Full text is available at the source.
Use of Continuous Glucose Monitoring With Machine Learning to Identify Metabolic Subphenotypes and Inform Precision Lifestyle Changes
Using Continuous Glucose Monitoring and Machine Learning to Identify Metabolic Types and Guide Personalized Lifestyle Changes
AI simplified
Abstract
Continuous glucose monitoring (CGM) may deconstruct early dysglycemia into distinct, actionable subphenotypes.
- Static glucose thresholds may obscure the heterogeneity of dysglycemia caused by insulin resistance, β-cell dysfunction, and incretin deficiency.
- Machine learning models can use high-resolution glucose data from at-home CGM tests to predict measures of muscle insulin resistance and β-cell function.
- Individuals' unique postprandial glycemic responses to specific foods, such as potatoes versus grapes, could serve as biomarkers for their metabolic subtype.
- Wearable data on diet, sleep, and physical activity patterns are associated with specific metabolic dysfunctions.
- The efficacy of dietary interventions to reduce postprandial glycemic responses appears to depend on the individual's metabolic phenotype.
AI simplified