Early stages of cognitive decline are characterized by different EEG microstates dynamics

Neural changes due to Alzheimer's Disease (AD) appear much earlier than objective cognitive decline symptoms. Subjective cognitive decline (SCD) is the first pre-clinical stage at which future development of the disease may be intercepted, possibly followed by mild cognitive impairment (MCI), the first step of clinical cognitive decline. Recent studies suggest that resting EEG analysis tools describing whole-brain activity, such as microstates and brain areas functional connectivity, might support a characterization of SCD and MCI conditions, but notwithstanding their clinical relevance, the cortical underpinnings of these AD prodromic states are far from being untangled. Here we verify whether the EEG activity modeled as a sequence of microstates is a better marker than simple spectral analyses in discriminating between SCD (n = 57), MCI (n = 45) and healthy controls (n = 17). While no significant differences were found in the spectral activity of our cohort, we identified effective markers of the prodromic changes of the disease in the reconstructed microstates dynamics. Specifically, cognitive decline alters the average duration, the coverage and the transition probabilities of several microstates, with the most prominent differences found for the canonical microstate C. Overall, this work advances the search for non-invasive EEG biomarkers charaterizing presymptomatic stages of AD, that may be used to automatically diagnose the future development of the disease.