Abstract MP32: Associations of Metabolites With Glycemic Outcomes Among Bogalusa Heart Study and Atherosclerosis Risk in Communities Study Participants

Background: Metabolomics may identify novel biomarkers for predicting, detecting, and managing diabetes. We evaluated the association of novel metabolites with diabetes, prediabetes, and continuous glycemic outcomes (fasting plasma glucose (FPG); hemoglobin A1c (HbA1c)) using data from the Bogalusa Heart Study (BHS) and Atherosclerosis Risk in Communities Study (ARIC). Methods: Untargeted ultrahigh-performance liquid chromatography-tandem mass spectroscopy was used to detect and quantify 1466 serum metabolites in 1194 BHS participants (2013-2016) with and without diabetes (defined as FPG ≥126 mg/dL or use of diabetes medication; prediabetes as FPG of 100 to <126 mg/dL). We assessed cross-sectional associations between all metabolites with FPG and HbA1c (linear regression) and with prediabetes and diabetes (multinomial logistic regression). Significant and novel metabolite associations in BHS were tested for replication (determined by statistical significance and consistent effect direction) among 1528 participants from ARIC (1990-1992). All analyses used Bonferroni-corrected alpha thresholds. Results: In BHS (mean age 48.1 y, 60% female), 16% of persons had diabetes and 32% had prediabetes. Mean FPG and HbA1c were 106.5 mg/dL and 5.9%, respectively. Among 163 novel glycemic-metabolite associations in BHS, 21 metabolites were tested for replication in ARIC (measured on a different platform): Six metabolites for diabetes, 9 for glucose, 12 for HbA1c, and 3 for prediabetes. Among metabolites tested, replication results for All (Black and White) and Black participants across ARIC subgroups identified glycylvaline as statistically significant for novel association replicated with HbA1c (Table). Conclusions: A significant novel metabolite association of glycylvaline, part of the dipeptide subpathway, with HbA1c was identified in BHS and replicated in ARIC. This dipeptide subpathway should be further explored as a potential subpathway to target for detection or management of hyperglycemia.