From Glucose Patterns to Health Outcomes: A Generalizable Foundation Model for Continuous Glucose Monitor Data Analysis
🤖 AI Summary
To address challenges in continuous glucose monitoring (CGM) data modeling—including poor generalizability, limited expressiveness of static clinical metrics, and difficulty capturing complex temporal dynamics—this study introduces GluFormer, the first generative foundation model specifically designed for CGM time-series data. Trained via large-scale self-supervised pretraining on millions of CGM records from non-diabetic adults, GluFormer employs an autoregressive token prediction architecture tailored to physiological glucose dynamics. It enables robust cross-population generalization across ethnicities, countries, CGM devices, and pathological states; supports multimodal CGM-diet representation learning; and facilitates intervention simulation. Evaluated on 19 external cohorts, GluFormer achieves 66% sensitivity in identifying incident diabetes cases within the top quartile of risk scores (vs. 7% in the bottom quartile) and concentrates 69% of cardiovascular mortality events in the top quartile (0% in the bottom). Its dietary-augmented variant accurately generates personalized CGM trajectories and predicts individualized postprandial glycemic responses.
Application Category
📝 Abstract
Recent advances in SSL enabled novel medical AI models, known as foundation models, offer great potential for better characterizing health from diverse biomedical data. CGM provides rich, temporal data on glycemic patterns, but its full potential for predicting broader health outcomes remains underutilized. Here, we present GluFormer, a generative foundation model for CGM data that learns nuanced glycemic patterns and translates them into predictive representations of metabolic health. Trained on over 10 million CGM measurements from 10,812 adults, primarily without diabetes, GluFormer uses autoregressive token prediction to capture longitudinal glucose dynamics. We show that GluFormer generalizes to 19 external cohorts (n=6,044) spanning different ethnicities and ages, 5 countries, 8 CGM devices, and diverse pathophysiological states. GluFormers representations exceed the performance of current CGM metrics, such as the Glucose Management Indicator (GMI), for forecasting clinical measures. In a longitudinal study of 580 adults with CGM data and 12-year follow-up, GluFormer identifies individuals at elevated risk of developing diabetes more effectively than blood HbA1C%, capturing 66% of all new-onset diabetes diagnoses in the top quartile versus 7% in the bottom quartile. Similarly, 69% of cardiovascular-death events occurred in the top quartile with none in the bottom quartile, demonstrating powerful risk stratification beyond traditional glycemic metrics. We also show that CGM representations from pre-intervention periods in Randomized Clinical Trials outperform other methods in predicting primary and secondary outcomes. When integrating dietary data into GluFormer, we show that the multi-modal version of the model can accurately generate CGM data based on dietary intake data, simulate outcomes of dietary interventions, and predict individual responses to specific foods.
Problem
Research questions and friction points this paper is trying to address.
Continuous Glucose Monitoring (CGM)
Metabolic Health
Dietary Impact on Blood Sugar
Innovation
Methods, ideas, or system contributions that make the work stand out.
GluFormer
Continuous Glucose Monitoring (CGM)
Metabolic Health Prediction
🔎 Similar Papers
No similar papers found.
