Emerging Neurobiomarkers in Neurodegenerative Disorders: From Molecular Mechanisms to Clinical Translation

1.1. Search Strategy

Databases such as PubMed, Scopus, Web of Science, and Google Scholar were searched using keywords like “neurodegenerative biomarkers,” “Alzheimer's fluid biomarkers,” “Parkinson's imaging biomarkers,” “neurofilament light chain,” “phosphorylated tau,” “digital biomarkers neurodegeneration,” and “multi-omics neurodegeneration” since January 1995 through January 2026, including peer-reviewed original research, systematic reviews, meta-analyses, and key clinical guidelines limited to the English language. Studies were selected based on relevance, recency, and contribution to the translational theme rather than through PRISMA-based screening. Conference abstracts without full publications, editorials without primary data, and non-English publications were excluded from the search.

1.2. Biomarker Readiness Classification

To consistently evaluate the translational maturity of the biomarkers in this review, each biomarker is classified according to a five-tier readiness scheme, including stages such as discovery, analytical validation, clinical validation, regulatory validation, and routine clinical use. Firstly, the discovery stage includes biomarkers identified in exploratory or preclinical studies, with limited or no validation in human cohorts. Secondly, the analytical validation stage encompasses biomarkers that have demonstrated assay reliability, reproducibility, and technical performance, but the diagnostic value of these biomarkers is not confirmed in independent cohorts. Thirdly, the clinical validation stage includes biomarkers that have demonstrated diagnostic or prognostic performance across independent, adequately powered clinical cohorts. Fourthly, the regulatory qualification stage includes biomarkers that have formally gained approval by regulatory agencies such as the Food and Drug Administration for use in clinical trial enrollment or as diagnostic aids. Lastly, the routine clinical use stage includes biomarkers that have been incorporated into standard clinical practice guidelines with established reimbursement and accessibility pathways.

2.1. Protein Misfolding and Aggregation (Tau, α-Synuclein, TDP-43)

A PET imaging study in healthy elderly individuals explored the uptake patterns of an in vivo biomarker for the accumulation of phosphorylated tau. Consistent anatomical patterns were observed with tau accumulation starting in the Medial Temporal Lobe (MTL), which would be correlated with specific Braak-like staging progressions that were measured using a computational method correlating closely with changes in cognitive measures [8]. Phosphorylated tau isoforms, particularly p-tau181 and p-tau217, accumulate as a downstream consequence of amyloid-driven tangle formation and are among the most extensively characterized fluid biomarkers of this pathway [9, 10]. Their diagnostic performance as CSF and plasma biomarkers is discussed in detail in Section 3.1 and Section 3.2. In patients with iRBD, CSF α-synuclein seeding assays demonstrated that individuals who tested positive had a higher correlation with the risk of developing PD or DLB than those who tested negative. This demonstration predicted, up to almost a decade before the clinical diagnosis of PD and DLB, in the reported individuals [11, 12]. These assays demonstrate the ability of CSF α-synuclein seeding assays to be translationally used as a biomarker in clinical settings. In studies concerning ALS and frontotemporal dementia, correlations between plasma TDP-43 and plasma in the disease group underscore an underlying pathology among these neurodegenerative diseases that requires further exploration [13, 14]. Among these biomarkers, plasma p-tau181, p-tau217, and CSF α-synuclein seeding assays have reached the clinical validation stage, while plasma TDP-43 remains at the discovery stage pending confirmation from larger, independent cohorts [9-14].

Overall, these findings underscore the importance of evaluating misfolded protein species not only as hallmarks of these disease states but also as disease-stage biomarkers.

2.2. Neuroinflammation and Glial Activation

Astrocytic reactivity in response to amyloid pathology is captured by GFAP, released into both CSF and plasma as a marker of glial activation [15]. GFAP's diagnostic performance across the AD continuum is discussed in Section 3.1 and Section 3.2. Microglial activation, as observed using translocator protein PET imaging, demonstrated increased binding in preclinical early AD [16]. This supports the implication of microglial activation being an early marker of AD. Inflammatory cytokines and chemokines in CSF and plasma are associated with neurodegenerative injury. Several inflammatory mediators, including interleukins and chemokines, have been described as elevated in CSF and plasma in association with neurodegenerative pathology, supporting a link between neuroinflammation and disease progression. Plasma and CSF GFAP have reached the clinical validation stage as early markers of astrocyte reactivity, whereas microglial activation, assessed by translocator protein PET imaging, and CSF cytokines remain at the discovery stage [15-18]. Neurodegeneration can be detected through neuroinflammation, an early biomarker that is demonstrated collectively through these findings.

2.3. Synaptic Dysfunction and Axonal Injury

Synaptic loss is observed to correlate with impairment across neurodegenerative disorders. Synaptic biomarkers such as neurogranin have been described as elevated in the CSF of patients with Alzheimer’s disease, supporting the evaluation of synaptic loss as a marker of neurodegeneration. Axonal injury releases Neurofilament Light Chain (NfL) into CSF and blood, and it is now one of the most widely measured markers of neurodegeneration across disease categories. Its diagnostic accuracy and, critically, its limitations for differential diagnosis are discussed in detail in Section 3.1 and Section 3.2.

In neurodegenerative disorders such as Huntington’s disease, elevated NfL concentrations have been reported years before clinical onset and correlate with subsequent motor and cognitive decline, supporting the notion that axonal injury is measurable during preclinical stages [19]. NfL has reached the clinical validation stage as a marker of axonal injury in PD. Collectively, these findings support NfL as a neurobiomarker of axonal injury that correlates with neurodegeneration in disease states, warranting its evaluation in clinical settings.

