Assessment for executive dysfunction in the early stages of Alzheimer's disease : an EEG study using event-related potential analysis

Abstract

Early and effective diagnosis is critical in the prevention of Alzheimer’s disease (AD); however, routinely available assessments are not sensitive enough to detect the subtle cognitive changes involved in the early stages of AD. Consequently, the early stages of AD, including mild cognitive impairment (MCI) and subjective cognitive decline (SCD), cannot be easily distinguished from normal control (NC). Previous studies indicated that executive function (EF) may be impaired in patients with AD, and EF dysfunction can be assessed using task performance as well as event-related potentials (ERPs) derived from an electroencephalograph (EEG). Therefore, a validated ERP protocol – containing three paradigms including Go, No-go, and prospective memory (PM) paradigms – was developed in assessing three core elements of EF: attention, inhibitory control and prospective memory., This study investigated whether the ERPs during task performance can effectively detect executive dysfunction in the early stages of AD. We recruited a total of 67 participants (30 patients with SCD, 17 with MCI and 20 NC) in the study. Each participant was required to complete both a set of neuropsychological battery and computerized tasks with EEG recording. From the neuropsychological battery results, we found that the global cognition level of MCI was lower than that of SCD and NC. Besides, MCI also yielded lower scores on Story Recall (immediate and delayed), Number Span (forward), Category Fluency (animal), Trail Making Task (part B), and Delayed Benson Figure Copy tasks. There were no significant group differences between SCD and NC on the performance of the neuropsychological battery. From the task performance results, we found significant decline in the behavioural performance (accuracy, reaction time and number of trials) in all three paradigms in MCI but not SCD, as compared with NC. From the ERP results using the Go/No-go paradigm, we found statistically significant increments in the amplitude of both the P300 (an essential cognitive parameter) and P200 (associated with sensory processing) in NC compared with SCD and MCI. We also identified higher N200 amplitudes in MCI and SCD compared with NC in both Go and No-go tasks; this might indicate differences in inhibitory control ability among groups. There was also a greater P300 latency in SCD and MCI, which might indicate a longer processing time for patients with early stages of AD. In the PM task, NC exhibited larger N300/frontal positivity and parietal positivity amplitudes over the entire brain region, which might indicate a decline in attention and prospective memory function for patients with early stages of AD. In summary, our findings of descending amplitudes may indicate both sensory and functional insufficiencies for patients with early stages of AD; and these may translate into specific dysfunctions of EF. In addition, ERP components present stronger correlation with general cognition level, executive ability and visual memory. In conclusion, we found that the amplitude of N300/frontal positivity and parietal positivity in PM task as well as N200 and P200 in Go/No-go task may reflect subtle executive function alternations, and may therefore be useful in distinguishing the earliest stages of AD from NC.