EVestG-derived action potential features as potential biomarkers for monitoring patients with Alzheimer’s disease in response to gamma tACS therapy

Abstract

Alzheimer's disease (AD) is a progressive neurodegenerative disorder for which no definitive cure currently exists. Transcranial alternating current stimulation (tACS) at gamma frequency (40 Hz), combined with cognitive exercises (CE), has emerged as a promising non-invasive neuromodulation approach, with clinical studies reporting improvements in cognitive function. The vestibular system, an evolutionarily ancient sensory system, has extensive projections to cortical regions involved in cognition, and growing evidence links vestibular dysfunction to cognitive decline in AD. Electrovestibulography (EVestG) is a non-invasive technique that records vestibulo-acoustic neural activity and enables extraction of mini action potential (AP) waveform features reflecting underlying neural dynamics. Previous studies have demonstrated that EVestG-derived features can distinguish between AD and AD with cerebrovascular symptomatology (AD-CVD), suggesting sensitivity to both disease state and cerebrovascular burden. This study investigates whether EVestG-derived AP features can capture differential neurophysiological responses to active versus sham tACS combined with cognitive exercises, using data from a double-blind, crossover clinical trial. Participants with AD and AD-CVD were assessed at baseline, post-treatment, and follow-up sessions. The analysis comprised two parts: AP area changes across pre-selected segments, followed by multivariate decomposition of AP waveform features using Principal Component Analysis (PCA) and Independent Component Analysis (ICA) to identify dominant and independent patterns across treatment conditions. Results suggest that active tACS+CE was associated with more structured neurophysiological patterns compared to sham tACS+CE, although most comparisons did not reach statistical significance given the limited sample size. Timing-related features, particularly depolarization and repolarization rise times, were especially sensitive to tACS intervention. Subgroup analysis based on the modified Hachinski Ischemic Score (HIS) revealed distinct response patterns, with participants having higher cerebrovascular burden showing delayed dynamics compared to those with lower HIS. Dynamic EVestG segments were more sensitive to intervention than static segments. These exploratory findings support the potential of EVestG-derived AP features as candidate biomarkers for monitoring neurophysiological responses to neuromodulation, emphasizing the need to consider cerebrovascular status in treatment evaluation.