2.4. Mitochondrial Dysfunction and Oxidative Stress

Mitochondrial dysfunction is a fundamental cause of neurodegeneration by compromising cellular energy attained through metabolism and producing reactive oxygen species. In individuals with AD and progressive supranuclear palsy, metabolic analyses found oxidative stress that was associated with lipid metabolism and mitochondrial dysfunction, which is connected to lysine metabolism. This demonstrates that disruption of mitochondrial and metabolic pathways is a hallmark of neurodegeneration [20].

In AD, metabolites involved in energy regulation, neurotransmission, and mitochondrial function are generally found to be affected. Significant changes were reported in alanine, aspartate, and glutamate metabolisms in most brain regions, which were linked to mitochondrial dysfunction, impaired energy metabolism, and changes in mitochondrial function. This shows direct and indirect connections to neurodegeneration in AD, which warrants evaluation for consideration as potential clinical biomarkers [21].

Oxidative stress-related metabolites and CSF-based inflammatory markers were associated with biomarkers of general AD pathology [22]. This finding demonstrates that oxidative stress metabolites are related to neurodegeneration in AD. Oxidative stress biomarkers in individuals with Mild Cognitive Impairment (MCI) were also shown to be associated with neurodegeneration, ultimately leading to cognitive decline. Although a personalized approach may be warranted due to significant changes in age and sex that influence oxidative stress levels. Biomarkers for mitochondrial dysfunction and oxidative stress are currently at the discovery stage [20-22].

These findings collectively demonstrate that mitochondrial dysfunction and oxidative stress, both due to associated metabolites, caused neurodegeneration through direct and indirect processes, which underscores the need for their clinical evaluation.

2.5. Genetic and Epigenetic Influences

Genetic mutations are correlated with disease onset and are also associated with related neurodegenerative biomarker outcomes. In Huntington’s disease mutation carriers, plasma NfL concentrations have been described as elevated well before the emergence of clinical symptoms, supporting its potential as a preclinical prognostic marker [19]. Elevations in biomarkers preclinically were also reported similarly in a familial ALS cohort analysis [23].

Large-scale epigenomic analysis demonstrates that methylation-related cfDNA biomarkers can distinguish individuals with ALS from healthy individuals with high sensitivity. cfDNA biomarkers may also be able to predict ALS, but further studies are warranted to confirm this [24].

Interactions between genetic and epigenetic factors may also influence the onset and progression of disease. A genomic analysis identified specific CpG-related mutations that modify the age of onset in individuals with ALS [25]. This demonstrates wide implications, as epigenetic influences may play significant roles in the control of neurodegenerative diseases such as ALS.

Exploring beyond disease-specific mutations, changes in DNA methylation, SIRT activity, and BDNF expression are observed to decline significantly in patients with dementia or PD [26]. Their utility as potential epibiomarkers in neurodegenerative disorders is warranted, and further studies are required to determine their efficacy in the diagnosis of neurodegenerative disorders. Most genetic and epigenetic biomarkers, such as cfDNA and DNA methylation, are in the discovery stage and require further study [23, 25, 26].

Overall, these findings collectively demonstrate that considerations of genetic mutations and epigenomics need to be evaluated clinically to serve as effective biomarkers for diagnosing neurodegenerative diseases.

Protein misfolding, neuroinflammation, synaptic dysfunction, axonal injury, and mitochondrial dysfunction are interlinked mechanisms that do not occur in isolation but represent converging pathways that drive neurodegeneration from pre-clinical early stages of these diseases. Each mechanism provides distinct but measurable biomarkers. Misfolded proteins and inflammatory mediators are detectable in CSF and blood; axonal injury is assessed via NfL; metabolic dysfunction and oxidative stress are interlinked mechanisms; and epigenetic disturbances are increasingly accessible via omics platforms.

Pathophysiological Pathways and Corresponding Biomarker Sources are indicated in Fig. (1).

3.1. Cerebrospinal Fluid Biomarkers

CSF biomarkers are among the most established indicators of neurodegeneration because they directly reflect changes in the central nervous system. They provide evidence of degenerative changes, including protein aggregation, neuronal injury, axonal dysfunction, and neuroinflammation, thereby promoting early detection of neurodegenerative disorders such as Alzheimer’s Disease, Parkinson’s Disease, Amyotrophic Lateral Sclerosis, and Frontotemporal Dementia (FTD). A comparative summary of the CSF, blood-based, and peripheral fluid biomarkers discussed throughout this section is presented in Table 1.

3.2. Blood-Based Biomarkers

The diagnosis and monitoring of neurodegenerative disorders have had significant advancements lately, from the emerging studies of blood-based biomarkers. Blood-based biomarkers may substitute for CSF biomarkers and neuroimaging due to their minimal intracranial invasion.

Blood-based biomarkers are major advancements in the diagnosis and monitoring of neurodegenerative disorders. They are minimally invasive, reliable alternatives to CSF and neuroimaging. Technological improvements in ultra-sensitive immunoassays have enabled reliable detection of less-abundant neuronal and glial proteins in plasma and serum. Blood-based biomarkers, therefore, reflect important pathological processes such as protein aggregation, axonal injury, and neuroinflammation.

3.3. Emerging Peripheral Biomarkers

Peripheral biomarkers are derived from sources such as saliva and urine, as well as other peripheral immune signatures. Detection of neurodegenerative disorders through peripheral biomarkers is gaining attention, as these biomarkers may be used in preclinical or early disease stages. Peripheral biomarkers are advantageous for their offers, such as minimal invasiveness, accessibility, and potential complementarity to CSF and plasma biomarkers. They may be used as a utility for insight into systemic and peripheral manifestations of neurodegenerative disorders.

4.1. Structural MRI Patterns

MRI patterns offer hope in the early detection of neurodegenerative disorders and provide targeted treatment through the localization of brain structures. Generally, it is combined with other imaging biomarkers such as Computed Tomography (CT) and Positron Emission Tomography (PET) to provide a more detailed understanding of the underlying pathology. The development of neuroimaging strategies is important to explain the inconsistencies in the consensus on disease progression.

