Beyond Scalar Metrics: Functional Data Analysis of Postprandial Continuous Glucose Monitoring in the AEGIS Study

Area of Science:

• Metabolic Health
• Data Science
• Personalized Nutrition

Background:

• Continuous Glucose Monitoring (CGM) provides valuable metabolic data but presents analytical challenges due to high within-person variability and multilevel structure.
• Existing analytical methods struggle with the complexity of CGM data, limiting personalized diet prescription.
• Understanding glucose response to meals is crucial for metabolic assessment in individuals without diabetes.

Purpose of the Study:

• To develop and apply a novel analytical framework for CGM data to examine glucose meal responses.
• To explore time-dependent associations between dietary intake, patient characteristics, and glucose fluctuations.
• To improve personalized diet recommendations by analyzing the full functional domain of glucose responses.

Main Methods:

• Utilized a multilevel functional modeling approach to analyze CGM data from the AEGIS study.
• Incorporated a new functional mixed R-square coefficient for model evaluation.
• Examined meal timing, nutrition, and patient characteristics (normoglycemic vs. prediabetic) in relation to glucose profiles.

Main Results:

• Demonstrated the significance of analyzing glucose responses across the entire functional domain for effective diet recommendations.
• Identified differential metabolic responses between normoglycemic and prediabetic individuals, particularly concerning lipid intake.
• Highlighted the necessity of including random, person-level effects in modeling CGM data for accurate scientific insights.

Conclusions:

• Multilevel functional models offer a robust framework for analyzing complex CGM data.
• Personalized metabolic responses to dietary intake vary significantly between individuals and glycemic status.
• Accurate modeling of CGM data requires accounting for individual variability through person-level effects to inform personalized nutrition strategies.