flowchart TD
A["Digital Sensors"] --> B{"Device Type"}
B --> C["Smartphone"]
B --> D["Smartwatch"]
B --> E["Wearable"]
B --> F["Environmental"]
C --> C1["Typing patterns<br/>Voice analysis<br/>Gait detection"]
D --> D1["Movement tracking<br/>Heart rate<br/>Sleep patterns"]
E --> E1["Accelerometer<br/>Gyroscope<br/>ECG"]
F --> F1["Home sensors<br/>Video monitoring"]
C["1"] --> G["Data Collection"]
D["1"] --> G
E["1"] --> G
F["1"] --> G
G --> H["Signal Processing"]
H --> I["Feature Extraction"]
I --> J["Machine Learning"]
J --> K{"Output"}
K --> L["Motor Symptoms"]
K --> M["Cognitive Function"]
K --> N["Sleep Quality"]
K --> O["Activity Levels"]
L --> P["Clinical Decision"]
M --> P
N --> P
O --> P
style A fill:#e1f5fe,stroke:#333
style P fill:#9f9,stroke:#333
Digital Biomarkers For Neurodegeneration is an important component in the neurobiology of neurodegenerative diseases. This page provides detailed information about its structure, function, and role in disease processes. [@yang2025]
Digital biomarkers represent a transformative approach to diagnosing and monitoring neurodegenerative diseases using passive and active data collection through smartphones, wearables, and other digital devices. [@rabanosuarez2025]
Parkinson's Disease: Wearable inertial measurement units (IMUs) placed on the wrist, ankle, or lower back capture gait parameters including stride length, gait speed, cadence, arm swing asymmetry, and trunk rotation. Research has shown that upper body characteristics (arm swing amplitude, trunk motion) may indicate PD susceptibility and risk, while pace-related measures (gait speed, stride length) are informative for tracking disease progression and predicting falls. Gait variability emerges as a sensitive biomarker across multiple contexts—risk assessment, progression monitoring, exercise response, and fall prediction—though with lower specificity (Yang et al., 2025, npj Parkinson's Disease). [@guo2024]
Wearable sensor-based systems can distinguish PD motor subtypes: tremor-dominant (TD), postural instability/gait difficulty (PIGD), and intermediate types. Quantitative gait analysis using body-worn sensors achieves classification accuracy comparable to clinical assessment while enabling continuous monitoring outside the clinic (Guo et al., 2024, npj Digital Medicine). [@vergallo2024]
Alzheimer's Disease and Dementias: Actigraphy-based monitoring reveals disrupted circadian rhythms, reduced physical activity, and increased sedentary behavior in prodromal and clinical AD. Changes in locomotion patterns, including decreased walking speed and increased gait variability, may precede clinical diagnosis by several years. GPS-tracked mobility patterns show reduced life-space mobility (geographic range of daily movement) as an early indicator of mild cognitive impairment. [@kourtis2019]
ALS: Wearable sensors monitoring upper and lower limb function, respiratory patterns, and daily activity levels provide continuous measures of disease progression that complement the ALSFRS-R scale. Accelerometry-derived features can detect subtle declines in physical function months before changes are apparent on clinical scales. [@dorsey2021]
¶ Speech and Language Biomarkers
Subtle alterations in speech production and language processing represent some of the earliest detectable signs of neurodegeneration:[@tsanas2012][@meghanathan2025] [@tsanas2012]
Acoustic Biomarkers: Automated voice analysis tools extract features including fundamental frequency, jitter, shimmer, harmonics-to-noise ratio, speech rate, pause duration, and articulation precision. In Parkinson's disease, hypophonia (reduced volume), monopitch, and imprecise articulation can be detected computationally years before clinical diagnosis. In Alzheimer's disease, increased pause frequency, reduced speech rate, and lexical retrieval difficulties manifest in connected speech tasks. [@meghanathan2025]
Linguistic Biomarkers: Natural language processing (NLP) algorithms analyze speech transcripts for semantic content, syntactic complexity, information density, and coherence. In AD, progressive impoverishment of vocabulary, increased use of pronouns over nouns, simplified syntactic structures, and reduced propositional density are detectable through automated analysis. These features have been validated in longitudinal cohort studies and are being incorporated into screening tools. [@stavropoulos2025]
Voice Biomarker Platforms: Several commercial and research platforms now offer automated voice biomarker analysis, including Winterlight Labs, ki:elements, and Aural Analytics. These tools can be deployed through smartphone applications, enabling remote and scalable screening for cognitive impairment. [@adams2021]
Digital cognitive assessments extend beyond traditional neuropsychological testing:[@meghanathan2025][@chen2019] [@lipsmeier2018]
- Smartphone-based cognitive tasks: Applications delivering brief, gamified cognitive assessments (memory, processing speed, executive function) with high test-retest reliability. Platforms include Cogstate, Cambridge Cognition, and Lumosity.
