Sweat reveals a new way to monitor health without needles
Xiaoyu Yin defended her PhD thesis at the Department of Electrical Engineering on October 20th.
Imagine checking your blood sugar, stress levels, or kidney function — without needles, blood draws, or hospital visits. Sweat-sensing technology promises exactly that: an effortless, non-invasive way to monitor health in real time. But to make sweat a reliable window into what’s happening in the bloodstream, scientists need to understand how different biomarkers travel from blood to sweat. In her PhD research, Xiaoyu Yin developed mathematical models that make it possible to accurately estimate blood concentrations of key biomarkers — glucose, urea, and cortisol — from sweat. The findings demonstrate that sweat can serve as a clinically meaningful and patient-friendly medium for long-term health monitoring, opening new possibilities for more personalized and comfortable healthcare solutions.
Sweat offers an appealing alternative to blood sampling: it’s easy to collect, completely non-invasive, and suitable for continuous measurement. However, translating sweat data into meaningful medical information has long been a challenge, as each biomarker behaves differently in how it moves from blood to sweat.
This research of Xiaoyu Yin addresses that problem through model-based strategies — mathematical descriptions that capture the complex transport mechanisms of various biomarkers and make it possible to infer blood concentrations accurately from sweat readings.
Glucose monitoring made easier
For people with diabetes, continuous glucose monitoring is crucial but often uncomfortable or costly. This research developed a kinetic model describing how glucose moves into sweat, taking into account individual differences in sweat rate due to both environmental and physiological factors.
By combining this model with an optimization algorithm, glucose levels in blood could be estimated with high accuracy from sweat measurements in both healthy individuals and diabetic patients. A simplified version — called the Local Density Random Walk model — reduced computational time to just 2.6 seconds per data point, enabling real-time glucose tracking. This breakthrough could make sweat-based glucose monitoring a practical and user-friendly alternative for diabetes management.
Sweat as a window into kidney health
For patients with end-stage renal disease, frequent measurement of blood urea levels is vital to guide dialysis treatment. This research proposed an innovative model of urea transport, incorporating the role of specific urea transporters in sweat glands.
Tests in dialysis patients showed that sweat urea concentrations could reliably reflect blood levels before and after treatment. This opens the door to home-based urea monitoring, helping patients manage their condition with fewer hospital visits while maintaining optimal treatment schedules.
Tracking stress through cortisol in sweat
Cortisol — often called the ‘stress hormone’ — is closely tied to stress-related disorders, but its measurement is complicated because much of it binds to proteins in the blood. To address this, the research developed a mathematical model capturing the transport of cortisol from blood to sweat.
When tested on data from cardiac surgery patients, the model showed a strong match between estimated and measured cortisol levels. This demonstrates the potential for sweat-based cortisol monitoring in assessing stress and related health conditions over time.
Towards personalized, painless health monitoring
Taken together, these findings show that blood concentrations of glucose, urea, and cortisol can be reliably estimated from sweat. This achievement marks a major step forward in making sweat a practical, clinically relevant diagnostic fluid.
By linking sweat readings to meaningful medical insights, this research of Xiaoyu Yin lays the foundation for non-invasive, continuous, and personalized health monitoring — improving patient comfort, enabling early detection, and transforming how chronic conditions are managed.
The research is part of the larger research project SEDAS TTW 18271 SPS: Sweat sensing device and data analytics for semi-continuous sepsis monitoring.
Title of PhD thesis: . Supervisors: Prof. Massimo Mischi, Prof. and Dr. Elisabetta Peri.