Generally, structural MRI is used to identify and understand patterns of regional brain atrophy from different neurodegenerative disorders. In AD, atrophy occurs in the MTL, entorhinal cortex, and parahippocampal gyrus. Structural changes in these regions are strongly associated with a decline in cognitive function and disease progression. A longitudinal study conducted by Jack et al. (1999) demonstrated that hippocampal atrophy in the brain predicted progression of the disease from MCI to AD and was closely associated with neuronal loss and degeneration of synapses reaching the routine clinical use stage [43].

The combination of MRI and PET has demonstrated that structural degeneration is strongly correlated with pathological protein accumulation. Regional neurodegeneration measured on structural MRI has been reported to follow the distribution of tau and amyloid pathology observed on PET across the Alzheimer’s disease spectrum, supporting the coupling between molecular pathology and downstream structural change. Structural MRI, therefore, may capture neurodegeneration that is correlated with protein accumulations from underlying pathology and provide a visual explanation for disease progression.

Longitudinal studies that combined MRI and tau PET imaging demonstrated that functionally connected regions are associated with tau accumulation. Tau accumulation was found to correlate with other regions with high-tau accumulation rates. Structural MRI, when used in combination with PET imaging, provides a visual representation of the underlying pathology that helps explain disease progression and staging. Further studies are required to examine combinations of such procedures through RCTs.

4.2. Functional Connectivity Alterations

Functional connectivity is the temporal dependence between the activity of different brain regions. Recent advances have enabled neuroimaging tools to investigate interactions among various brain regions. Resting-state functional MRI (rs-fMRI) can be used to assess brain integrity and functional connectivity. Alterations of functional connectivity have recently been demonstrated to be strongly correlated with recognition of protein deposition and neurodegenerative progression.

Connectome-based predictive modelling used in AD demonstrated functional connectivity patterns throughout the brain that predicted regional tau PET deposition. It was generally found in regions affected during advanced disease stages. These predictive models generalized to external datasets, including patients with both symptomatic and prodromal AD, and predicted tau-positive patients. Spatial distinctions between the positively and negatively correlated nodes were suggestive [44].

Brain connectivity has been reported to influence the regional distribution of tau accumulation in Alzheimer’s disease, with combined models that incorporate amyloid deposition and regional molecular vulnerability providing a more complete account of tau's spatial distribution than connectivity alone. Overall, resting-state fMRI-based connectivity remains at the analytical validation stage and requires further standardization. This highlights functional connectivity models as key determinants of tau deposition and progression in neurodegenerative disorders.

Functional connectivity has been described as predictive of tau progression and reflective of network-level disease vulnerability, supporting functional MRI as a candidate imaging biomarker of disease progression.

4.3. PET Imaging (Amyloid, Tau, Dopaminergic Imaging)

PET is an imaging technique that visualizes and assesses multiple functions at the molecular pathological level by tracing a compound known as a radiopharmaceutical. PET allows for noninvasive assessment of changes in levels of neurobiomarker proteins such as amyloid-beta peptides and tau proteins.

A longitudinal study conducted by Hanseeuw et al. (2019) demonstrated that elevated amyloid PET signals were associated with tau accumulation and neurodegeneration in preclinical AD. Abnormalities in amyloid PET generally precede structural atrophy and cognitive decline, underscoring its capability to be utilized for early disease detection and risk assessment [45].

Tau PET is capable of predicting cortical atrophy, along with the accuracy of severity and its distribution. A longitudinal study by La Joie et al. (2020) used ¹⁸F-flortaucipir PET imaging in individuals with AD spectrum and showed that baseline tau deposition was strongly correlated with cortical atrophy and concurrent neurodegeneration [46]. Tau PET could be used to quantify the extent of atrophy, thereby maximizing sensitivity in future studies. The use of MRI to detect disease progression is somewhat limited by inter-individual heterogeneity in atrophy patterns. Future studies are required to assess the value of baseline tau PET for prognosis with longer follow-ups.

Single-photon Emission Computed Tomography (SPECT) and PET imaging allow for the assessment of the dopaminergic pathway and represent major advancements in clinical diagnosis and the understanding of mechanistic pathways of neurodegeneration, especially in PD and related syndromes. The evaluation of PD utilizes dopaminergic PET and SPECT imaging, where a normal presynaptic SPECT scan is used as an exclusion criterion. In patients with PD, Dopamine Transporter (DAT) availability in the striatum can be detected early in the disease and declines over time. A longitudinal cohort study by Simuni et al. (2018) demonstrated that DAT binding at lower baseline levels was able to predict worsening of symptoms and functional decline in PD subjects [47]. Collectively, amyloid PET, tau PET, and SPECT imaging have reached the clinical validation stage and have transitioned to the regulatory qualification stage [45-47]. Dopaminergic imaging has previously been established as a reliable biomarker for disease severity and progression. Differentiating between PD and other movement disorders can be done using these imaging techniques.

4.4. Multimodal Imaging Approaches

Multimodal imaging has multifaceted roles in neurodegenerative disorders, such as diagnosis and evaluation of treatment. Techniques such as MRI, CT, PET, and ultrasound are collectively combined to address the limitations imposed by the others. MRI provides images of neuroanatomy, whereas PET shows changes in metabolism and molecular levels. Advanced MRI techniques such as Diffusion Tensor Imaging (DTI) and fMRI are utilized for the assessment of microstructural integrity and neural connectivity. Ultrasound is utilized for the evaluation of muscle morphology and nerve abnormalities. The collective use of these imaging techniques will demonstrate an in-depth understanding of neurodegeneration and better treatment outcomes. Integration of Imaging and fluid biomarkers is shown in Fig. (2).

