Summary
Wearable devices are increasingly being used in clinical trials to collect continuous physiological data outside research centres. Their broader adoption depends on measurement accuracy, technical reliability, meaningful clinical endpoints and responsible data governance.
Wearable devices are moving beyond fitness tracking into clinical-trial research, where they can record physiological data continuously as participants live at home. A Nature Spotlight published on 16 September 2026 describes growing use of smartwatches, rings, patches, chest straps and sensor-equipped clothing, while showing that adoption remains gradual because the measurements must be reliable, clinically useful and acceptable to regulators and participants.
The evidence described is a synthesis of registry data, clinical-trial examples and expert assessments rather than a new intervention study. Its central finding is that wearables can produce richer data than occasional visits to a research centre, but collecting more data does not by itself make a trial stronger.
How wearables are being used in trials
Different sensors determine what a device can measure. Accelerometers and gyroscopes track movement. Photoplethysmography, or PPG, uses changes in light absorption and reflection from blood vessels near the skin to estimate heart rate, heart-rate variability, sleep stages and peripheral blood oxygenation. Barometric-pressure sensors can indicate vertical movement and posture, while temperature sensors can contribute to monitoring sleep–wake cycles, fever, inflammation and hormonal changes during the menstrual cycle.
The number of drug trials using wearable data has increased, although slowly. A ClinicalTrials.gov search described in the article found that such trials rose from fewer than 30 in 2014 to 128 in 2024. A 2025 review reported the greatest use of smartwatches in cardiology, which accounted for 28.7% of applications, followed by neurology at 21.8% and oncology at 11.5%.
Cardiology provides one example of how continuous measurements can contribute to a trial. In a study of people with permanent atrial fibrillation and symptoms of heart failure, Fitbit trackers were used alongside other methods to compare digoxin with the beta-blocker bisoprolol. The study found that the two drugs were equally effective at controlling heart rate. Physical-activity data from the Fitbit were also as good as standard clinical tests at predicting heart-failure severity. The study collected about 150 million heart-rate data points, according to the article.
Why continuous data matter
A clinic visit captures a limited period under unfamiliar conditions. A wearable can record changes while participants sleep, exercise, work and carry out ordinary activities. This may provide a more representative picture of how a disease or treatment affects daily life and can reduce dependence on scheduled measurements.
The data could also support faster responses during a trial. A clinical data-management specialist quoted by Nature says artificial-intelligence tools could identify readings outside safe limits and generate alerts so that participants can receive emergency care. Wearables may also reduce some visits and follow-ups, although the devices bring their own costs for distribution, training, support and data management.
The technology could eventually expand beyond measurements already common in consumer devices. The article identifies possible future applications for substances such as glutamate and cortisol. These are potential extensions of wearable sensing, not routine capabilities described for current clinical-trial devices.
Reliability is the gateway to wider adoption
Technical failures can directly affect a trial. In an epilepsy study, an unexpected firmware update caused the app recording seizure data to crash. Devices had to be recalled, participants needed individual technical support and the disruption threatened continued participation. Around-the-clock assistance and stable software are therefore part of the trial infrastructure, not optional conveniences.
Regulatory requirements add another test. Researchers must show that a device measures a physiological parameter accurately and reliably, works appropriately in the target population and supports a measurable clinical endpoint, such as decreased pain. A clinical endpoint is an outcome used to judge whether an intervention has produced a meaningful effect.
Privacy and data security are also central because wearables can generate detailed, continuous records of health and behaviour. The article identifies ethical concerns about the role of large technology companies in healthcare alongside the scientific and regulatory questions. The path to wider use will therefore depend on linking sensor data to valid trial outcomes while maintaining dependable devices, participant support and responsible handling of personal information.