What Are Digital Biomarkers?
- Koneksa Health
- Apr 15
- 8 min read
Healthcare systems worldwide are increasingly shifting toward data-driven models integrated with digital health technologies (DHTs) and artificial intelligence (AI). This shift is reflected in the rapid growth of the global digital biomarkers market, which is valued at US$7.33 billion in 2026 and projected to increase at a compound annual growth rate (CAGR) of 13.80% to US$18.1 billion by 2033. Increasing investments in digital therapeutics, growing awareness about preventive healthcare, and advancements with smart devices are collectively contributing to the market’s expansion.
The term “biomarker” is certainly not a new one, considering it has existed since the 1980s. As modern healthcare places a growing emphasis on precision medicine and patient-centric care models, the demand for reliable and scalable digital biomarker solutions continues to rise.
This article aims to provide an in-depth understanding of digital biomarkers and will offer answers to the following key questions:
What are digital biomarkers and how do they work?
What are examples of digital biomarkers in clinical trials?
How are digital biomarkers used in clinical research?
How do digital biomarkers differ from wearables and digital trial endpoints?
What are the advantages and limitations?
Key Definitions
Traditional Biomarker
The joint United States (US) Food and Drug Administration (FDA) and National Health Institute (NIH) working group first defined a biomarker as “a marker of a biological process which can be normal, pathological, or as a response to therapy or exposure.” Biomarkers can be captured using objective measurement methods, such as laboratory instrumentation, in vivo imaging techniques, or electrophysiological methods. Several types exist, including molecular, histologic, radiographic, digital, or physiologic biomarkers.
Digital Biomarker
Digital biomarkers refer to the measurable and quantifiable physiological, behavioral, and environmental parameters collected for an individual through innovative DHTs, including wearables, smart devices, and medical sensors. These biomarkers can offer key health insights that indicate health status, disease state, or treatment response.
How Do Digital Biomarkers Work?
In today’s modern, digitalized world, human health parameters can be quantified and collected through innovative DHTs, including wearables, smart devices, and medical sensors. For example, smartwatches or fitness trackers can measure an individual’s heart rate, physical activity, and resting patterns, while sensors in smart devices can recognize subtle movements and record speech. These measurements all represent well-known modulators of personal health or correlate strongly with it and can potentially be used to support optimal healthcare and precision medicine for patients.
The Importance of Thoughtful Measurement Design
It is important to emphasize that digital biomarkers are a strategic type of measurement that can be used to influence clinical decisions, not simply “more data.” Health insights captured using biomarkers are only useful when the measurement is strategically designed, aligning both biologically and with the clinical intent. For digital biomarkers to be applicable in a healthcare context, they must be developed from patient-level data and validated to ensure their reliability. Therefore, the accuracy and clarity of signals captured by digital biomarkers, determined by a robust measurement strategy, are substantially more crucial than sheer data volume.
How are Digital Biomarkers Validated?
A reliable, clinically meaningful digital biomarker must be proven accurate and meaningful through both analytical and clinical validation.
Analytical validation determines whether the digital technology reliably measures its target outcome with accuracy, precision, and consistency of data, ensuring replicability, particularly against gold standards, and a clean signal.
Clinical validation demonstrates whether the digital biomarker provides clinically meaningful insights into the health outcome of interest in people.
Both types of validation must be met in order for a digital biomarker to be considered “fit-for-purpose,” confirming that it produces clear, stable signals that are consistently representative of real-world patient outcomes.
Types of Digital Biomarkers
Several types of digital biomarkers are available, each with key benefits and limitations, as summarized in Table 1.