Okazawa et al. (2024) demonstrated that amyloid PET and DTI showed that deposition of amyloid peptides was significantly correlated with glymphatic clearance and structural brain alterations in patients with early AD, meeting the analytical validation requirements and warranting further replication [48]. The complementary use of these imaging techniques resulted in robust findings for the identification of neurodegeneration at early disease stages.

The combination of structural MRI and amyloid PET has been shown to further improve the accuracy of disease classification compared with single-modality imaging. A multi-centre study conducted by Bao et al. (2024) showed that the integration of amyloid PET with structural MRI measures significantly improved the differentiation between AD, MCI, and healthy individuals [49]. The use of structural MRI along with PET imaging provides complementary analyses that may support differential diagnosis in clinical settings, but analytical validation is still required for standardization. Further well-designed prospective studies are required to establish the findings observed.

The value of multi-modalities can also be extrapolated to PD, where differentiation may be established compared with healthy individuals. Li et al. (2025) showed that 18F-FDG-PET/MR can effectively reflect metabolic and functional alterations in PD patients. The combination of FDG-PET metabolic imaging and functional connectivity analysis yielded results that significantly improved differentiation between PD and control subjects, reaching the clinical validation stage [50]. The metabolic network connectivity may be represented as an imaging biomarker that can reflect neurodegeneration. Future multi-centre studies with a larger sample size are required to validate these findings.

Multimodal radiomics and machine learning have further shown diagnostic performance in distinguishing PD from other neurologic diseases. Sun et al. (2023) demonstrated that multimodal PET/MRI radiomics, combined with machine learning models, successfully distinguished PD from multiple system atrophy. This combination of PET/MRI radiomics-clinical machine learning models showed promising performance in discriminating PD from MSA, indicating scope for the development of future models for clinical use and positioning them at the discovery stage [51].

PET imaging is limited by accessibility and cost due to limited regional radiotracer availability. Additionally, repeated scanning may lead to cumulative radiation exposure. Quantification may be limited due to tau PET tracers binding to non-tau structures [46].

Imaging biomarkers provide structural context for the molecular and fluid biomarkers mentioned previously. CSF and plasma biomarkers provide biochemical insights into pathology that, when paired with structural MRI, functional connectivity, and PET imaging, localize that pathology anatomically and help track its spatial progression by visualizing processes such as tau spread, amyloid deposition, and dopaminergic decline. The consistent finding across multimodal studies is that combining structural, functional, and molecular imaging modalities improves diagnostic and prognostic accuracy compared with any single modality.

5.1. Risk Genes and Pathogenic Mutations

Mutations in genes and susceptibility genes play important roles in the pathogenesis of neurodegenerative disorders. They influence protein aggregation, mitochondrial dysfunction, and neuroinflammation. Disease susceptibility is driven by monogenic mutations and polygenic risk variants.

In AD, mutations in the APP, PSEN1, and PSEN2 have been associated with familial autosomal-dominant early-onset disease that results in the production of amyloid peptides at an increased rate, which accumulate as plaques [52]. The APOEε4 allele is considered a strong risk factor for late-onset AD. Further, this contributes to impaired amyloid clearance and increased neuroinflammation [53].

In PD, gene mutations of SNCA lead to the accumulation of α-synuclein and the formation of Lewy bodies, which are representative hallmarks of PD pathology [54]. Gene mutations that are associated with mitochondrial function and impaired protein homeostasis, clearance, or α-synuclein aggregation, such as PARK2, PINK1, and DJ-1, collectively contribute to dopaminergic neuronal degeneration through oxidative stress and mitochondrial pathways [55]. Mutations in all the mentioned genes have reached the routine clinical use stage [53-55].

Genomic analyses in recent years have identified novel candidate genes that are associated with amyloid pathology, including CD180 and LY86, through a combination of transcriptomic and proteomic datasets. This suggests new molecular targets for exploration in neurodegenerative disorders [56].

5.2. Transcriptomics and microRNAs

New transcriptomic technologies allow a comprehensive analysis of changes in gene expression that are associated with neurodegenerative disorders. Wang et al. (2018) demonstrated that there are widespread alterations in genes that are associated with neuroinflammation, synaptic function, and mitochondrial metabolism [57].

RNAs such as microRNAs are being recognized as important regulators of gene expression in neurodegeneration at progressive rates. mRNA stability and translation are modulated by microRNAs, which thereby influence pathways that are linked to synaptic plasticity and neuronal availability. MicroRNA expression profiles being dysregulated have been detected in CSF and peripheral blood of patients with neurodegenerative disorders. The potential for microRNAs, therefore, warrants further investigation as minimally invasive biomarkers [58].

Transcriptomic studies in individual-level multi-omics have recently demonstrated that specific cell types, such as astrocytes and microglia, express distinct transcriptional and metabolic markers that are correlated with the progression of AD. The importance of specific cell-type gene regulation is therefore emphasized for consideration in neurodegeneration [59]. Transcriptomics and microRNAs are currently at the discovery stage [58, 59].

5.3. Epigenetic Signatures

Gene expression that are regulated by epigenetic mechanisms without changing the underlying DNA sequence are crucial to the pathogenesis of neurodegenerative disorders. DNA methylation and histone modifications are major epigenetic modifications in these disorders.

Significant epigenetic alterations in brain tissue and peripheral blood of patients have been expressed through studies of DNA methylation. Epigenetic alterations can affect the expression of genes that are involved in inflammation and neuronal availability. Further evaluation of these changes is required for assessment in neurodegeneration [60].

