Abstract
Background/Objective: Timely and successful treatments for Alzheimer's disease (AD) depend on early detection. The Multimer Detection System (MDS-OAβ) for quantifying plasma oligomeric amyloid-β (OAβ) has shown promise as a biomarker of amyloid disease. The theta-to-alpha ratio (TAR) and theta-to-beta ratio (TBR) are two examples of spectral power metrics that can be used in resting-state quantitative EEG (qEEG) to evaluate brain function non-invasively. This study used resting-state EEG (rEEG) recordings obtained while the subjects were both eyes-open (EO) and eyes-closed (EC) to investigate the relationship between regional qEEG power ratios and plasma MDS-OAβ levels in older adults. Methods: The analysis comprised 174 patients between the ages of 60 and 85, with 2 in the low-MDS-OAβ group and 82 in the high-MDS-OAβ group. The clinical plasma cutoff was 0.78 ng/mL. All participants underwent rEEG recordings and plasma OAβ quantification. EEG pre-processing included bandpass filtering (0.5-100 Hz), average re-referencing, artifact rejection using independent component analysis (ICA), and spectral power estimation using Welch's method. The TAR and TBR were calculated across five lobar regions (frontal, central, parietal, occipital, and temporal) during both EO and EC conditions. To normalize data distributions, EEG ratio variables were log-transformed prior to statistical analysis. Group comparisons and linear regression analyses were conducted to evaluate the associations between EEG power ratios and MDS-OAβ levels. Adjusted regression models included age, years of education, and neuropsychological test scores as covariates. Statistical significance was set at p < 0.05. Results: No significant associations were found between TAR and plasma MDS-OAβ levels across any lobar regions under either EO or EC conditions. In contrast, TBR exhibited consistent and significant negative associations with MDS-OAβ levels, particularly under EC conditions. Adjusted regression models revealed that higher MDS-OAβ levels were associated with lower TBR values in the central (β = -0.059, p = 0.015), parietal (β = -0.072, p = 0.006), occipital (β = -0.067, p = 0.040), and temporal (β = -0.053, p = 0.018) lobes, with the strongest inverse relationship observed in the parietal lobe. A similar, though slightly weaker, pattern was observed during EO conditions, with significant inverse associations in the frontal, central, and temporal lobes. Conclusions: Our findings indicate that, after adjusting for covariates, increased plasma MDS-OAβ levels are significantly associated with a reduced TBR, particularly in the parietal and central lobes, under both EO and EC resting-state conditions. In contrast, no significant associations were observed with TAR. These results suggest that a lower TBR may reflect an increased peripheral amyloid burden and highlight its potential as a sensitive qEEG biomarker for early amyloid-related brain changes in older adults.