Table 1: Overview of key types of digital biomarkers
Type | Description | Pros | Cons |
Active | Requires a participant action (e.g., guided walk test on a phone) | Greater control, clearer signal | Higher patient burden, adherence issues, site training or coaching needs |
Passive | Captured continuously in the background (e.g., activity, sleep) | Lower patient burden, offers longitudinal patterns | Higher noise, context may be confounding, prone to missing data |
Continuous | Streaming or high-frequency measurement that captures patterns over time | Real-time data, increased sensitivity, and reduced manual effort | Higher noise, context may be confounding, higher data management burden, may require specialized equipment |
Point-in-time | Scheduled, episodic measurements taken at specific moments | Easier to standardize, may be simpler to interpret | Intermittent data, labor-intensive, weaker sensitivity for fluctuations and transient events between assessments |
Direct | Closer to the biological or clinical outcome intended to be measured | Clearer clinical meaning, may offer greater insights for decision making | May be difficult to measure outside the clinic, may require more standardization |
Proxy | Correlated signal used as an indirect indicator of the target outcome | Can be feasible at scale, useful when direct measurement is impractical | Must be interpreted with caution, particularly if the evidence base for clinical meaning is weak, higher susceptibility to confounding |
How do Digital Biomarkers Differ from Wearables, Digital Endpoints, or ePRO?
Although these concepts can complement each other to improve clinical research results, digital biomarkers are not the same as wearables or smart technology, digital trial endpoints, or electronic patient-reported outcomes (ePROs). Digital biomarkers are a type of measurement strategy, whereas wearables or smart technologies are tools. Data collected by these tools can only be considered digital biomarkers once they are analytically and clinically validated, representing a clear clinical outcome. Similarly, in clinical trials, a digital endpoint is the outcome for which data from digital biomarker measurements are used to test the study’s hypothesis. Lastly, ePROs capture self-reported patient health and experiences directly via digital devices over the duration of a trial, whereas digital biomarker measurements are sensor-derived.
The Role and Key Advantages of Digital Biomarkers in Clinical Trials
The primary advantage of digital biomarkers is that they can address several limitations of traditional clinical trial outcome assessments. These assessments are typically point-in-time measurements that represent only snapshots in time, which make the resulting data prone to high variability, depending on patient motivation at the exact time of assessment, and do not necessarily reflect what is happening to patients in their natural environment. Additionally, clinical trial assessments are often conducted in-clinic, posing a key barrier for patients that may contribute to lower participation and adherence rates.
Designing and developing digital biomarkers offers an opportunity to address these challenges by providing more frequent or semicontinuous real-world monitoring of patients’ functioning in their work and home settings with DHTs. Outcomes measured by these tools may constitute digital biomarkers or electronic clinical outcome assessments (eCOA), which can be deployed with greater convenience for patients, while complementing traditional in-clinic assessments with novel datasets. Other benefits of using digital biomarkers in clinical research include allowing for objective measurements of signs and symptoms, earlier detection of change than with periodic in-clinic assessments, and improved understanding of disease features and subtypes.
Although the use of digital biomarkers offers valuable insights for clinicians and researchers, it is important to pair these measurement strategies with data collected from clinical assessments, imaging or lab markers, and PROs. Triangulation with several data collection approaches improves interpretability and reduces the risk of false positives, resulting in a clearer signal. In turn, this enables potentially faster clinical trials with smaller sample sizes, efficient decision-making, and better allocation of Research and Development resources for sponsors.
For more information about how digital biomarkers can help expedite the drug development process, visit our page here.
Applications and Use Cases of Digital Biomarkers in Clinical Research
The interactions between individuals and innovative DHTs produce numerous types of measurable data that can be useful in clinical trials. For example, Figure 1 represents examples of key domains relevant to digital measurement programs in central nervous system (CNS) research, with commonly studied digital measures including gait, balance, tremor, activity, sleep, cognition, and speech. However, as described earlier, the specific modalities used to collect digital health data should only be viewed as tools. The real value of these biomarkers is determined by whether the measurement represents a clinically meaningful signal when evaluating trial outcomes. For specific use-case driven examples of digital biomarkers in clinical research settings, see Table 2.
Interested in learning how digital endpoints are transforming CNS clinical trials? Read our white paper here.
