Wearable health technology: adoption, capabilities and limits

A watch that records a walk, a ring that summarizes sleep and a glucose sensor used in diabetes care can all be called wearable health technology. They collect different information, however, and their results need different kinds of interpretation. A useful record of daily activity is not the same service as clinical monitoring.

Wearables bring together sensors, software, connected phones and sometimes a healthcare team. Understanding that arrangement matters more than the promise of a new sensor. Here is a look at adoption, algorithms, regulation and the practical questions facing patients, caregivers and clinicians.

Adoption: what the numbers actually describe

In its 2023 survey of US adults, Rock Health reported that 44% owned wearable health tracking devices. The finding appeared in an August 2024 article. It describes ownership in that survey, rather than daily use, clinical use or a current estimate for 2026.

Those distinctions matter when reading a claim that wearables have become mainstream. Owning a tracker does not tell us whether someone still wears it, finds its information useful or shares it with a clinician. Rock Health's work considers adoption over time; an ownership figure alone cannot explain what happens after a device is bought.

Market forecasts answer another question: how much business researchers expect a defined category to generate. Check which products, regions and starting year a forecast includes. A projection for medical devices should not be treated as a measured improvement in health, or silently applied to every consumer fitness tracker.

Algorithms: estimates need a stated purpose

Software can organize readings, look for patterns and produce a notification or score. That creates opportunities to make a long record easier to review. It also raises a question: what was the particular function evaluated to do?

Apple Watch and Fitbit are familiar examples in discussions of heart and activity features. A brand name or an AI label does not establish that every model has the same sensors, software, indications or supported population. Check the documentation for the exact function instead of transferring a claim from one feature to an entire product family.

The FDA's information on AI-enabled medical devices concerns specific medical uses. Its transparency principles call for clear information about intended use, performance and limitations. This is a useful standard for reading claims about an algorithm: look for its validation, the people represented in that work and known failure modes.

A stress-management score does not, by itself, demonstrate that a device predicts an anxiety episode. Likewise, an early-disease-detection claim needs evidence for that condition and function. Do not assume a consumer score can diagnose a problem or select treatment.

Sensors: separate the measurement from the interpretation

Movement, pulse signals and other sensor inputs can feed several estimates. A dashboard can therefore display more categories than the device directly measures. Ask which information comes from a sensor, which is calculated and which you enter yourself.

Continuous glucose monitoring illustrates why the distinction matters. NIDDK explains that CGM sensors estimate glucose in the fluid between cells. The system can send readings to a phone, receiver or compatible equipment. That is a particular sensing arrangement, not evidence that an ordinary watch or ring can independently measure glucose.

The FDA warns against smartwatches and rings claiming to measure blood glucose on their own without piercing the skin. Displaying readings from a separate glucose sensor is different from measuring glucose at the wrist.

Apply the same scrutiny to claims about hydration, blood pressure or respiratory health. Which sensor and method are involved? Is the feature intended for general wellness, research or a particular medical use? A developing technology should not be described as a capability available on every wearable today.

Regulation: cleared, approved and listed are different

FDA status belongs to a specific device or function and its intended use. It is not a blanket description of a company's entire product range.

The 510(k) clearance process considers substantial equivalence to a legally marketed device. Premarket approval is a different pathway. Registration or listing does not establish either: the FDA explains these distinctions for consumers.

For an ECG, oxygen-related feature or another medical function, check the exact decision and labeling rather than relying on a general claim that wearables are FDA-approved. Review the population, use conditions and instructions. A regulatory decision does not mean every score on the same device has the same status, or that it detects every problem.

Connected platforms also depend on continuing support

An app can help organize information from more than one source. Apple's Health documentation, for example, explains how to manage data sources and access. Check the current settings and supported integrations rather than assuming every record is shared automatically.

Apple, Google's Fitbit and Amazon have all appeared in discussions of consumer health technology, but their services do not share the same history. Amazon ended Halo support on July 31, 2023. Halo should not be presented as an expanding active platform in a current overview.

Connections to a clinic or electronic health record need their own checks. Find out what the clinic accepts, whether anyone reviews the information and how it becomes part of care. A working app connection is not a promise of continuous clinical observation.

Prevention and other opportunities

Activity records, sleep summaries and reminders can give people something concrete to discuss or review. The opportunity is to connect that information to a useful action. The action and its benefits still need evaluation; recording a metric does not establish prevention of a disease.

Several possibilities deserve separate consideration:

• Personalized care: A clinician may consider a person's record alongside symptoms, history and other evidence. More measurements do not automatically produce a better treatment plan.
• Patient participation: A person can bring questions about a pattern to an appointment. The record needs enough context to be interpreted, rather than presented as a diagnosis.
• Remote monitoring: An agreed service can specify what is collected, who reviews it and how concerns are handled. It should also explain what happens when connectivity or equipment fails.
• Research: Repeated measurements can support defined research questions. Consent, data quality and the participants represented in a study remain part of assessing its findings.

Claims that wearables generally reduce hospitalizations or healthcare costs require evidence for the particular service, population and outcome. Those benefits should not be promised simply because data can be collected remotely.

Challenges that affect everyday use

Privacy is part of the arrangement. Review which information leaves the device, who receives it, what optional sharing is enabled and how records can be exported or deleted. Consider caregiver access separately from permission for advertising or unrelated services.

Accuracy also needs a defined task. Read about missing readings, known limitations and the circumstances in which a result should be checked another way. A confident-looking chart can still contain gaps or estimates.

Access involves more than the purchase price. A compatible phone, internet connection, replacement supplies, subscriptions, charging, comfort and help with setup can affect whether someone can use a service. Check the actual requirements and costs before relying on it.

For developers and healthcare organizations, regulatory work, validation and ongoing support are practical obligations. Adding a medical claim changes the evidence needed. Calling a feature general wellness does not supply evidence for a disease-related promise.

What it means for patients, caregivers and clinicians

For patients

Start with the job you want the information to serve. Is it a record of activity, a reminder you choose to use or a medical function discussed with your clinician? Check the exact feature, compatibility and limitations before relying on a product description.

If a device is part of your care, agree on which readings matter and what to do when information is missing or conflicts with how you feel. Do not change treatment on the strength of a general wellness score.

For caregivers

Agree on sharing with the person using the device. Decide who receives notifications, what they mean and what to do if no reading arrives. Review the limits of any fall-related feature or reminder rather than assuming it guarantees safety or an immediate response.

A shared account or app connection should come with clear expectations. It should not quietly become a substitute for the contact or support a person needs.

For clinicians

Assess the specific device, function, evidence and intended use. Decide how incoming information will be reviewed, documented and acted on before adding it to a care workflow. A regulatory status check is one part of that assessment, not a substitute for it.

Also consider the work expected of patients and staff: setup, training, incomplete data, follow-up and support. A proposal for remote monitoring should describe that service as carefully as the sensor.

A clearer role for connected health

Wearable health technology offers several different tools, not one uniform kind of care. Adoption numbers describe who owns devices; sensor documentation explains what is measured; validation and labeling define a particular function's evidence and limits.

The practical question is what happens after a reading is collected. Who can interpret it, what action is appropriate and can the person using the service get help when it fails? Those questions keep the focus on a supported, understandable use of the technology.