Associations between Source-space EEG Functional Connectivity and Cognitive Function across the Alzheimer’s Disease Continuum

NTRODUCTION:

    Alzheimer’s disease (AD) is increasingly recognized as a disconnection syndrome.

    Identifying early network-level biomarkers of AD remains a key challenge.

    EEG functional connectivity offers a scalable, non-invasive approach to characterize brain network alterations across the AD cognitive continuum.

METHODS:

    2059 participants in MIND-China, age ≥60 y

    A 3-min resting-state EEG (19 channels), with only the last 1-min analyzed

    Functional connectivity: PLV, wPLI, ImC, AEC, and AEC-c

    Frequency bands: δ, θ, α, and β

    Data analysis: ANCOVA + Bonferroni correction

    Classification: BrainNetCNN + demographic features.

RESULTS:

    MCI was associated with increased theta-band connectivity (frontoparietal & DMN regions)

    AD was associated with widespread beta-band connectivity reduction

    Reduced beta connectivity was correlated with increased MMSE scores

    EEG FC + demographics yielded AUC = 0.85 for differentiating AD from non-dementia.

CONCLUSIONS:

    EEG functional connectivity reflects progressive network disruption across the AD cognitive spectrum

    Increases in theta and decreases in beta represent key frequency-specific signatures

    EEG connectivity + demographics enables robust AD classification

    EEG FC is a promising scalable biomarker for AD.