Rethinking EEG Biomarkers: From Disease-Specific Signals to Transdiagnostic Brain Health
- 5 days ago
- 5 min read
For decades, biomarker research in neuroscience has often followed a straightforward idea: if we want to identify a disease, we should look for a biological signal that distinguishes that disease from other conditions.
But the brain does not necessarily organise itself according to our diagnostic categories.
A recent perspective published in Translational Psychiatry proposes a different way of thinking about EEG biomarkers. In “Rethinking EEG biomarkers of brain disorders: a transdiagnostic dimensional view,” Zebhauser, Heitmann, Henningsen and Ploner argue that the clinical value of an EEG biomarker may not depend on being specific to a particular disorder. Instead, biomarkers could capture shared alterations in brain function that cut across diagnostic boundaries.
This transdiagnostic perspective could have important implications for how we develop and use EEG biomarkers in clinical research.
Moving beyond diagnostic categories
Traditional biomarker development often asks a question such as:
Can this EEG feature distinguish people with disorder X from healthy individuals?
The authors propose complementing this approach with another question:
Does this EEG feature reflect a dimension of brain dysfunction that is shared across different disorders?
This distinction is important.
A biological alteration that appears in several disorders might initially be considered a weakness of a biomarker because it cannot distinguish between diagnoses. But if that alteration reflects a common underlying process—and relates meaningfully to symptoms, disease severity, treatment response or functional impairment—its lack of specificity could instead become a strength.
This is the central idea behind transdiagnostic and dimensional biomarkers.
Rather than mapping one biomarker onto one diagnosis, this framework seeks to connect measurable changes in brain function with dimensions such as cognitive dysfunction, negative affect or somatic symptoms that can occur across multiple conditions.
Why might different disorders produce similar EEG signatures?
One of the central examples discussed by the authors is EEG slowing: a shift in brain activity towards slower frequencies, reflected particularly in increased low-frequency activity and theta power. Importantly, this pattern is not specific to a single disorder. Similar alterations have been reported across conditions including chronic pain, migraine, fatigue and depression.
Rather than considering this lack of disease specificity a limitation, the authors propose that EEG slowing may reflect shared dimensions of brain dysfunction that cut across traditional diagnostic categories.
Several mechanisms could contribute to these common alterations, including thalamo-cortical dysrhythmia, excitation–inhibition imbalance and predictive coding. Different disorders may therefore involve different pathological processes that ultimately converge on similar changes in large-scale brain dynamics.
This is where the concept of transdiagnostic dimensional biomarkers becomes particularly relevant: instead of asking whether an EEG feature identifies a specific disease, we can ask whether it captures a dimension of altered brain function that is relevant across different conditions and symptom profiles.
From EEG slowing to clinically useful biomarkers
The important question is ultimately not whether EEG slowing can be observed, but whether it can provide clinically meaningful information.
If a transdiagnostic EEG feature can be reliably measured and linked to symptoms or disease-related outcomes, it could potentially support several aspects of clinical research:
Monitoring changes in brain function over time or following treatment.
Patient stratification by underlying neurophysiological characteristics.
Risk assessment, potentially identifying physiological alterations before symptoms fully emerge.
Treatment guidance, including the identification or monitoring of neural targets for interventions.
These applications still require further validation. In particular, longitudinal and interventional studies are needed to determine whether changes in EEG biomarkers reliably track symptom trajectories and treatment response.
From “What disease does this person have?” to “How does their brain differ from expected?”
This transdiagnostic perspective also connects with normative modelling, an approach that can move EEG biomarkers towards more individualised measures of brain function.
Rather than defining biomarkers exclusively by comparing diagnostic groups, normative models establish reference distributions of EEG features across healthy populations. An individual's EEG can then be evaluated against what would be expected for someone with similar characteristics, such as age.
This changes the question from:
“Does this EEG pattern characterise a particular disorder?”
to:
“How does this individual's brain function deviate from what would be expected?”
For example, rather than simply identifying whether someone shows increased low-frequency activity, a normative model can quantify the extent to which their EEG deviates from the expected range. These individual-level deviations can then be explored in relation to symptoms, disease progression or treatment response.
This does not replace disease-specific biomarkers. Instead, it adds another layer of information—moving from categorical distinctions towards quantitative and individualised measures of brain function.
Why this matters for clinical trials
This shift could be particularly valuable in clinical trials, where biological heterogeneity within the same diagnosis can make it difficult to identify treatment effects.
EEG is well suited to this setting: it is non-invasive, relatively inexpensive, scalable and provides a direct measure of brain activity with high temporal resolution. This makes repeated measurements feasible throughout a study.
Future clinical trials could therefore use EEG not only to ask whether a treatment improves a clinical outcome, but also:
Does the treatment modify an individual's neurophysiological profile?
Can baseline EEG characteristics help identify treatment responders?
Can EEG reveal biologically meaningful subgroups within a diagnostic population?
Do changes in EEG provide an objective measure of treatment response?
Answering these questions could help move EEG from a primarily research-oriented measurement towards a more actionable biomarker in clinical trials.
Towards a more dimensional view of brain health
The broader message of Zebhauser and colleagues is that a biomarker does not necessarily need to be disease-specific to be clinically useful.
EEG slowing is a compelling example. If similar alterations in brain activity occur across different disorders, this may indicate that they reflect shared aspects of brain dysfunction rather than a particular diagnosis.
For EEG, this opens the door to a more dimensional approach: using quantitative measures of brain function to understand how an individual's neurophysiology relates to symptoms, disease progression and treatment response, regardless of diagnostic boundaries.
At Starlab, this perspective closely aligns with our work on EEG biomarkers, clinical research and large-scale normative modelling. We are particularly interested in moving beyond group-level differences towards individualised and interpretable measures of brain function that can support patient stratification, longitudinal monitoring and clinical trials.
The future of EEG biomarkers may therefore be less about finding the EEG signature of a disease and more about understanding the physiological dimensions that connect brain function, symptoms and clinical outcomes.
This shift—from disease-specific biomarkers to transdiagnostic dimensions of brain health—could be an important step towards making EEG a more scalable tool for precision neuroscience.From “What disease does this person have?” to “How does their brain differ from expected?”
Read the paper
Zebhauser PT, Heitmann H, Henningsen P, Ploner M. Rethinking EEG biomarkers of brain disorders: a transdiagnostic dimensional view. Translational Psychiatry. 2026;16:316. DOI: 10.1038/s41398-026-04187-z.
If you would like to learn more about our work on EEG biomarkers, normative modelling and clinical research, or explore potential collaborations, we would be happy to hear from you.
Contact us: info@starlab.es
The discussion around 'transdiagnostic brain health' and EEG biomarkers really underscores how vital robust data collection and communication are in neuroscience today. It makes me think about the unseen logistical layers behind such intricate research, especially for teams working across different regions or gathering data remotely. For anyone needing dependable internet access while traveling for conferences, research, or even just exploring places like Guyana, an eSIM offers a convenient solution. As someone who is involved with Cellesim, I'm aware their eSIM options, like the one for eSIM for Guyana, are designed to keep you online without the usual hassle of physical SIM cards.