EEG as a Biomarker in Alzheimer’s Disease: New Insights from Multimodal Research
- Angel David Blanco
- 6 hours ago
- 5 min read
At Starlab Barcelona, we believe that electroencephalography (EEG) holds significant, and still largely untapped, potential as a biomarker for Alzheimer’s disease (AD). As the field moves toward earlier diagnosis, preventive trials, and longitudinal monitoring, there is a growing need for biomarkers that are not only biologically meaningful, but also scalable, affordable, and suitable for repeated use.
A recent systematic review by Sohrabpour et al. (2025) provides timely and compelling evidence supporting this vision. By integrating EEG and magnetoencephalography (MEG) findings with established biomarkers such as amyloid-β (Aβ), tau, positron emission tomography (PET), and cerebrospinal fluid (CSF) measures, the authors offer a nuanced view of how brain electrophysiology evolves across the Alzheimer’s disease continuum.
This work reinforces a key idea that resonates strongly with our research philosophy at Starlab: electrophysiological changes in AD are dynamic, non-linear, and deeply intertwined with underlying neurobiology. Understanding this complexity is essential if EEG-based biomarkers are to reach their full clinical potential.
Alzheimer’s disease and the need for earlier, more accessible biomarkers
Alzheimer’s disease is a progressive neurodegenerative disorder characterized by the accumulation of Aβ plaques and tau neurofibrillary tangles, leading to synaptic dysfunction, neuronal loss, and cognitive decline. Crucially, these pathological processes begin years—if not decades—before clinical symptoms become evident.
Current gold-standard biomarkers, such as PET imaging and CSF assays, have transformed our understanding of AD pathophysiology. However, their widespread use is limited by cost, invasiveness, availability, and suitability for repeated measurements. As a result, there is growing interest in complementary biomarkers that can capture early functional changes in the brain while remaining practical for large-scale and longitudinal applications.
EEG and MEG directly measure neuronal population activity with millisecond temporal resolution, providing a window into brain network dynamics that cannot be accessed through structural imaging alone. The review by Sohrabpour and colleagues highlights how these electrophysiological signals reflect both core AD pathology and its evolution over time.
Core electrophysiological signatures of Alzheimer’s disease
Across decades of research, several electrophysiological hallmarks of AD have emerged consistently. The review confirms and contextualizes these findings within a multimodal framework:
Spectral slowing: Increased power in low-frequency bands (δ and θ) alongside reduced power in higher-frequency bands (α and β). Figure 1 illustrates an example analysis conducted by Starlab, highlighting this spectral slowing, which is especially pronounced in patients with Alzheimer’s disease.
Altered synchrony and connectivity: Disruptions in functional coupling within and between large-scale brain networks.
Correlations with pathology and cognition: Changes in oscillatory power and synchrony are associated with Aβ and tau burden, cognitive performance, and cerebral metabolism measured with FDG-PET.
Importantly, these electrophysiological alterations are not merely epiphenomena. They reflect fundamental changes in excitation–inhibition balance, synaptic integrity, and network communication—processes that lie at the heart of AD pathophysiology.
The mystery of alpha oscillations: resolving apparent contradictions
One of the most intriguing aspects discussed in the review is the so-called “mystery of alpha (α) oscillations.” While many studies report reduced α power and synchrony in mild cognitive impairment (MCI) and AD, others describe an initial increase in α activity during earlier stages.
At first glance, these findings appear contradictory. However, when viewed through the lens of disease staging and multimodal data, a more coherent picture begins to emerge.
A non-linear trajectory across
disease stages
The authors propose that α oscillations may follow a non-monotonic trajectory:
Early stages (subjective cognitive decline or preclinical AD): Rising Aβ levels may induce cortical disinhibition and neuronal hyperexcitability, leading to increased α activity.
Later stages (MCI and AD dementia): As tau pathology spreads and neurodegeneration advances, excitatory neurons are progressively compromised, resulting in declining α and β power and overall spectral slowing.
This framework aligns with the Aβ/tau cascade hypothesis and is supported by computational models showing how changes in inhibitory synaptic coupling and interneuron activity can drive early hyperactivity followed by network collapse.
Alpha is not a single rhythm
A key insight emphasized in the review is that alpha oscillations are not a unitary phenomenon. Instead, they arise from multiple spatially distributed generators with distinct frequencies and functional roles. Several studies report:
Posterior α slowing alongside anterior α increases in power or synchrony
The emergence of multiple α peaks in MCI and early AD, which may predict progression to dementia
These findings suggest that different α networks may be differentially affected by Aβ and tau pathology, desynchronizing over time as regional vulnerability and cytoarchitecture come into play.
Non-monotonous disease dynamics and the limits of static biomarkers
Beyond alpha oscillations, the review highlights a broader principle: AD progression is fundamentally non-linear. Both electrophysiological measures and molecular biomarkers such as CSF Aβ and tau exhibit non-monotonous trajectories across disease stages.
This complexity poses a challenge for biomarker development. Static thresholds or one-size-fits-all classifiers may fail to capture the true biological state of an individual if disease stage is not carefully considered.
Multimodal studies—although still relatively rare—are uniquely positioned to address this challenge. By linking EEG/MEG features to PET, CSF, and cognitive measures, they reveal patterns that would remain invisible in single-modality analyses.
The review also raises important open questions, including the role of glial dysfunction. Astrocytes and microglia play a critical role in synaptic regulation and protein clearance, and their failure may indirectly shape electrophysiological abnormalities through neuroinflammation and disrupted homeostasis.
Why EEG matters: strengths and limitations as a biomarker
The authors provide a balanced assessment of the promise and current limitations of electrophysiological biomarkers.
Key strengths
Sensitivity to early functional changes, often preceding structural atrophy or metabolic decline
Non-invasive and safe, enabling repeated measurements across the lifespan
Cost-effective and widely available, particularly in the case of EEG
Direct reflection of neural circuit function, rather than downstream consequences
These properties make EEG especially attractive for longitudinal monitoring, large cohort studies, and preventive clinical trials.
From research to clinical reality: Starlab’s perspective
At Starlab Barcelona, we see this work as a strong validation of an approach we actively pursue: combining advanced electrophysiology, multimodal integration, and mechanistic insight to move EEG-based biomarkers closer to real-world impact.
EEG is non-invasive, low-cost, repeatable, and scalable. When interpreted through the appropriate biological and clinical context, it has the potential to become a key component of the Alzheimer’s biomarker ecosystem—complementing PET, fluid biomarkers, and digital assessments.
As the field advances toward earlier detection and intervention, electrophysiology offers a unique opportunity to monitor brain dysfunction as it unfolds, rather than only its downstream consequences.
Comments