Developing a Simple and Non-Invasive Method for Continuous Glucose Monitoring and Early Diabetes Prediction in Healthy Individuals
A promising breakthrough has been achieved by researchers at the University of Tokyo, who have created a wearable technology capable of continuously tracking blood glucose levels. This innovation aims to facilitate early diagnosis of diabetes and improve risk assessment without the need for traditional blood draws or complex medical procedures. The study detailing this method was published in Communications Medicine, outlining the technical approach and its potential applications.
The Challenge of Managing Diabetes
Controlling and understanding blood glucose levels remains one of the most critical and challenging goals for public health worldwide. The stakes are high: in Italy alone, the National Institute of Health reports that nearly 5% of the adult population, approximately 4 million people aged between 18 and 69, are living with type 2 diabetes. Additionally, around 1.5 million Italians are unaware they have the condition—yet remain at risk.
Globally, the numbers are even more alarming, with approximately 530 million people affected by diabetes as of recent estimates. Projections indicate that by 2030, this figure could rise to 642 million, and by 2045, up to 738 million. The sheer scale of this epidemic underscores the importance of developing effective screening and management tools. So, why focus on creating an algorithm?
The Limits of Traditional Testing
According to the research team, standard diagnostic tests for diabetes often fall short in capturing the dynamic nature of glucose regulation within the body. Conventional tests, like fasting blood glucose, HbA1c, or oral glucose tolerance tests (OGTT), provide snapshots that may miss early or subtle impairments in glucose handling. A more nuanced and continuous understanding of how blood sugar fluctuates over time could enable earlier intervention and better risk prediction—something that wearable technology might fulfill.
The Study
The research was designed to assess the usefulness of indices derived from Continuous Glucose Monitoring (CGM), a technology that measures blood glucose levels at regular time intervals. Unlike traditional, resource-intensive tests, these indices are relatively easy to obtain and interpret, offering a practical way to monitor glucose regulation.
The wearable device was tested on 64 individuals with no prior diagnosis of diabetes. These participants underwent CGM alongside other tests such as the oral glucose tolerance test (OGTT) and clamp procedures—specifically, hyperglycemic and hyperinsulinemic-euglycemic clamps—which are considered gold standards for assessing glucose metabolism. By comparing data from different sources, researchers validated the predictive accuracy of indices derived from CGM in an independent dataset from another country, supported by mathematical models using simulated data.
A key finding from this prospective study was that an index based on CGM data, which measures the autocorrelation function of blood glucose levels (called AC_Var), correlates strongly with the Disposition Index (DI)—a well-established measure of the body’s capacity to handle glucose and a predictor of future diabetes development.
This model, which combines the standard deviation of glucose levels measured by CGM with AC_Var, appears to outperform traditional diagnostic markers like fasting glucose, HbA1c, or OGTT in predicting DI derived from clamp tests. Additional computer simulations confirmed a strong association between AC_Var and the DI, suggesting that this index might be a powerful tool for early detection.
In Summary
The new algorithm has shown promise in identifying individuals with impaired glucose regulation even when standard diagnostic tests show normal results. This indicates a potential for detecting early signs of diabetes before clinical diagnosis, allowing for preventive measures to be implemented proactively.
Indices derived from CGM, especially AC_Var, could become essential tools for managing blood sugar in populations without diagnosed diabetes. Moreover, to facilitate broader use and validation, the researchers are currently developing a web application that calculates these indices based on CGM data, making this innovative approach accessible and easy to implement.
Sources
Sugimoto H, Hironaka K, Nakamura T et al. "Improved detection of decreased glucose handling capacities via continuous glucose monitoring-derived indices." Communications Medicine, 2025, Volume 5, Article 103. Link to the article