In PD, alterations in epigenetics have been associated with different methylation patterns in genes that are correlated with α-synuclein expression and mitochondrial function [61]. Increased α-synuclein expression and disease susceptibility have been correlated with hypomethylation of the SNCA promoter region. Currently, SNCA promoter hypomethylation is at the discovery stage. Overall, epigenetic alterations require further studies for evaluating their use as potential biomarkers for early disease detection and monitoring.

5.4. Proteomics and Metabolomics

Proteomics and metabolomics are crucial for providing insights into molecular alterations that reflect functional changes in neurodegenerative disorders.

Proteins that are associated with complement activation and neuroinflammation are correlated with amyloid accumulation and neurodegeneration. Wingo et al. (2019) demonstrated that circulating protein signatures are associated with AD pathology and subsequent cognitive decline. Proteins involved in mitochondrial activities or synaptic functions had increased abundance among subjects with stable cognition, with or without amyloid plaques or neurofibrillary tangles [62].

Emerging plasma proteomic panels, combined with machine learning, have been shown to differentiate dementia etiologies, including Alzheimer’s disease, frontotemporal lobar degeneration, and Lewy body disease, which is particularly relevant in clinical settings where these disorders present with overlapping symptoms [63].

Current alterations of metabolic pathways are associated with mitochondrial metabolism, lipid biosynthesis, amino acid metabolism, neurotransmitter metabolism, and hormonal metabolism in neurodegenerative disorders. Trushina et al. (2013) demonstrated that in MCI and AD, pathways related to mitochondrial function, lipid biosynthesis, amino acid metabolism, neurotransmitter metabolism, and hormone metabolism were altered canonically [64]. Circulating proteomic signatures and metabolomic panels are in the analytical validation stage and require further standardization. These changes can be reflected as indicators for early disease progression.

5.5. Systems Biology and Multi-Omics Integration

Systems biology and multi-omics datasets are combined to provide a comprehensive understanding of the mechanisms of neurodegenerative disorders. Integration of multi-omics enables the investigation of interactions among genes, transcripts, proteins, metabolites, and other substrates, and the identification of disease-specific molecular networks.

Integrated transcriptomic, proteomic, and neuropathological analyses have identified molecular signatures associated with amyloid pathology, neuroinflammation, microglial metabolism, and synaptic dysfunction in Alzheimer’s disease, positioning microglial biology as a candidate mechanistic driver and biomarker source at the discovery stage.

Multi-omics studies similarly integrated genetic and proteomic datasets, leading to the identification of biomarkers that are shared across multiple neurodegenerative disorders [62]. This suggests that altered molecular pathways, such as mitochondrial metabolism, immune activation, synaptic biosynthesis, astrocyte metabolism, and biosynthesis, may overlap.

Multi-omics datasets are increasingly being used by recent approaches such as machine learning and artificial intelligence for the identification of biomarker patterns, which can be used to detect early disease onset and enable progression suspension. Integrated approaches towards precision medicine and personalized therapeutic strategies in neurodegenerative disorders can be crucial for progression.

A comparative summary of omics-based biomarkers and their clinical readiness is presented in Table 3.

Despite promising outcomes, multi-omics integration is limited by inconsistent data standardization across platforms and substantial computational demands. These models are generally trained on small cohorts, raising overfitting concerns. There are limited panels that have undergone prospective validation in independent populations.

Multi-omics and genetic biomarkers add additional layers of mechanistic and risk stratification. While CSF, plasma, and imaging biomarkers provide comprehensive insights into the consequences of neurodegeneration, genetic and multi-omics data can offer insight into individual susceptibility, disease subtype, and the molecular drivers of pathology. However, the clinical readiness of these omics categories varies considerably, while proteomics and select genomic markers (APOE, SNCA) are approaching clinical-translational use, transcriptomics, epigenomics, and multi-omics integration approaches remain largely at an early emerging stage. Multi-omics provide value through precision stratification by identifying molecular subtypes within clinically similar patient populations that may respond differently to therapeutic interventions, rather than by replacing fluid and imaging biomarkers.

6.1. EEG and Electrophysiological Markers

EEG is a method for recording the brain's spontaneous electrical activity. It is a non-invasive method for measuring neuronal activity in the brain and has been widely explored as a biomarker in neurodegenerative disorders. In AD and PD, alterations in EEG spectral power and network connectivity have been observed. EEG recordings in AD frequently demonstrate abnormal activity and network dysfunction, is evidenced by increased delta and theta power and reduced alpha and beta activity. This is associated with synaptic dysfunction and neuronal loss [65].

In PD, electrophysiological biomarkers are being concurrently investigated. Decline in motor function has been linked to abnormalities in beta oscillations within cortico-basal ganglia circuits. Neumann et al. (2016) demonstrated that increased subthalamic 8- 35 Hz band power was strongly linked to clinical signs in PD patients. Low-beta frequency bands show a significant correlation with levodopa-induced reduction compared with high-frequency bands, suggesting variable functional roles for individual frequency ranges. Subthalamic oscillatory activity correlates with motor impairment, indicating that electrophysiological markers may provide informative insights into motor decline in neurodegenerative disorders [66]. EEG and electrophysiological markers are at the discovery stage.

High-density EEG and source localisation techniques have recently enabled the identification of disease-specific network disruptions by improving the resolution of electrophysiological analyses. Electrophysiological biomarkers, especially EEG-based, may complement imaging and molecular biomarkers for the detection and monitoring of diseases at early stages.

6.2. Wearable Technologies and Remote Monitoring

Monitoring of physiological and behavioral parameters has been enabled by wearable technologies in patients with neurodegenerative disorders. Detailed information on movement patterns and sleep behavior can be captured through devices equipped with accelerometers and physiological sensors. Motor symptoms in PD, such as tremor and bradykinesia, can be evaluated using wearable sensors. Del Din et al. (2019) demonstrated that motor activity obtained from devices worn on wrists was able to detect subtle motor impairments and predict disease progression in PD patients [67].

