Wearable Biometric Devices: What the Research Says

Fitness watches and rings can reveal useful patterns in sleep, heart rate, and activity—but how much should you trust the numbers?

October 21, 2026 · By Kirstin, Licensed Naturopathic Doctor

Biometric Tracking: What’s the Big Deal?

Wearable devices have moved beyond counting steps. Many watches and rings now estimate heart rate, sleep, activity, skin-temperature trends, and other measures. These data can help people notice changes over time—information that may be hard to capture during a single office visit.

There is a downside, too: a single “poor recovery” score can create more anxiety than insight. This guide looks at what wearables measure well, where the evidence is less certain, and how to use the information constructively with a healthcare professional.

What do these devices measure?

Most consumer wearables combine several sensors:

• Light-based sensors (photoplethysmography, or PPG): shine light into the skin and detect changes in blood volume. Devices use these signals to estimate heart rate and, with additional processing, heart-rate variability (HRV) and sometimes blood oxygen saturation.
• Motion sensors (accelerometers): detect movement and help estimate steps, activity, and sleep versus wakefulness.
• Electrical sensors: some watches record a single-lead electrocardiogram (ECG), which can help assess certain heart-rhythm patterns.
• Temperature sensors: some devices track skin-temperature changes. These are trends in skin temperature, not the same as a clinical body-temperature measurement.

The basic principle is simple: a device is generally better at measuring a signal directly than at estimating something indirectly from several signals. Software algorithms also affect the result.

What wearables do well—and where to be cautious

1. Heart-rate tracking

Wrist-worn devices can be useful for following resting heart rate and trends during steady activity. In a 2025 validation study, several wrist-worn devices performed reasonably well overall, but accuracy declined during rapid changes in heart rate. Results also varied by device.

2. Sleep tracking

Consumer devices estimate sleep from movement and physiological signals; they do not measure sleep in the same way as a sleep-laboratory study. Many are better at estimating sleep duration and broad sleep patterns than at identifying the exact stage of sleep you were in at a given moment.

In a 2024 study of the Oura Ring Generation 3 involving 96 participants and comparisons with polysomnography (a clinical sleep study), the ring detected sleep with high sensitivity, but it was less reliable at identifying wakefulness. Accuracy also varied by sleep stage. These results apply to the specific device, algorithm, study population, and methods—not automatically to every wearable.

3. Heart-rate variability and “recovery” scores

Heart-rate variability (HRV) describes variation in the time between heartbeats. It can provide useful information about changes in the body, but it is influenced by many factors, including sleep, illness, stress, physical activity, and measurement conditions.

Readiness, recovery, and stress scores are calculated using company-specific algorithms. There is no single universal formula or scale, so a score from one brand should not be assumed to mean the same thing as a score from another. Research on WHOOP's metrics includes a 2024 preprint systematic review; because that review was posted as a preprint, its conclusions should be considered preliminary until peer review and publication status are confirmed.

4. Atrial fibrillation alerts

Atrial fibrillation (AF) is an irregular heart rhythm that can increase the risk of stroke. Some consumer devices can record an ECG or notify the wearer of a possible irregular rhythm. Evidence for certain features is encouraging, but performance varies by device, method, and study population.

A 2024 systematic review and meta-analysis of Apple Watch ECG studies reported pooled sensitivity of 94.8% and specificity of 95% for detecting AF. Sensitivity describes how often a test identifies people who have the condition; specificity describes how often it correctly identifies people who do not. These pooled results do not mean every watch reading is correct, nor do they apply to every brand or to every kind of notification. Study results can also differ depending on who is tested and how the device is used.

A wearable alert is not a complete medical evaluation. A clinician may need to review the actual tracing and arrange a standard ECG or other monitoring. A lack of an alert does not rule out an arrhythmia.

5. Activity and step counts

Step counts can be useful for seeing whether your activity is increasing, decreasing, or staying steady. They are not perfectly precise: results can differ by device, walking speed, movement pattern, and how the device is worn. A study comparing the Fitbit Inspire 3 with a research-grade device found a difference in average daily step counts, illustrating why two devices may not report identical totals.

