Author perspective
“The honest reading of this piece is that the sensors stopped being the hard part a while ago. What remains is validation at scale, a credible regulatory path, an ethical framework, and a business capable of actually manufacturing and deploying these systems. Those challenges may be less interesting to write papers about, but they are what ultimately determine whether any of this reaches a patient.”
— Roozbeh Ghaffari, PhD · CEO & Co-Founder, Epicore Biosystems
Study findings
Consumer wearables count steps and read heart rate. They measure physics, not chemistry, and that ceiling is the subject of this Perspective: seventeen researchers across Switzerland, the United States, Germany, the United Kingdom and Tanzania assessing what it would take for wearables to measure actual biochemical molecules in the body, and what stands in the way.
The article surveys the accessible fluids one by one. Sweat, which is easy to reach, chemically simpler than blood, and uniquely able to be stimulated on demand. Interstitial fluid, whose concentrations track blood closely for small molecules and which already supports continuous glucose monitoring. Breath, containing more than 3,500 substances but at concentrations a thousand to ten thousand times lower than blood. Then saliva, tears, wound fluid, urine and stool, each with its own collection problem.
The conclusion is that the sensors are no longer the hard part. What’s missing is everything around them: large-scale validation across real conditions, a regulatory path that treats these as in vitro diagnostics rather than fitness trackers, an ethical framework for continuous molecular surveillance of a person, sociocultural acceptance, and commercialisation, which the authors note plainly is what makes reimbursement possible.
Scientific highlights
Sweat has advantages the other fluids don’t
Easier access, lower matrix complexity, less contamination risk, and the only biofluid that can be actively stimulated on demand through iontophoresis.
The bottleneck is validation, not sensing
Widespread adoption requires large-scale validation across varied conditions, plus demonstrated clinical value to patients, clinicians and payers alike.
These count as diagnostics, not fitness devices
In Europe they fall under the In Vitro Diagnostics Regulation (EU) 2017/746, a different and heavier regulatory path than the one biophysical wearables follow.
Commercialisation is named as a prerequisite
The authors list commercialisation among the five conditions for mainstream adoption, because it is what makes reimbursement possible.
Why this research matters
This is an independent assessment of the field, led from ETH Zurich, published in Nature as a Perspective rather than as new experimental work. Its value is the vantage point: a group with no single commercial stake laying out what separates a laboratory demonstration from clinical practice.
Among the references the authors single out for comment is the Gx Sweat Patch paper, which they describe as an example of successful translation and commercialisation of a body-fluid analysing device, and as interdisciplinary collaboration between engineers, physiologists and a business partner. In an article largely concerned with why so little has reached real use, that is a specific nod to work that did.
Applications:
- Clinical diagnostics
- Paediatric care
- Global health access
- Therapeutic drug monitoring
- Public health
How this connects to Epicore’s solutions
Epicore CEO Roozbeh Ghaffari is one of the seventeen authors, contributing to the section on the environment required for mainstream adoption, alongside Epicore co-founder and board member John A. Rogers. The article is led by researchers at ETH Zurich’s Institute of Translational Medicine, so this is Epicore contributing to an independent assessment of the field rather than presenting its own work.
The barriers the article identifies are the ones Epicore’s commercial products has cleared in practice: validating a measurement against laboratory reference methods, manufacturing at a cost that supports real deployment, and producing something people will actually wear through a full shift. The article’s framing of sweat as the most accessible biofluid, and the only one that can be stimulated on demand, is the premise the company is built on.
Connected Hydration
Real-time sweat and hydration sensing for industrial workers, with personalized intervention before heat stress sets in.
GX Sweat Patch
A wearable microfluidic patch that gives athletes sweat rate and electrolyte feedback to tailor a personalized action plan.
Discovery Patch
An FDA-registered collection system for multi-biomarker sweat analysis in clinical studies, the foundation for cortisol work.
FAQs
A wearable device can measure more than physical signals. While a smartwatch tracks things like pulse and movement, a biosensor measures molecules, including glucose, sodium, cortisol, or proteins, associated with inflammation. It does this using a sensing element, such as an enzyme or antibody, that recognizes a specific molecule and converts that interaction into a measurable signal, such as a color change or electrical response.
A fitness tracker measures biophysical signals: heart rate, motion, skin temperature, blood oxygen. A biochemical wearable measures what molecules are present and at what concentration. The distinction matters for regulation as much as for capability, because in Europe a biochemical wearable is an in vitro diagnostic device and a heart-rate monitor is not.
This article covers eight. Sweat and interstitial fluid are the most developed, followed by exhaled breath. Saliva, tears, wound fluid, urine and stool are all being worked on, usually in specific forms such as a smart contact lens, an instrumented bandage, or a sensor built into a diaper.
Blood is the clinical gold standard, but accessing it requires breaking the skin, making continuous measurement outside a hospital or clinic impractical. Sweat, by contrast, reaches the skin’s surface naturally and carries electrolytes, metabolites, hormones, and proteins. Its simpler chemistry can also make these signals easier to measure. Most uniquely, sweat can be produced on demand: a small electrical current can deliver a sweat-stimulating drug into the skin, creating a controllable, non-invasive window into human physiology.
The fluid that surrounds cells in your tissues, just beneath the skin. It exchanges chemicals with the bloodstream continuously, so for small molecules its concentrations closely follow blood concentrations. Continuous glucose monitors measure it, which makes it the one clear commercial success in wearable biochemical sensing so far.
The ones that measure molecules generally are. In the European Union they fall under the In Vitro Diagnostics Regulation (EU) 2017/746, which requires verification of the individual components, validation of the whole device in real-world use, testing of any software or algorithms, and review by a notified body. Devices measuring only physical signals face a lighter path.
Several things at once, according to this article. Data has to be reproducible across varied real conditions, not just in a laboratory. Clinical value has to be demonstrated separately to patients, clinicians and the payers who would reimburse it. False positives create costly and burdensome follow-up visits. And a practical point the authors make bluntly: more data does not automatically produce better outcomes if the sampling frequency doesn’t match how fast the condition actually changes.
A health measurement captured by a digital device rather than a laboratory. A digital endpoint is the stricter version: a digital biomarker validated well enough to serve as an official outcome measure in a clinical trial. Getting from one to the other is the validation work this article argues the field has been skipping.
The article argues they could, and treats it as one of the strongest arguments for the technology, since underdiagnosis and late presentation are the core problems in those settings. But it is specific about the conditions: local policy support, stable electricity and network coverage, adaptation to different climates, sociocultural acceptance, and, most importantly in the authors’ view, diagnostics that are actually connected to available treatment.