AD and MCI can also be monitored through remote monitoring technologies. Sleep disturbances and daily activity patterns being monitored continuously have been shown to correlate with cognitive decline and neurodegenerative pathology. Clinical symptoms of dementia preceding diagnosis can be detected through passive digital monitoring that is integrated into homes and observes daily routines. These wearable technologies require further validation and are currently in the analytical validation stage.

Advantages of wearable technologies, including objective quantification of symptoms and longitudinal monitoring, may allow for a reduction in reliance on clinical visits. Digital biomarkers collectively may prove to have utility that is particularly valuable for clinical trials with large populations and longitudinal disease monitoring.

6.3. Speech, Gait, and Cognitive Digital Phenotyping

Behavioral and physiological data generated through interactions with digital devices can be collected for analysis, which is referred to as digital phenotyping. Speech patterns and gait characteristics can serve as indicators of disease progression for assessment in neurodegenerative disorders.

Recently, speech analysis has emerged as a potential biomarker for neurodegenerative disorders. Symptoms such as reduced fluency and impaired articulation are generally observed as characteristic features in patients with AD and PD. The detection of cognitive and motor impairments in the early stages of the disease states is possible through computational analyses of acoustic features and speech timings. Lexical and syntactic features from narrative speech can distinguish individuals with AD from healthy individuals [68]. Early detection and disease monitoring through potential biomarkers, such as speech-based phenotyping, should be further evaluated.

Digital biomarkers such as gait analysis are also important. Motor coordination and balance generally lead to the development of measurable gait abnormalities in neurodegenerative disorders. Wearable sensors and motion capture systems identified specific gait signatures associated with PD, including reduced stride length and impaired postural stability [69]. Digital biomarkers that detect speech patterns and cognitive phenotyping require further real-world validation and are therefore at the analytical validation stage.

Cognitive tests based on smartphones have been shown to have promising results in the detection of early cognitive decline. Assessment of reaction time and memory performance, along with other cognitive assessments, may serve as objective indicators of cognitive impairment, enabling early detection of neurodegenerative disorders.

6.4. Artificial Intelligence in Biomarker Discovery

Artificial intelligence and machine learning approaches have recently emerged for the identification of complex biomarker patterns in neurodegenerative disorders. These approaches may be capable of combining multimodal datasets like neuroimaging and electrophysiological records for the identification of predictive signatures of neurodegeneration and disease progression.

Early AD pathology can be detected with high accuracy by analyzing neuroimaging data using deep learning algorithms. Neural networks that were trained on brain imaging datasets were capable of identifying individuals with preclinical AD multiple years before the onset of symptoms [70]. The analysis of motor activity decline in patients with PD for the assessment of disease severity can be performed through wearable sensors with machine learning models applied to them.

Biomarker discovery through artificial intelligence also enables the identification of previously unrecognized relationships between physiological signals and disease mechanisms. By analyzing data obtained through digital health technologies, machine learning models may aid in the discovery of subtle patterns that were undetectable through models based on conventional statistical approaches. Artificial intelligence and deep learning algorithms require further validation in realistic scenarios and are therefore classified as being at the analytical validation stage.

As access to data and computational power increases, artificial intelligence is expected to play a crucial role in the development of biomarkers for neurodegenerative disorders in the near future, allowing early-stage diagnoses of these diseases.

Data privacy concerns and regulatory pathways raise concerns for digital biomarkers. Interpretability may be subjective in artificial intelligence-based algorithms. Additionally, device heterogeneity limits cross-study comparability, and variable patient compliance introduces missing data.

Digital and neurophysiological biomarkers provide a niche role within the previously proposed multimodal framework. Fluid, imaging, and omics biomarkers capture a single time-point snapshot of underlying pathology, often obtained through invasive or resource-intensive procedures. Digital biomarkers obtained from EEG, wearable sensors, speech, and gait enable continuous, valid monitoring of motor, cognitive, and behavioral function in real-world settings. This complementary insight makes digital biomarkers well suited for longitudinal disease monitoring, remote clinical trial endpoints, and early detection of subtle functional decline between clinical visits that molecular or imaging biomarkers may not capture. Analysis is provided through artificial intelligence to combine multimodal datasets for detecting signatures of neurodegeneration that can be difficult to detect through conventional statistical approaches.

7.1. Diagnostic and Prognostic Applications

Neurobiomarkers have gained increasing importance for their proficiency in diagnostic accuracy and specificity across neurodegenerative disorders, particularly at prodromal stages. Conventionally, diagnosis substantially relies on clinical symptoms, which appear at later stages of the disease when significant neurodegeneration has occurred. Duly, diagnostic approaches can overcome these challenges through the detection of these underlying pathologies at the preclinical or prodromal stages of disease.

As previously established, blood-based biomarkers can accurately identify pathological hallmarks of neurodegeneration. Plasma phosphorylated tau (p-tau217) produced high diagnostic accuracy for the identification of amyloid and tau pathology that directly correlated with AD. p-tau217 was demonstrated to be the highest predictor of cognitive decline in the group with Alzheimer’s clinical syndrome [71].

NfL can predict neuronal injury and disease progression across multiple neurodegenerative disorders such as MS, AD, FTD, ALS, atypical parkinsonian disorders, and traumatic brain injury. NfL should be complemented by other disease-specific biomarkers and brain imaging findings in clinical settings to realize its potential as a biomarker.

