Researchers at the Swiss Federal Institute of Technology (ETH), Zurich, built a handheld breathalyzer that can show whether a person’s body is burning fat or carbohydrates (Device 2026, DOI: 10.1016/j.device.2026.101226).
The device measures acetone in exhaled breath, which can indicate when the body is burning fat. When the liver runs short of carbohydrates, it breaks down fatty acids and produces ketones. Acetone is a volatile by-product of this process, so it diffuses into the lungs and is released in breath.
“We wanted to make metabolic monitoring as simple and accessible as stepping onto a scale,” says Andreas Güntner, who led the work. “People respond differently to the same diet or exercise program, yet they currently have few practical ways to see when their metabolism switches from using carbohydrates to burning fat.”
The device’s sensor is made of a chemoresistive material, whose electrical resistance changes when chemicals interact with its surface. The chosen sensor material has a high baseline resistance when exposed to air. When the exhaled breath contains acetone, it binds to the sensor’s surface and injects electrons into the material, lowering its resistance. The device measures this drop, which corresponds to acetone in the exhaled breath.
But the sensor also reacts to water, just as it does to acetone. Because human breath is saturated with water vapor, this would drastically affect the reading. To solve this problem, Güntner’s team added a sorption column in front of the sensor. This tube is packed with a porous polymer called Tenax TA, which attracts acetone 90 times as strongly as water. As a result, water vapor passes through immediately and leaves the device, while the tube traps the acetone and slowly releases it to the sensor. By the time acetone arrives at the sensor, the water is long gone, so the sensor reads a clean acetone signal.
The handheld device was validated against PTR-MS (proton transfer reaction/mass spectrometer)—a large lab instrument used for measuring trace gas concentrations. When both instruments measured the same breath sample at the same moment, the new device’s reading differed from the PTR-MS results by only 7.5% at 2 ppm, a concentration consistent with someone burning mostly carbohydrates. This level of precision is sufficient to detect changes in acetone concentration when the body burns fat.
The device connects to a smartphone app via Bluetooth to guide the user in giving consistent samples. Wolfram Miekisch, a breath analysis researcher at Rostock University Medical Center who was not involved in the study, thinks this sampling approach is very innovative and critical to the sensor’s success.
Alivion, an ETH Zurich spin-off company involving Güntner and the paper’s coauthor Jan van den Broek, commercialized the device. It has not yet been validated or approved as a diagnostic device for clinical use. Güntner says the next steps are larger clinical validation studies and the development of tailored solutions for different use cases. He says someone optimizing athletic performance, trying to lose weight, or receiving glucagon-like peptide 1 therapy will need customized clinical interpretation and decision support.
The team is also collaborating with University Children’s Hospital Zurich to test whether the device can help children with epilepsy monitor ketogenic diets without blood tests.
“This looks promising,” Miekisch says. “The introduction of easily accessible, highly specific sensors in the field of breath research will be a game changer.”
