amplitude as the standard trials progressed suggests the impaired attention and inhibitory processes in the MCI group compared to the HC group. The aMCI subgroup showed high variability in COH values, while the naMCI subgroup showed impairments in their overall behavioral performance.

Conclusion: These findings highlight the variability and connectivity measures can be used as markers of early cognitive decline; such measures can be assessed with simple and fast two-channel prefrontal ERP signals evoked by both standard and target stimuli. Our study provides deeper insight of cognitive impairment and the potential use of the prefrontal ERP connectivity measures to assess early cognitive decline.

Keywords: EEG beta band; auditory oddball paradigm; brain connectivity; event-related potential; mild cognitive impairment; synchronization; two-channel prefrontal EEG.

FIGURE 1

Experimental workflow in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement. HC, healthy control; CN, cognitively normal; SCD, subjective cognitive declines; MCI, mild cognitive impairment; aMCI, amnestic mild cognitive impairment; naMCI, non-amnestic mild cognitive impairment.

FIGURE 2

The ANCOVA results of the 4 subgroups. Only the variables that were analyzed and showed significant differences are displayed. Age, sex, and education level were used as covariates, and the Bonferroni method was employed for post-hoc analysis. *p < 0.05, **p < 0.01.

FIGURE 3

Grand average ERP waveforms for both standard and target epochs in the (a) 2 main groups and (b) 4 subgroups.

FIGURE 4

Grand average PLV waveforms for both standard and target epochs in the (a) beta band, (b) alpha band, and (c) theta band with significant difference time intervals (marked with an asterisk) based on the comparison of averages across seven 100 ms intervals between the 2 main groups and in the (d) beta band among the 4 subgroups. All time points on the grand average PLV waveform, where significant differences between the 2 main groups were observed, are marked by the yellow area. *P < 0.05.

FIGURE 5

Grand average COH waveforms for both standard and target epochs with significant difference frequency ranges (marked with an asterisk) based on the comparison of averages across seven frequency ranges between the (a) 2 main groups and (b) 4 subgroups. All frequency ranges on the grand average COH waveform, where significant differences between the 2 main groups were observed, are marked by the yellow area. *P < 0.05.

FIGURE 6

Grand average TF plots of ERSP for standard epochs in the (a) 2 main groups and (b) 4 subgroups, and for target epochs in the (c) 2 main groups and (d) 4 subgroups. Significant values compared to the baseline window are indicated in colors, while non-significant values are plotted in green.

FIGURE 7

Grand average TF plots of ITC for both standard and target epochs in the 2 main groups. Significant values compared to the baseline window are indicated in colors, while non-significant values are plotted in green.

FIGURE 8

Grand average TT plots for standard epochs in the (a) 2 main groups and (b) 4 subgroups, and for target epochs in the (c) 2 main groups and (d) 4 subgroups.