Figure 1: Digital measurement components spanning domains relevant to CNS clinical research
Across CNS programs, commonly studied digital measures include:
Gait and balance
(speed, variability, freezing)
Postural tremor
(amplitude, frequency)
Upper limb function (pronation/supination, finger tapping)
Lower limb function
(mobility, balance)
Sleep
(time asleep, number of awakenings)
Cognition
(memory, executive function)
Speech
(phonation, prosody, articulation)
Table 2: Examples digital biomarkers used across various therapeutic areas of clinical research
Therapeutic Area | Parameters Requiring DBM | Description of DB |
Diabetes Care | Blood glucose levels | CGM systems track glucose levels in real-time using patches (skin-implanted sensors). |
Medication adherence | DBs track medication adherence by monitoring when and if prescribed medications are taken through smart pill containers, medication reminder apps or insulin pen caps. | |
Cardiovascular Diseases | Physical activity and sedentary behavior | Accelerometers and gyroscope sensors integrated into wearable devices track an individual’s physical activity levels and sedentary behavior patterns. |
Blood pressure | Ambulatory blood pressure monitoring through wearable devices provides continuous blood pressure measurements throughout the day. | |
COPD | Oxygen saturation | Wearable pulse oximeters continuously measure oxygen saturation, alerting individuals and healthcare providers to any fluctuations or abnormalities. |
Cough | Smartphone applications and wearable devices equipped with microphones can provide valuable data on cough frequency, intensity, and pattern. | |
Cancers | Radiomics and pathomics application | AI algorithms utilize data from digital images from radiology and pathology to diagnose cancer at earlier stages |
Cancer-related symptoms | Smartphone applications and wearable devices can track and report cancer-related symptoms such as pain, nausea, fatigue, and depression. |
Abbreviations: CGM = continuous glucose monitoring; COPD = chronic obstructive pulmonary disease; DB = digital biomarker; DMB = digital biomarker monitoring; ECG = electrocardiogram.
Key Challenges of Digital Biomarkers
Although the potential of digital biomarkers in the context of clinical research and disease management is promising, several challenges need to be addressed. For example, digital biomarkers collect sensitive patient health information that must be protected. Given that their use in clinical research is still evolving, developing additional measures to ensure standardized informed consent, data anonymization, and protection of ownership rights is ongoing.
The sheer volume of data that continuous digital health monitoring produces also presents challenges with validation, which are further complicated by the lack of common data standards. Further, the use of machine learning models for digital biomarkers can pose inherent biases from the training dataset, limiting clinical relevance, emphasizing the importance of building models on representative samples of the target disease population. Other potential challenges include noise and variability in datasets, as well as navigating missing data.
The Future of Digital Biomarkers in Clinical Research
Despite these challenges, the future of digital biomarkers in the management of non-communicable chronic diseases is promising, with ongoing research collaborations between technology and healthcare sectors driving advancements in AI and machine learning algorithms to enhance the accuracy, reliability, and clinical utility of digital biomarkers. Therefore, digital biomarkers are well-poised to provide new opportunities for patient-centered clinical research, as well as deeper insights with remote health monitoring.
However, researchers must consider a purposeful design, conduct rigorous validation, and complement insights from digital biomarkers with other traditional clinical assessment methods. When successfully implemented, digital biomarkers can provide accurate, more continuous, real-world insights that improve trial efficiency and clinically relevant decision-making.
Digital biomarkers for earlier signals, better compliance, and richer clinical insights
Koneksa is a measurement-science partner supporting clinical development through biologically grounded measurement strategy and high-reliability multi-measure data collection for digital biomarkers across multiple therapeutic areas, including oncology and Parkinson's disease. By reducing variability and clarifying treatment effects, Koneksa enables earlier insight, stronger evidence, and more confident decisions across trials.
Visit our website to learn more about our advanced modeling and analysis services and explore how to design digital biomarkers that reveal clearer signals in clinical trials.
If you’re evaluating digital biomarkers in your own studies, connect with Koneksa to discuss your measurement strategy.