Beta to Theta Power Ratio in EEG Periodic Components as a Potential Biomarker in Mild Cognitive Impairment and Alzheimer's Dementia

Neurology Biochemistry Basic Medical Sciences Medical Imaging Life Sciences Medicine
Alzheimer's dementia mild cognitive impairment EEG periodic EEG power components biomarkers cognitive function

Alzheimer’s Disease Research and Treatment Frontiers: Beta/Theta Power Ratio in EEG Periodic Components as a Potential Biomarker

Background Introduction

Alzheimer’s dementia (AD) is a progressively developing disease, accounting for 60% to 80% of all dementia cases [1]. In the early stages of AD, mild cognitive impairment (MCI) typically occurs, during which individuals can still live independently [2]. Identifying biomarkers that distinguish MCI from AD or healthy aging is crucial for developing preventive interventions, improving quality of life, reducing caregiver burden, and lowering care costs [3].

Electroencephalogram (EEG) is a non-invasive and cost-effective tool that evaluates neural ionic currents through voltage differences on different spatial scales, featuring high temporal resolution [4-6]. Most studies focusing on EEG in AD and MCI analyze power spectral density (PSD), especially resting-state EEG [5,7-12]. These studies usually find increased delta and theta waves and decreased alpha and beta waves in AD individuals, particularly in the temporal and occipital regions [5,7-14]. The beta/theta power ratio has also been shown to differentiate between AD and normal cognitive individuals, believed to signify cognitive processing capabilities [4-8].

In contrast, EEG’s effectiveness in distinguishing MCI from healthy individuals shows fewer and less consistent differences [9,15-17]. This could be due to these studies not excluding the non-periodic components of EEG. The EEG power spectrum generally comprises two main parts: a background aperiodic component (non-rhythmic component - 1/f-like component) and periodic or rhythmic neural oscillations [19]. The aperiodic component, also known as fractal or “scale-free” activity, exhibits self-similarity across multiple time scales [20]. Several studies have highlighted the benefits of focusing on the periodic components of EEG [15,19,21,22].

Research Source

This study was conducted by Hamed Azami and colleagues from five academic hospitals in Toronto. The research was published in the journal “Alzheimer’s Research & Therapy” (2023, Volume 15, Issue 133). The data were sourced from two interventional studies (clinicaltrials.gov identifiers: nct01847586 and nct02537496) and funded by Brain Canada, Canada Research Chair, and the Labatt Family Network for Research on the Biology of Depression.

Research Design and Process

A) Research Process

  1. Participant Recruitment:

    • The study recruited 44 healthy controls (HC) (average age 69.1), 114 participants with MCI (average age 72.2), and 41 participants with AD (average age 75.7). All participants provided written informed consent, complying with the ethics approval of the Center for Addiction and Mental Health (CAMH) in Toronto.
    • HC and MCI participants were recruited from the AD prevention trial (nct02386670), while AD participants were sourced from two other interventional studies.

  2. EEG Data Collection and Processing:

    • EEG data were recorded for each participant using a 64-channel Synamps 2 EEG device and a 10-10 montage system, recording 10 minutes each at a sampling frequency of 1000 Hz. Participants were seated in a relaxed state with eyes closed, avoiding head or eye movements or falling asleep.
    • Data were processed offline using MATLAB and the EEGLAB toolbox. The EEG data were first visually inspected to ensure no apparent delta and theta waves, then noisily segments and channels severely affected by artifacts were removed. Independent component analysis (ICA) was applied to eliminate components related to eye movements and muscle activity.

  3. Data Analysis:

    • The total power spectrum was calculated using the Welch method, and the results were parameterized using the “FOOOF” toolbox, decomposing the power spectrum into its aperiodic and periodic components, optimizing the model spectrum using the least-squares error method.
    • The beta/theta ratio was obtained by comparing the EEG spectra of the three groups in the overall brain (average across all EEG electrodes) and the frontal, temporal, central, parietal, and occipital regions.

B) Research Results

  1. Aperiodic Components:

    • The study found no significant differences in the aperiodic components of EEG among the HC, MCI, and AD groups (ANCOVA f(3,195)=0.55, p=0.56).

  2. Periodic Components and Overall Power Spectrum:

    • AD participants showed an increase in relative power within the delta, theta, and gamma ranges and a decrease in beta waves, especially in the occipital region.
    • Notably, the beta/theta ratio in the occipital region, measured by the periodic spectrum, showed significant differences between MCI and healthy controls (Bonferroni correction p=0.036), outperforming the beta/theta ratio based on the overall spectrum in classification tasks.

C) Conclusion and Significance

This study demonstrated the potential of the periodic beta/theta power ratio in EEG for distinguishing between healthy individuals, MCI, and AD, suggesting that focusing on periodic components can be more precise in detecting subtle changes in mild disease states. This finding may have significant value in developing more effective preventive measures and mechanistic research.

D) Study Highlights

  1. Findings and Evidence:

    • Focusing on the periodic components of EEG, MCI individuals showed a significantly lower beta/theta power ratio in certain brain regions compared to healthy aging.
    • Classifiers based on periodic components outperformed those based on the overall spectrum in distinguishing AD and MCI individuals.

  2. Methodological and Procedural Innovations:

    • The “FOOOF” method was used to separate the periodic and aperiodic components of EEG, confirming its capability to reveal clearer disease-related changes upon removing aperiodic components.
    • The study showed that EEG periodic component analysis, excluding the aperiodic components, is more reliable in distinguishing MCI and healthy aging.

Study Limitations and Future Prospects

The study has limitations such as insufficient sample size for specific subgroup analyses and the lack of biomarker-based diagnostics. Future research can further validate these findings with larger samples and combined biomarkers.

Conclusion

This study supports further exploration of the beta/theta power ratio in EEG periodic components as a potential neurophysiological indicator distinguishing MCI and AD patients from healthy individuals. Future work needs to link this indicator with biomarkers of other neurodegenerative diseases to further elucidate its mechanistic role.