Media release

From: University of Technology Sydney (UTS)

Sweat contains a wealth of biological information that, with the help of artificial intelligence and next-generation sensors, could transform how we monitor our health and wellbeing, a new study suggests.

The study, published in the Journal of Pharmaceutical Analysis, examines sweat's potential for real-time monitoring of hormones and other biomarkers, medication doses, and early detection of diseases such as diabetes, cancer, Parkinson's and Alzheimer's.

“Collecting sweat is painless, simple and non-invasive,” said co-author Dr Dayanne Bordin, an analytical chemist at the University of Technology Sydney (UTS). “It’s an attractive alternative to blood or urine, especially for continuous monitoring in real-time.”

“Anyone who is already interested in tracking their health using wearables such as an Apple watch – for example their heart rate, step count or blood pressure – would be interested in the information sweat can provide.

“There are already sweat monitoring devices on the market such as the Gatorade sweat patch, which is a single-use, wearable sticker that pairs with an app to analyse your sweat rate and sodium loss, and provide tailored advice.”

Recent advances in microfluidics, stretchable electronics and wireless communications have led to a new generation of wearable sensors. These thin, flexible patches adhere to the skin and continuously sample sweat.

Combined with artificial intelligence, these devices could detect specific metabolites and interpret complex biochemical patterns, offering users personalised health insights and early warning for a range of diseases.

Athletes could monitor electrolyte loss during training, and provide proof that they are drug free before competitions. Diabetic patients might one day wear a patch that detects glucose changes through sweat instead of blood tests.

“Sweat is an under-used diagnostic fluid,” said co-author Dr Janice McCauley from the UTS Faculty of Science.

“The ability to measure multiple biomarkers simultaneously, and transmit that data wirelessly, provides enormous potential for preventive health care.

“The year 2023 was marked by an evolutionary step in artificial intelligence, opening the door for improved pattern analysis and classification algorithms to improve diagnostic precision and therapeutic accuracy,” she said.

AI can now process huge datasets to link subtle molecular signals in sweat to specific physiological states. The next step, the authors suggest, is integrating this with compact, low-power devices with secure data transmission.

UTS researchers are currently working on understanding the baseline physiological aspects of sweat. They are also developing microfluidic devices that are sensitive enough to detect trace amounts of biomarkers such as glucose and cortisol.

While much of the research remains at the prototype stage, commercial interest is growing.

“We’re not far from a future where your wearable can tell you when you’ve got high stress hormone levels, and by monitoring this over time, whether you are at risk of chronic health conditions,” Dr Bordin said.