Can Smartwatches and Blood Tests Predict the Risk of Diabetes Before It’s Too Late?
Type 2 diabetes now affects over 500 million adults worldwide, a number that could rise to 640 million by 2030. In nine out of ten cases, this disease is linked to a phenomenon called insulin resistance. This silent disorder occurs when the body’s cells respond less effectively to insulin, a hormone essential for regulating blood sugar levels. Without intervention, it can progress to full-blown diabetes or serious complications such as heart disease or liver damage.
However, detecting this resistance remains challenging. Current methods, such as laboratory blood tests, are expensive and not easily accessible. A recent study shows that it is possible to detect it more simply by combining data from smartwatches and routine blood tests. Researchers used information such as resting heart rate, daily step count, sleep duration, and triglyceride and cholesterol levels. Using artificial intelligence, this data can identify at-risk individuals with nearly 80% accuracy.
The study involved over 1,000 participants in the United States. The results reveal that certain lifestyle habits, such as low physical activity or insufficient sleep, are closely linked to increased insulin resistance. For example, overweight or obese individuals have a much higher risk, but even those with a normal weight can be affected. Among the participants, one in five had insulin resistance without knowing it, even though their blood sugar levels appeared normal.
The advantage of this approach is its simplicity. Smartwatches continuously measure indicators such as heart rate and physical activity, while standard blood tests provide metabolic data. By cross-referencing this information, scientists have developed a model capable of predicting risk long before the first symptoms appear. Early detection opens the door to targeted interventions: weight loss, regular exercise, or a tailored diet can reverse the trend.
This method could revolutionize diabetes prevention. It avoids complex and costly exams while offering a scalable solution accessible to millions of people. In the long term, it could even be integrated into voice assistants or health apps, providing personalized recommendations to reduce risks. The stakes are high, as early action can prevent heavy treatments and irreversible complications. Technology, combined with medicine, thus becomes a valuable tool for proactive health.
About Our Sources
Original Publication
DOI: https://doi.org/10.1038/s41586-026-10179-2
Title: Insulin resistance prediction from wearables and routine blood biomarkers
Journal: Nature
Publisher: Springer Science and Business Media LLC
Authors: Ahmed A. Metwally; A. Ali Heydari; Daniel McDuff; Alexandru Solot; Zeinab Esmaeilpour; Anthony Z. Faranesh; Menglian Zhou; Girish Narayanswamy; Maxwell A. Xu; Xin Liu; Yuzhe Yang; David B. Savage; Mark Malhotra; Conor Heneghan; Shwetak Patel; Cathy Speed; Javier L. Prieto