DOI: 10.2337/db23-188-or ISSN: 0012-1797

188-OR: Circulating Metabolites and Type 2 Diabetes (T2D) Risk—An Integrated Metabolomics and Genetics Study of ~23,000 Adults

JUN LI, JIE HU, LIMING LIANG, MARTA GUASCH, JORDI MERINO, MIGUEL RUIZ-CANELA, KAI LUO, CASEY REBHOLZ, BIANCA PORNEALA, JOSEE DUPUIS, ELIZABETH SELVIN, SHILPA BHUPATHIRAJU, JENNIFER A. BRODY, YONGMEI LIU, A. HEATHER ELIASSEN, JOANN E. MANSON, CLARY B. CLISH, ROZENN N. LEMAITRE, KATHERINE L. TUCKER, JEROME I. ROTTER, MIGUEL A. MARTINEZ-GONZALEZ, KATHRYN M. REXRODE, JAMES B. MEIGS, ERIC BOERWINKLE, ROBERT KAPLAN, FRANK HU, BING YU, QIBIN QI, THE NHLBI TOPMED METABOLOMICS AND PROTEOMICS WORKING GROUP
  • Endocrinology, Diabetes and Metabolism
  • Internal Medicine

Metabolome captures complex states of metabolic homeostasis affected by genetic and environmental factors. A systematic understanding of metabolites associated with T2D risk may offer potentials for personalized risk monitoring and prevention.

Here, we integrated data for 467 harmonized circulating metabolites, genetics, and diet/lifestyle factors in up to 23,616 racially diverse individuals from 10 cohorts with up to 26-y of follow-up. Through multivariable-adjusted metabolome-wide association analyses, we identified 235 metabolites associated with incident T2D (4,000 incident cases; FDR<0.05). Genetic determinants of these metabolites were mapped to genes implicated in pathways of nutrient/energy metabolism and glucose/lipid dysregulation (pathway enrichment FDR<0.05) and highly expressed in T2D-relevent tissues e.g., liver, adipose tissues, pancreas, and muscles (tissue-specific expression enrichment FDR<0.05). Mendelian randomization analyses further corroborated potential causal associations between 93 of the 235 metabolites and T2D. These included previously known (e.g., valine and isoleucine) and novel associations (e.g., α-ketoisocaproate in leucine metabolism, N-acetylaspartate in glutamate metabolism, and odd-chain fatty acid heptadecanoate). Furthermore, we analyzed metabolome-wide associations with 18 mutually adjusted modifiable risk factors. Several factors e.g., physical activity and whole grain (WG) intake, explained higher proportions of inter-individual variation of T2D-associated vs. other metabolites. We identified specific metabolites as potential mediators linking each risk factor to T2D (e.g., 4 metabolites mediated 22% of the WG-T2D association).

In conclusion, we found known and novel metabolites associated with T2D risk, indicating potential biological pathways through which genetic and lifestyle factors contribute to T2D. Our findings may inform future personalized T2D prevention.

Disclosure

J.Li: None. J.Dupuis: None. E.Selvin: None. S.Bhupathiraju: None. J.A.Brody: None. Y.Liu: None. A.Eliassen: None. J.E.Manson: None. C.B.Clish: None. R.N.Lemaitre: None. K.L.Tucker: None. J.Hu: None. J.I.Rotter: None. M.A.Martinez-gonzalez: None. K.M.Rexrode: None. J.B.Meigs: Consultant; Quest Diagnostics. E.Boerwinkle: None. R.Kaplan: None. F.Hu: None. B.Yu: None. Q.Qi: None. N.The nhlbi topmed metabolomics and proteomics worki: n/a. L.Liang: None. M.Guasch: None. J.Merino: None. M.Ruiz-canela: None. K.Luo: None. C.Rebholz: None. B.Porneala: None.

Funding

National Institutes of Health (R00DK122128 to J.L.)

More from our Archive