Glucodensity Functional Profiles Outperform Traditional Continuous Glucose Monitoring Metrics

Marcos Matabuena,Rahul Ghosal,Javier Enrique Aguilar,Robert Wagner,Carmen Fernández Merino,Juan Sánchez Castro,Vadim Zipunnikov,Jukka-Pekka Onnela,Francisco Gude

DOI: https://doi.org/10.48550/arXiv.2410.00912

2024-10-02

Abstract:Continuous glucose monitoring (CGM) data has revolutionized the management of type 1 diabetes, particularly when integrated with insulin pumps to mitigate clinical events such as hypoglycemia. Recently, there has been growing interest in utilizing CGM devices in clinical studies involving healthy and diabetes populations. However, efficiently exploiting the high temporal resolution of CGM profiles remains a significant challenge. Numerous indices -- such as time-in-range metrics and glucose variability measures -- have been proposed, but evidence suggests these metrics overlook critical aspects of glucose dynamic homeostasis. As an alternative method, this paper explores the clinical value of glucodensity metrics in capturing glucose dynamics -- specifically the speed and acceleration of CGM time series -- as new biomarkers for predicting long-term glucose outcomes. Our results demonstrate significant information gains, exceeding 20\% in terms of adjusted $R^2$, in forecasting glycosylated hemoglobin (HbA1c) and fasting plasma glucose (FPG) at five and eight years from baseline AEGIS data, compared to traditional non-CGM and CGM glucose biomarkers. These findings underscore the importance of incorporating more complex CGM functional metrics, such as the glucodensity approach, to fully capture continuous glucose fluctuations across different time-scale resolutions.

Applications

What problem does this paper attempt to address?

### The problems the paper attempts to solve

The paper aims to explore how to use the glucodensity feature in continuous glucose monitoring (CGM) data to capture glucose kinetics, especially the velocity (rate of change) and acceleration (rate of change of the rate of change) of glucose, in order to predict long - term blood - glucose outcomes. Specifically, the authors hope to improve the prediction accuracy of fasting plasma glucose (FPG) and glycated hemoglobin (HbA1c), which are key biomarkers for diabetes control and progression, by introducing these new functional indicators.

### Background and motivation

1. **Development of CGM technology**:
- Continuous glucose monitoring (CGM) technology has significantly improved the management of type 1 diabetes, especially when combined with insulin pump systems, and can effectively reduce high - risk clinical events such as hypoglycemia.
- CGM devices are increasingly used in healthy people, pre - diabetic and diabetic patients, not only for blood - glucose control, but also for evaluating the effects of new drugs and personalized nutritional interventions.

2. **Limitations of existing CGM indicators**:
- Currently commonly used CGM indicators, such as time - in - range and blood - glucose variability measurements, are useful but fail to fully capture key aspects of glucose dynamic homeostasis.
- These traditional indicators often overlook dynamic features such as the velocity and acceleration of glucose fluctuations, which are very important for understanding blood - glucose control and predicting complications.

### Research objectives

1. **Introduce new CGM data analysis methods**:
- Based on the glucodensity method and distribution data analysis, capture glucose kinetics on different time scales, with a focus on the velocity and acceleration of glucose.

2. **Verify the predictive ability of new indicators**:
- Use the baseline data of the AEGIS study to predict FPG and HbA1c levels in the next 5 and 8 years, and evaluate the predictive performance of new indicators relative to traditional CGM and non - CGM indicators.

3. **Develop multivariate glucodensity methods**:
- Introduce multivariate glucodensity methods to simultaneously capture the distribution characteristics of glucose concentration, velocity and acceleration, and construct a unified functional profile.

### Methods

1. **Univariate glucodensity method**:
- Convert CGM data into marginal density functions and automatically integrate traditional CGM indicators.

2. **Multivariate glucodensity method**:
- Extend the glucodensity method to include the first - order derivative (velocity) and second - order derivative (acceleration) of glucose concentration, and model by smoothing techniques and kernel estimation methods.

3. **Regression modeling**:
- Use an additive scalar distribution regression model to capture the nonlinear distribution effects of glucose concentration, velocity and acceleration, and predict FPG and HbA1c levels.

### Results and significance

1. **Improvement in prediction performance**:
- The results show that, compared with traditional CGM and non - CGM indicators, the new glucodensity method has an adjusted R² value increased by more than 20% when predicting FPG and HbA1c levels.

2. **Clinical value**:
- The new method provides a more comprehensive and accurate representation of CGM data, which helps to better understand blood - glucose dynamics, optimize diabetes management and drug development.

### Conclusion

The paper extends the traditional glucodensity method by introducing new indicators of glucose velocity and acceleration, and significantly improves the ability to predict long - term blood - glucose outcomes. This method is not only of great clinical value, but also provides new tools and perspectives for future diabetes research.