As discussed in Section 3.2, plasma p-tau217 has demonstrated the strongest cross-sectional diagnostic accuracy among current blood-based biomarkers for identifying AD pathology (AUC up to 0.96), while plasma p-tau181 has independently demonstrated both strong cross-sectional diagnostic accuracy and superior longitudinal prognostic value compared with T-tau, Aβ42/Aβ40, and NfL [9, 10]. This differential is clinically consequential: while NfL reliably signals that neurodegeneration is occurring, Section 3.2 shows it lacks the specificity needed to distinguish AD from other neurodegenerative diseases or to independently predict AD-specific risk, whereas p-tau species and CSF-based combinatorial algorithms (Section 3.1) can meaningfully reduce reliance on lumbar puncture and PET imaging for diagnostic confirmation [9, 10].

7.2. Biomarkers in Clinical Trials

Biomarkers play a crucial role in the design and execution of clinical trials in neurodegenerative disorders. Selection of participants, stratification, pharmacodynamic monitoring, and outcome measures are continuous use cases for biomarkers.

Biomarkers are often used to confirm the presence of amyloid or tau pathology for the selection of participants in AD-based clinical trials. This ensures that appropriate patient populations are selected for testing investigational therapies targeting specific pathological mechanisms. P-tau217 reduces the need for Aβ-PET imaging or CSF biomarkers in most individuals and minimizes the need for tau PET to identify individuals who will benefit from antiamyloid immunotherapies, improving trial efficiency and reducing costs [71].

NfL has been widely used as an indicator of neuroaxonal damage and as an endpoint for treatment in ALS-based clinical trials. Correspondingly, amyloid PET and tau PET are increasingly used as surrogate endpoints in trials evaluating therapies for disease modification before measurable clinical improvement occurs [45, 72].

Many clinical trials require biomarker evidence, such as Aβ pathology, for inclusion in AD trials. However, there may be interindividual variations due to different brain pathologies; therefore, it is important to determine the cause of symptoms for each individual. Subgroups that are presumably outlined by certain biomarkers may respond more to therapies.

Longitudinal studies focusing on these biomarkers that span from pre-symptomatic or prodromal stages of the disease are crucial for understanding how pathologies may overlap and manifest observationally, as well as their correlation with clinical symptoms and their differences.

7.3. Standardization and Regulatory Pathways

The standardization and validation of these biomarkers are necessary for their implementation in clinical practice. Changes or differences in laboratory techniques, sample collection methods, assay platforms, and imaging protocols may result in substantial variability across multiple studies. Inconsistencies in biomarker measurements can be produced due to the aforementioned variations, which may limit the reproducibility and clinical utility of these biomarker measurements.

International regulatory agencies such as the U.S. Food and Drug Administration and the European Medicines Agency have developed biomarker validation frameworks to assess the analytical validity and clinical validity of these biomarkers before they can be used in drug development or decision-making.

The importance of analytical validation in biomarker assays requires emphasis. Partial parallelism was generally inconsistent in its observation and reporting for most biomarker assays for neurodegeneration. Generalizability is limited due to the narrow range of dilution and inconsistent methodologies. Dilution protocols and adherence to attested analytical validation guidelines are of significant importance [73].

Cross-cohort validation of a plasma p-tau217 assay has been described as showing consistent identification of amyloid pathology across independent community-based cohorts, which is supportive of clinical implementation [74]. Standardization efforts, as previously demonstrated, need to be integrated into the validation of these biomarkers used for aid clinical diagnosis.

7.4. Cost-Effectiveness and Global Accessibility

Significant progress has been made in biomarker research. Despite this, the widespread implementation of biomarker technologies remains hindered by cost and global accessibility. Specialized equipment and trained personnel are required for techniques like PET imaging and Multimodal imaging, which increases their difficulty for implementation in healthcare settings.

Alternatives such as blood-based biomarkers are promising due to their affordability, scalability, minimal invasiveness, and diagnostic accuracy. Large diagnostic studies have shown that combined plasma Aβ ratio and phosphorylated tau assays can accurately detect Alzheimer’s disease pathology in both primary care and specialist settings, supporting their scalability outside dedicated memory clinics.

Recent developments have also highlighted the importance of blood-based biomarker tests for the transformation of clinical practice. p-tau217 has demonstrated high diagnostic accuracy for distinguishing Alzheimer’s disease from other neurodegenerative disorders in multicenter cohort studies [10]. This highlights important underlying frameworks that may aid with quality of life and patient-centred financial outcomes.

Digital biomarkers derived from wearable devices and remote monitoring technologies may improve accessibility through continuous monitoring of motor and cognitive decline outside clinical settings. Wearable sensors have demonstrated the ability to detect changes in gait and motor activity in patients with PD, providing valuable insights for disease monitoring and treatment [67]. Diagnostics based on biomarkers, including these technologies, have the potential to improve disease monitoring in healthcare settings with limited resources.

8.1. Biological Heterogeneity

Biological heterogeneity is a significant obstacle to the development of biomarkers for neurodegenerative disorders. Genetic susceptibility, complex neuropathological mechanisms, clinical presentations, and disease progression are confounding factors that need consideration in patients diagnosed with the same neurodegenerative disorder. The generalizability of candidate biomarkers may be limited by variability found in small discovery cohorts.

Large-scale multi-omics investigations have recently demonstrated that multiple molecular subtypes exist within neurodegenerative disorders, each characterized by distinct biological pathways. Neuroinflammation and synaptic dysfunction have been described as molecular signatures that are both disease-specific and shared across several neurodegenerative disorders, as suggested by cross-disease proteomic analyses. These findings suggest that underlying biological processes may develop seemingly similar clinical patterns of disorders.

Heterogeneity is evident, particularly in PD, where genetic mutations and neuropathological differences produce diverse phenotypes that may vary in motor symptoms, non-motor symptoms, cognitive impairment, and other clinical manifestations [75]. Neuropathological studies of Alzheimer’s disease have described distinct subtypes based on the regional distribution of tau pathology, differing in age at onset, genotype, and clinical trajectory [76].