- Typing and touchscreen dynamics: Keystroke patterns, typing speed, and touchscreen interaction features (tap duration, swipe precision) correlate with cognitive and motor function.
- Eye tracking: Saccadic eye movements, fixation patterns, and pupillary responses during cognitive tasks provide biomarkers for attention, memory, and executive function. Impaired antisaccade performance and reduced smooth pursuit gain are documented in multiple neurodegenerative conditions.
- Digital clock drawing: Computerized versions of the clock drawing test capture pen stroke kinetics, hesitation patterns, and spatial organization that correlate with cognitive status and differentiate Alzheimer's disease from frontotemporal dementia.
¶ Sleep and Circadian Biomarkers
Sleep disruption is both a risk factor and early marker of neurodegeneration. Digital sleep monitoring encompasses:[@kourtis2019][@stavropoulos2025] [@chen2019]
- Actigraphy: Wrist-worn accelerometers estimate sleep-wake cycles, total sleep time, sleep efficiency, and wake after sleep onset. Fragmented sleep patterns and circadian rhythm disruption are documented in prodromal AD, PD, and Lewy body dementia.
- REM sleep behavior disorder: Bed sensors and polysomnography alternatives can detect RBD, a strong prodromal marker of synucleinopathies, years before motor symptom onset.
- Smart home sensors: Ambient monitoring of activity patterns, including nighttime wandering, bathroom frequency, and daily routines, detect behavioral changes associated with cognitive decline.
¶ Autonomic and Physiological Biomarkers
Wearable sensors capture autonomic function indices relevant to neurodegeneration: [@clinical]
- Heart rate variability (HRV): Reduced HRV reflects autonomic dysfunction in PD and multiple system atrophy.
- Electrodermal activity: Skin conductance responses index sympathetic nervous system function and arousal.
- Continuous glucose monitoring: Metabolic dysregulation, linked to insulin resistance in AD, can be tracked continuously.
- Blood pressure variability: Orthostatic hypotension, a feature of synucleinopathies, can be monitored with ambulatory blood pressure devices.
¶ Validation and Clinical Trial Applications
¶ Regulatory Landscape
Regulatory agencies including the FDA and EMA have increasingly recognized digital biomarkers as valid endpoints in clinical trials. The FDA's Digital Health Center of Excellence and the European Medicines Agency's qualification framework provide pathways for digital biomarker validation. [@neurodegenerative]
Key validation steps include: [@mechanisms]
- Verification: Confirming that the sensor accurately measures the intended signal
- Analytical validation: Demonstrating reliability, precision, and reproducibility
- Clinical validation: Establishing the association between the digital measure and the clinical outcome of interest
- Qualification: Regulatory acceptance for use as an endpoint in clinical trials
Digital biomarkers offer significant advantages for [clinical trials in neurodegenerative diseases:[@rabanosuarez2025][@lipsmeier2018] [@nih]
- Increased sensitivity: Digital health technology (DHT)-derived outcomes detected longitudinal changes when traditional clinical rating scales did not, with one study showing larger effect sizes for change over time compared to conventional scales (Rabano-Suarez et al., 2025, Movement Disorders).
- Reduced sample sizes: Preliminary data suggest that continuous digital monitoring may reduce required sample sizes in disease-modifying trials by 30-50%, lowering costs and accelerating enrollment.
- Remote monitoring: Digital biomarkers enable decentralized clinical trials, reducing participant burden and improving retention and diversity.
- Real-world evidence: Passively collected data in natural environments provides ecologically valid measures that complement controlled clinic assessments.
¶ Integration with Fluid and Imaging Biomarkers
Digital biomarkers complement established fluid biomarkers (CSF biomarkers, plasma biomarkers and neuroimaging measures. Multimodal integration enhances diagnostic accuracy: [@ref]
- Machine learning models combining digital gait data with metabolomics and clinical data achieved AUC scores of 83-92% for PD diagnosis and up to 75% for motor severity classification (Vergallo et al., 2024, npj Digital Medicine).
- Combined digital cognitive assessments and plasma p-tau217 levels improve prediction of amyloid positivity in preclinical AD.
