Individual alpha peak frequency tracks Alzheimer's disease progression: A longitudinal pilot study

Abstract

Background Early and accurate tracking of Alzheimer's disease (AD) progression is critical for timely intervention. However, electrophysiological biomarkers capable of capturing long-term neurodegenerative changes remain largely underexplored. Objective We investigated whether individual alpha peak frequency (IAPF), an electroencephalography (EEG)-derived measure of dominant neural oscillatory activity, could serve as a longitudinal biomarker of AD progression. Methods Twenty-seven patients with AD aged 63–91 years underwent annual EEG and cognitive assessments over 2–7 years. IAPF was extracted from eyes-closed resting-state EEG. Longitudinal associations among IAPF, Mini-Mental State Examination (MMSE) scores, age, and follow-up time were evaluated using repeated-measures correlation and linear mixed-effects models. Annual IAPF changes were compared with those of healthy controls (HC) aged 20–70 years, stratified by decade-based age subgroups. Longitudinal changes in relative spectral power were also analyzed. Results Patients with AD showed significant longitudinal declines in both IAPF and MMSE scores, with a positive longitudinal association between the two measures. Mixed-effects models indicated that these declines were better explained by follow-up time after accounting for baseline age than by age at assessment alone. Compared with all healthy-control age subgroups, patients with AD exhibited a significantly steeper annual IAPF decline. Relative theta power increased and alpha/beta power decreased over follow-up, consistent with spectral slowing. However, annualized spectral-power changes showed limited disease specificity, with significant AD–HC differences only for delta and alpha power relative to the oldest HC subgroup. Conclusions These findings support IAPF as a non-invasive, temporally sensitive, and clinically accessible biomarker for monitoring AD progression.