Important challenges in biomarker development, such as heterogeneity, can be problematic, as diagnostic biomarkers must accurately detect disease pathology across diverse patient populations. Molecular subtypes are increasingly used for the classification of patients for stratification approaches, rather than relying solely on traditional clinical diagnoses. Integration of genomic, proteomic, metabolic, and imaging biomarkers may enable precise characterisation of disease subgroups and improve biomarker performance.

8.2. Validation and Reproducibility

The limited reproducibility of many candidate biomarkers may cause critical barriers to clinical translation. Numerous biomarkers demonstrate diagnostic accuracy in preliminary studies but may fail to replicate this when evaluated in independent cohorts or in practical clinical populations. Study design, sample collection, protocol adherence, and analytical methods may significantly influence biomarker measurements and contribute to inconsistent findings.

Factors such as standard operating procedures, sampling time points and methodologies, sample handling, and storage conditions have been shown to affect the stability and quantification of fluid biomarkers such as p-tau and NfL. Multicenter investigations have therefore emphasized the importance of adherence to standardised sample-collection protocols and randomised sample analysis to improve reproducibility across laboratories.

Computational approaches may also cause reproducibility challenges when used for biomarker discovery. Machine learning models derived from neuroimaging or multi-omics datasets using a single dataset are not replicated using other datasets. Dataset integration can result in a decrease in classification accuracy but also enhance the reproducibility of potential biomarkers. The average lack of reproducibility is significantly reduced by dataset integration, which is particularly important [77].

Addressing these limitations, international initiatives have established large collaborative datasets designed to validate biomarkers across diverse populations. Integration of datasets like imaging data, clinical phenotypes and genotypes, molecular biomarkers, and variance can produce robust resources for evaluating candidate biomarkers and improving their reproducibility.

8.3. Ethical Considerations of Early Diagnosis

Early detection of neurodegenerative disorders during preclinical stages, often multiple years before the onset of clinical symptoms, was made possible by advancements in biomarker technology. Recruitment of individuals into clinical trials that are generally high-risk for the detection of diseases raises important ethical and social concerns that need to be addressed.

The conceptualization of multiple neurodegenerative disorders has been fundamentally changed due to diagnostic frameworks based on biomarkers. The National Institute of Aging and Alzheimer’s Association (NIA-AA) has proposed a research framework that introduced a biological construct to define AD based on biomarkers that are specific to hallmark AD proteinopathies, such as amyloid deposition and tau pathology, rather than purely clinical criteria [78]. Similarly, recent proposals for redefining synucleinopathies emphasize that the accumulation of α-synuclein pathology should be the focus for considering synuclein diseases, even if they meet the clinical definitions of PD or DLB.

Complex ethical questions may still be raised despite the improvement in diagnostic precision. Individuals who may be positive for specific biomarkers may remain asymptomatic for several years or may never develop the indicated clinical disease. Expressing the existence of preclinical disease risk may result in psychological distress or social stigmatisation.

The potential misuse of biomarker information in employment or insurance settings may also raise additional concerns. Ethical guidelines should emphasize the importance of informed consent and genetic counselling. The benefits of early diagnosis need to be weighed against potential psychological and societal consequences led by mental disturbances, which may pose a central challenge to the implementation of diagnosis based on biomarkers.

8.4. Toward Precision and Preventive Neurology

The development of strategies for precision and preventive approaches to neurodegenerative disorders is guided by future advances in biomarker research. Precision neurology integrates molecular, imaging, digital, and fluid biomarkers with clinical data to identify disease subtypes and predict progression for individual patients.

Genomics and proteomics across diverse populations have been integrated to generate extensive biomarker datasets through global collaboration. Large datasets based on proteomics have been designed to identify biomarkers and therapeutic targets across multiple neurodegenerative disorders [79]. Large-scale resources, therefore, provide opportunities to investigate disease mechanisms and further develop accurate diagnostic tools.

Artificial intelligence and machine learning approaches are expected to accelerate biomarker discovery by enabling the integration of complex multimodal datasets. Combining data from neuroimaging, fluid biomarkers, genomics, proteomics, and digital monitoring technologies may allow clinicians to detect neurodegenerative disorders earlier and identify patients at higher risk of disease progression.

Finally, multimodal biomarkers combined with precision medicine may shift clinical practice from reactive treatment based on symptoms of the associated neurodegenerative disorder to preventive approaches that may delay or prevent disease onset. Achieving this, however, may require efforts for validate biomarkers and ensure ethical implementation on an ongoing basis.

8.5. Limitations of This Review

This review has several limitations that should be considered when interpreting its conclusions. As a narrative rather than a systematic review, study selection was guided by relevance and translational significance rather than by a predefined, reproducible screening protocol, introducing potential for selection bias in which studies were included and emphasized. The search was restricted to English-language publications, meaning relevant findings published in other languages may have been omitted, particularly from research communities where biomarker studies are conducted but not always translated into English-language journals. Coverage across neurodegenerative disorders is also uneven: because Alzheimer's disease has generated a substantially larger body of biomarker literature than Parkinson's disease, amyotrophic lateral sclerosis, frontotemporal dementia, or Huntington's disease, AD-related biomarkers are more extensively represented and more thoroughly characterized throughout this review than those for other conditions, which may understate the relative maturity of biomarker research in less-studied disorders. Furthermore, this review does not include a formal meta-analysis or forest-plot synthesis of quantitative findings; where diagnostic performance metrics are reported, they are drawn directly from individual studies rather than pooled across the literature, and readers should be cautious in generalizing single-study estimates, particularly where cohort size, population characteristics, or assay platform differ. Finally, given the pace of publication in this field, coverage of the most recent literature, including a number of 2025 publications, may be incomplete due to the lag between publication and indexing, and biomarkers or findings published after the literature search was conducted are not reflected here.