Where wearables fall short

They show patterns, not causes

A wearable might show that your resting heart rate has risen over several weeks or that your sleep appears more fragmented. It cannot determine why. Possible explanations may include illness, stress, alcohol, medication changes, changes in activity, or measurement error. The data can point to a question worth exploring, but it cannot answer that question on its own.

Accuracy varies by metric

• Heart rate: often useful for resting measurements and trends, with limitations during movement or rapid changes.
• Sleep stages: estimates can differ from clinical sleep testing and from one device to another.
• Calories burned: estimates depend on indirect calculations and can be substantially inaccurate for an individual.
• Blood oxygen: readings from consumer devices may be affected by fit, circulation, movement, skin characteristics, and other factors. Do not rely on a wearable alone to assess a medical problem.

Accuracy is not one property that a device either has or lacks. It depends on what is being measured, the device and algorithm, the person wearing it, and the conditions in which the measurement is taken.

When tracking itself becomes stressful

In 2017, researchers described “orthosomnia”—a term for preoccupation with achieving ideal sleep-tracker results—in a case series of patients seeking care for sleep concerns. The report raised an important point: for some people, trying to perfect a sleep score can add anxiety rather than improve rest. This does not mean sleep tracking is harmful for everyone; it means the effect on the individual matters.

If checking your data makes you anxious, disrupts sleep, or pushes you to ignore how you actually feel, consider taking a break or turning off selected notifications. You can also discuss that experience with a healthcare professional.

How to use wearable data with your healthcare provider

1. Bring trends, not just one reading. A few weeks of information is often more useful than a single unusual night.
2. Include context. Note illness, travel, alcohol, medication changes, stressful periods, and changes in exercise.
3. Describe what you noticed. For example: “My resting heart rate has increased from my usual range over three weeks, and I've also felt more tired.”
4. Ask a specific question. “Is this pattern worth investigating?” is more useful than asking whether a score is simply “good” or “bad.”
5. Say if tracking is making you anxious. That is useful information, not a personal failing.
6. Do not self-diagnose or self-treat based on a wearable alone. New, persistent, or concerning readings deserve appropriate clinical context.

The bottom line

Wearables can make patterns in heart rate, sleep, and activity easier to see. Their value is greatest when you understand what a particular metric can—and cannot—tell you. Accuracy differs across devices and measurements, and an estimate cannot explain the cause of a change.

Used thoughtfully, your data can be a helpful starting point for a better conversation with a healthcare professional—not a verdict about your health.

References

The research below supports the examples discussed in this article. Findings from a specific device or study should not be generalized to every wearable or every user.

1. Van Oost CN, et al. “Accuracy of Heart Rate Measurement Under Transient States: A Validation Study of Wearables for Real-Life Monitoring.” Sensors. 2025;25(20):6319. Read the study.
2. Schyvens AM, et al. “Validity and reliability of the Oura Ring Generation 3 (Gen3) with Oura sleep staging algorithm 2.0 (OSSA 2.0) when compared to multi-night ambulatory polysomnography.” Sleep Medicine. 2024;115:251–263. Read the study.
3. “Accuracy, utility and applicability of the WHOOP wearable monitoring device: a systematic review.” medRxiv preprint. 2024. Read the preprint.
4. “Diagnostic Accuracy of Apple Watch Electrocardiogram for Atrial Fibrillation: A Systematic Review and Meta-Analysis.” JACC: Advances. 2024. Read the review.
5. “Clinical validation of 5 direct-to-consumer wearable smart devices to detect atrial fibrillation: BASEL Wearable Study.” JACC: Clinical Electrophysiology. 2023. Read the study.
6. “Detection of atrial fibrillation in a large population using wearable devices.” Circulation. 2022. Read the study.
7. “Fitbit's accuracy to measure short bouts of stepping and sedentary behavior.” 2024. Read the study.
8. “Comparison of daily step count between the Fitbit Inspire 3 and a research-grade device.” Frontiers in Rehabilitation Sciences. 2024. Read the study.
9. Baron KG, et al. “Orthosomnia: are some patients taking the quantified self too far?” Journal of Clinical Sleep Medicine. 2017. Read the case series.