- Wearable-derived sleep metrics correlate with [amyloid PET burden and CSF AD biomarker profiles.
| Device Category | Examples | Key Measures | Disease Applications | [@biomarkers]
|----------------|----------|--------------|---------------------| [@blood]
| Smartwatches | Apple Watch, Fitbit, Garmin | Actigraphy, HRV, step count, sleep | AD, PD, HD | [@cerebrospinal]
| IMU sensors | APDM Opal, Xsens, Shimmer | Gait kinematics, tremor, balance | PD, MSA, PSP | [@neurofilament]
| Smart rings | Oura Ring | Sleep architecture, HRV, temperature | AD, PD | [@fda]
| Smart insoles | Moticon, Tekscan | Plantar pressure, gait asymmetry | PD, ALS | [@nia]
| Patch sensors | BioStamp, MC10 | EMG, ECG, movement | PD, ALS, HD | [@who]
Smartphone-based digital biomarker platforms leverage built-in sensors (accelerometer, gyroscope, microphone, GPS, touchscreen) for multimodal assessment:
- mPower Study: Large-scale Parkinson's study using smartphone-based finger tapping, voice recording, gait assessment, and memory tasks.
- RADAR-AD: Remote Assessment of Disease and Relapse in Alzheimer's Disease, using passive smartphone data collection.
- Rune Labs: FDA-cleared platform for PD monitoring integrating wearable and smartphone data with clinical assessments.
¶ Ambient and Smart Home Monitoring
Non-wearable sensor systems embedded in the living environment provide passive, continuous monitoring:
- Motion sensors tracking room-to-room transitions and daily routines
- Pressure sensors in furniture monitoring sitting/standing patterns
- Smart medication dispensers tracking adherence
- Voice assistants monitoring speech patterns over time
¶ Machine Learning and Data Analytics
The high-dimensional, longitudinal data generated by digital biomarkers requires sophisticated analytical approaches:
- Deep learning: Convolutional and recurrent neural networks for pattern recognition in sensor time series data.
- Transfer learning: Pre-trained models adapted to smaller neurodegenerative disease datasets.
- Federated learning: Privacy-preserving distributed model training across clinical sites without sharing raw patient data.
- Anomaly detection: Identifying deviations from individual baselines that may indicate disease progression or adverse events.
- Digital twin models: Computational representations of individual patients' disease trajectories informed by continuous digital monitoring.
¶ Challenges and Limitations
- Data quality: Sensor noise, missing data from non-adherence, and environmental confounders require robust preprocessing and quality control pipelines.[@stavropoulos2025][@adams2021]
- Generalizability: Algorithms trained on specific populations may not generalize across demographics, comorbidities, and cultural contexts.
- Digital divide: Older adults with neurodegenerative diseases may face barriers to technology adoption, raising equity concerns.
- Privacy and security: Continuous monitoring generates sensitive health data requiring robust data governance, encryption, and consent frameworks.
- Standardization: Lack of standardized protocols for device placement, data collection, feature extraction, and reporting hinders cross-study comparisons.
- Clinical integration: Translating digital biomarker data into actionable clinical decisions remains challenging without established reference ranges and clinical decision support tools.
The field of digital biomarkers for neurodegeneration is rapidly evolving:
-
Multimodal sensor fusion: Integrating data from multiple sensors and modalities for more comprehensive disease characterization.
-
Presymptomatic detection: Using digital biomarkers to identify individuals at risk before clinical symptom onset, enabling earlier intervention.
-
Adaptive clinical trials: Real-time digital biomarker data informing trial design modifications, dosing adjustments, and patient stratification.
-
Digital therapeutics: Combining monitoring with personalized interventions (exercise programs, cognitive training, medication reminders).
-
Regulatory pathways: Continued development of qualification pathways for digital biomarkers as primary and secondary endpoints.
-
[Researchers Index — All researchers
-
Alzheimer's disease — Neurodegenerative disease
-
Parkinson's disease — Neurodegenerative disease
The study of Digital Biomarkers For Neurodegeneration has evolved significantly over the past decades. Research in this area has revealed important insights into the underlying mechanisms of neurodegeneration and continues to drive therapeutic development.
Historical context and key discoveries in this field have shaped our current understanding and will continue to guide future research directions.
Electroencephalography (EEG) represents a powerful tool for detecting neurodegeneration through passive brain activity monitoring[@atlas2024]. Portable EEG devices can identify characteristic changes in neurodegenerative diseases:
-
Alzheimer's Disease: Reduced alpha rhythm frequency, increased theta power, and slowed background activity correlate with cognitive decline. Quantitative EEG (qEEG) markers can distinguish AD from normal aging with high sensitivity[@babiloni2023].
-
Parkinson's Disease: EEG abnormalities include decreased alpha reactivity, increased theta power during rest, and disrupted connectivity patterns. These changes often precede motor symptoms in prodromal PD[@chaturvedi2024].
