What does this research mean for the field?

Continuous glucose monitoring (CGM) variability metrics, specifically the mean amplitude of glycemic excursions, are associated with distinct metabolic pathways and metabolites that are not captured by traditional glycemic measures like fasting glucose or A1c. Novelty: ClaimNovelty.NOVEL_FINDING. Consensus alignment: ConsensusAlignment.ESTABLISHES_NEW_DIRECTION.

What question did this study set out to answer?

This research aims to characterize the metabolomic profiles associated with continuous glucose monitoring (CGM) variability and its relationship with traditional glycemic measures.

June 7, 2026

1918-P: MWAS of CGM Variability Reveals Distinct Patterns of Both Known and Unknown Metabolites beyond Traditional Glycemic Measures

Key Points

This research aims to characterize the metabolomic profiles associated with continuous glucose monitoring (CGM) variability and its relationship with traditional glycemic measures.
Conducted a MWAS of 4145 metabolites among 1379 participants with and without T2D.
Adjusted for age, sex, and race/ethnicity, utilizing separate models for T2D and non-diabetes.
Compared identified metabolites to traditional measures like fasting glucose and A1C.
Identified 129 metabolites significantly associated with at least one CGM trait (FDR-corrected P<.05).
Of these, 47 metabolites showed a significant association with mean amplitude of glycemic excursions (MAGE) but not with fasting glucose or A1C.
Highlights the presence of both known and unidentified metabolites, suggesting novel physiological pathways related to dysglycemia.

Abstract

Introduction and Objective: A reliance on measures such as fasting glucose (FG) and A1c has left many aspects of dysglycemia, like variability, poorly characterized. We leveraged CGM for new insight into metabolic health and metabolomics as a connection to mechanistic pathways, a combination rarely available in diverse population-based studies. Methods: We conducted an MWAS (n=4145 metabolites) of CGM traits among 1379 participants with and without T2D in the Multi-Ethnic Study of Atherosclerosis. We adjusted for age, sex, and race/ethnicity, used separate models for T2D and non-diabetes, and compared to FG and A1C results. Results: Among 129 metabolites significantly associated (FDR-corrected P.05) with at least one CGM trait, 47 metabolites were significantly associated with mean amplitude of glycemic excursions (MAGE) that were NOT associated with FG or A1c. While many of these metabolites are recognized, a considerable number remain unidentified, suggesting novel pathways separate from known clinical metabolomics. Next steps include investigation of metabolomic overlap between CGM metrics and MWAS with more novel CGM metrics. Conclusion: These results showing overlapping and unique metabolite signals between clinical biomarkers and MAGE from CGM across the glucose range strongly indicates the potential of CGM to identify physiological pathways to dysglycemia and T2D risk as novel targets for prevention. Disclosure B. Bakhshi: None. J. Yao: None. S. Chen: None. H. Chen: None. C.B. Clish: None. X. Guo: None. H. Lin: None. P.Y. Liu: None. C. Schaich: None. N. Spartano: Research Support; Current; Dexcom, Inc. J. Tanley: None. M.E. Walker: None. A. Wood: Research Support; Current; Beef Checkoff. Consultant; Ended; Lundquist Institute for Biomedical Innovation. J.I. Rotter: None. M. Mongraw-Chaffin: None. Funding NIA (R01AG070881)

Mark Helpful

Bookmark

Relay