The Promise of Hypoglycemia Risk Prediction

In this issue of J Diabetes Sci Technol, several publications discuss the burden of the unpredicted nature of glucose levels at day and night for people with diabetes (PwD) and the potential benefits in reducing their fear of hypoglycemia and diabetes distress offered by a newly commercially available app: the Accu-Chek SmartGuide Predict app.^1-5 This digital application was designed to predict the risk of hypoglycemia and glucose, using a novel real-time continuous glucose monitoring (CGM) sensor. ⁶

Although there is good evidence in the literature that extended hypoglycemia and glucose prediction is possible,^7,8 the new CGM solution is the first one available to PwD with this advanced technology. The algorithms powering the glucose predictions of the app were developed using machine learning applying data from CGM readings, insulin boluses, meal carbohydrate information, and user settings from the PREDICT study, which enrolled 221 adults with type 1 diabetes on multiple daily injections (MDIs) of insulin. ¹ The app includes three functionalities: Low Glucose Predict (LGP), which alerts for low glucose episodes within the next 30 minutes; Glucose Predict (GP), a two-hour prediction of glucose levels; and Night Low Predict (NLP), which estimates the likelihood of experiencing hypoglycemia overnight. ¹

The algorithms underwent extensive testing using various clinical and real-world datasets, including data from individuals with type 1 and type 2 diabetes on MDI or insulin pump therapy. ¹ The demonstrated high performance should translate into valuable real-world usage of this digital tool, benefiting PwD in their daily diabetes management. ¹

Hypoglycemia is a frequent complication of diabetes treatment with insulin or insulin secretagogues, ⁹ and one of the greatest burdens of this condition. Despite advances in CGM technology, daytime and nocturnal hypoglycemia remain clinically relevant and distressing problems for PwD who inject insulin.^2-4 Particularly nocturnal hypoglycemia poses a serious threat because the autonomic and symptomatic response to hypoglycemia is diminished during sleep. ²

Nocturnal hypoglycemia is common among individuals with type 1 diabetes on MDI therapy, ³ but it can also occur among individuals with insulin-treated type 2 diabetes. ¹⁰ Besides type 1 diabetes, other risk factors in PwD for such hypoglycemic events include advanced age, adolescence, female gender, a longer duration of diabetes, high glycemic variability, and the presence of comorbidities like autonomic neuropathy. ²

Hypoglycemic events during the night also tend to last longer than daytime hypoglycemia and are associated with a higher risk of hypoglycemic events on the next day. ³ In addition, the hormonal circadian rhythm causes an alternating period of increased and decreased insulin sensitivity during the night, ² making achievement of stable glycemic control harder for PwD, and unawareness or fear of nocturnal hypoglycemia distressing.

Even though technological advances in CGM have enabled PwD to better monitor and manage glucose control, they have also brought new challenges, like stress and fatigue from alarms, fear of being spotted with hypoglycemia in public, and sleep disturbance. ² Following the recommendations for using the app for daily diabetes management holds the promise of better risk management for hypoglycemia, which can be of value for PwD and their diabetes teams. ⁵

According to a survey published in this journal issue, ⁴ PwD perceive the prospect of receiving low glucose predictions for up to 30 minutes, a continuous glucose prediction for up to two hours, and a nighttime low glucose prediction as an added benefit in being a step ahead of hypoglycemic episodes and reducing the burden of diabetes. ⁴ This includes reducing the fear of hypoglycemia and concomitant self-restrictions in attending social events and performing daily life activities such as driving and cycling. ⁴

The new app ¹ therefore addresses an unmet need for improved glucose control in hypoglycemia. By helping to prevent hypo-events, it aims to empower PwD, reduce diabetes burden and increase their self-esteem and confidence, and to improve daily clinical care. The novel CGM solution includes this innovative application, which leverages the potential use of CGM data and may pave the way for future developments in CGM technology.