How Movement Affects Blood Sugar: A New Tool Helps Predict Glucose Changes

People with Type 1 diabetes know that movement affects their blood glucose—but predicting exactly how much, and when, remains challenging. Physical activity lowers blood glucose levels, and this effect depends on both the intensity and duration of the activity. However, the real world is messy. Unlike structured exercise sessions, daily life includes variable activities that are hard to anticipate and plan for.

Understanding these glucose changes can help reduce the risk of hypoglycemia—dangerously low blood sugar—which is a major concern for people managing Type 1 diabetes. That's why researchers are working to build better tools to predict how different types of movement will affect blood glucose in everyday situations.

Researchers in this study created a personalized prediction model called ACTIVE-GLU to forecast glucose changes based on physical activity. The model works by combining data from two wearable devices: a continuous glucose monitor that tracks blood sugar in real time, and a Garmin smartwatch that tracks movement.

Fourteen adults with Type 1 diabetes participated in the study. Researchers categorized their activities into four levels: low, medium, high, and very high intensity. The model then calculated what happened to each person's blood glucose at 15-minute intervals after each type of activity. By analyzing each person's unique patterns, the researchers built a personalized model rather than applying one-size-fits-all predictions.

The key advantage of this approach is that it's based on real life. Participants wore the devices during their normal days, not during controlled laboratory exercise sessions. This means the model captures how the body responds to activity in genuine, messy, everyday conditions.

The ACTIVE-GLU model successfully predicted blood glucose changes across all activity intensity levels. The prediction accuracy was strongest during low intensity activity, with an average accuracy of 81.67%. During very high intensity activity, the model's predictions had a larger margin of error, with a mean absolute error of 0.44 mmol/L—but this is still clinically useful for diabetes management.

When researchers compared the model's predictions to actual glucose values using a statistical method called Bland-Altman analysis, they found close agreement. Importantly, the predictions showed negligible bias, meaning the model didn't consistently overpredict or underpredict across different activity intensities.

These results suggest that personalizing predictions to each individual's unique glucose response patterns can provide accurate, real-world forecasting of how movement will affect blood sugar.

This exploratory study demonstrates that combining activity tracking and continuous glucose monitoring data can help predict blood glucose responses to everyday movement. If this approach is validated in larger studies, it could eventually become a practical tool for people with Type 1 diabetes.

Such a tool might help users anticipate when low blood sugar could occur and adjust food intake, insulin dosing, or activity timing accordingly. However, this research is preliminary. The study included only 14 participants, and more research is needed to confirm these findings in larger, more diverse groups and to determine whether this model can be reliably used to guide real-world diabetes management decisions.

Source: Digital health. Evidence type: PubMed indexed literature. Type1Cure is an information and intelligence hub, not a medical advice service. This article summarizes published research and does not provide diagnosis, treatment, or personal medical guidance. Always talk to your own care team before changing anything about your Type 1 diabetes management.