Longitudinal changes in resting‐state EEG are correlated with cognitive decline in Alzheimer's disease and related dementia: Biomarkers (non‐neuroimaging) / Longitudinal change over time. (7th December 2020)

• Record Type: Journal Article
• Authors: Meghdadi, Amir H.; Berka, Chris; Richard, Christian; Karic, Marija Stevanovic; Rupp, Greg; Mcconnell, Marissa; Poole, Jared; Fink, Abigail; Hamilton, Joanne M; Salat, David H
• Is Part Of: Alzheimer's & dementia. Volume 16(2020)Supplement 5
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-12-07
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.044113
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: Resting‐state EEG measures such as spectral power in Theta (3‐7 Hz) and Alpha (8‐13 Hz) bands have been previously linked to cognitive decline in Alzheimer's disease (AD) and other dementias. However, routine use of EEG in clinical settings has not been widely practiced. In this ongoing longitudinal study, we demonstrate how changes in EEG measures after 1‐year are correlated with actual changes in cognitive decline as measured by MMSE score of cognitive assessment. Method: Five minutes of eyes‐closed resting‐state EEG were recorded during both an initial and one‐year follow‐up visits from Healthy Controls (n=67), individuals with Mild Cognitive Impairment (n=16), Alzheimer's Disease (n=4), Dementia with Lewy‐body (n=1), Parkinson's disease dementia (n=1), and subjects reporting memory problems without a clear diagnosis (n=7). A generalized‐linear‐model (GLM) was used to regress the relationship between predictors (MMSE and age at initial visit as well as longitudinal changes in 9 EEG measures that were selected a priori ) and the outcome variable (actual changes in MMSE score after 1‐year). EEG measures included Theta‐Alpha Ratio (TAR) and relative power at 5 Hz (theta5 ) at temporal channels as well as relative power at 10Hz at POz channel. Result: MMSE scores fell in 32% of participants and improved in 8% (Figure 1). A GLM using logarithmic link‐function with 11 predictors and 21 terms (EEG predictors and their interaction with age) was fitted to the data (F=4.24, p=2.44x10 ‐6 ). The most significant (p<0.01) predictors were age, theta5, TAR and their interaction with age, all at channel T6. Predicted and actual decline in MMSE were highly correlated (r=0.73, p=10 ‐5 ) (Figure 2). A simpler model with 12 terms (no interaction between predictors) resulted in (F=4.34, p=3.98x10 ‐5 ). Conclusion: Predictive power of EEG in modeling cognitive decline was demonstrated in a cohort of patients with known or possible dementia diagnosis as well as controls. These results support the utility of EEG as a biomarker of cognitive decline. These markers could supplement other neuropathological biomarkers (such as beta‐amyloid and tau), particularly because changes in cognition may not necessarily happen at the same rate as pathological changes.