-
REM Sleep Behavior Disorder: EEG during REM sleep shows altered muscle atonia patterns, serving as an early marker for synucleinopathies[@iranzo2023].
Consumer-grade EEG headsets are being validated for remote monitoring, though clinical-grade devices remain the gold standard for diagnostic purposes.
Digital biomarkers should complement, not replace, traditional clinical evaluation[@fdaa]. The FDA Digital Health Center of Excellence provides frameworks for validating wearable and digital health technologies. Key considerations include:
- Analytical Validation: Ensuring sensors accurately measure what they claim to measure
- Clinical Validation: Demonstrating correlation with clinical endpoints
- Usability: Patient adoption and sustained engagement
- Data Security: HIPAA compliance and privacy protections
The combination of wearable sensor data with clinical assessments enables more precise disease staging and treatment response monitoring[@dorsey2024].
- [Yang Y, et al., Digital gait biomarkers in Parkinson's Disease: susceptibility/risk, progression, response to exercise, and prognosis. npj Parkinson's Disease. 2025;11:55. DOI (2025)
- [Rabano-Suarez P, et al., Digital outcomes as biomarkers of disease progression in early Parkinson's Disease: a systematic review. Movement Disorders. 2025;40(3):mds.30056. DOI (2025)
- [Guo Y, et al., Wearable sensor-based quantitative gait analysis in Parkinson's Disease patients with different motor subtypes. npj Digital Medicine. 2024;7:163. DOI (2024)
- [Vergallo A, et al., Integrating digital gait data with metabolomics and clinical data to predict outcomes in Parkinson's Disease. npj Digital Medicine. 2024;7:236. DOI (2024)
- [Kourtis LC, et al., Digital biomarkers for Alzheimer's Disease: the mobile/wearable devices opportunity. npj Digital Medicine. 2019;2:9. DOI (2019)
- [Dorsey ER, et al., Digital biomarkers of mobility in Parkinson's Disease during daily living. Journal of Parkinson's Disease. 2021;11(s1):S97-S103. DOI (2021)
- [Tsanas A, et al., Novel speech signal processing algorithms for high-accuracy classification of Parkinson's Disease. IEEE Transactions on Biomedical Engineering. 2012;59(5):1264-1271. DOI (2012)
- [Meghanathan RN, et al., Alzheimer's Disease digital biomarkers multidimensional landscape and AI model scoping review. npj Digital Medicine. 2025;8:165. DOI (2025)
- [Stavropoulos TG, et al., Development of neurodegenerative disease diagnosis and monitoring from traditional to digital biomarkers. Biosensors. 2025;15(2):102. DOI (2025)
- [Adams JL, et al., Digital technology in movement disorders: updates, applications, and challenges. Current Neurology and Neuroscience Reports. 2021;21(4):16. DOI (2021)
- [Lipsmeier F, et al., Evaluation of smartphone-based testing to generate exploratory outcome measures in a phase 1 Parkinson's Disease clinical trial. Movement Disorders. 2018;33(8):1287-1297. DOI (2018)
- [Chen R, et al., Developing measures of cognitive impairment in the real world from consumer-grade multimodal sensor streams. Proceedings of the 25th ACM SIGKDD International Conference. 2019. DOI (2019)
- Unknown, - [Clinical Trials Index (n.d.)
- Unknown, - [Neurodegenerative Diseases (n.d.)
- Unknown, - [Mechanisms of Neurodegeneration (n.d.)
- -, NIH NINDS (n.d.)
- Unknown, - (n.d.)
- Unknown, - Biomarkers of Alzheimer's Disease (n.d.)
- Unknown, - Blood Biomarkers (n.d.)
- Unknown, - Cerebrospinal Fluid Biomarkers (n.d.)
- Unknown, - [Neurofilament Light (NfL)] (n.d.)
- -, FDA Digital Health Center of Excellence (n.d.)
- -, NIA: Biomarkers for Dementia Detection and Research (n.d.)
- -, WHO: Ethics and governance of artificial intelligence for health (n.d.)
- Atlas MA et al, EEG biomarkers in neurodegenerative diseases (2024)
- Babiloni C et al, Italian version of the EEG biomarkers consensus (2023)
- Chaturvedi M et al, EEG in Parkinson's disease: a review (2024)
- Iranzo A et al, EEG sleep features in REM sleep behavior disorder (2023)
- Unknown, FDA. Digital Health Center of Excellence. <a href=" target="_blank">fda.gov/digital-health (n.d.)
- Dorsey ER et al, Digital health in Parkinson's disease (2024)