Diabetes (DM) can be classified into 5 subtypes characterized by distinct progression in dysglycaemia and complications. Using 5 clinical variables, we categorized 7017 participants from the Outcome Reduction with an Initial Glargine Intervention (ORIGIN) trial into 1/auto-immune DM (n=241) , 2/insulin-deficient DM (n=1594) , 3/insulin-resistant DM (n=914) , 4/obesity-related DM (n=1595) , 5/age-related DM (n=2673) .Yet, whether blood biomarkers are associated with these subtypes is unknown. Forward-selection logistic regression models were used to identify biomarkers that were each independent determinant of one cluster versus the others, among 233 selected cardiometabolic proteins measured at baseline. Models were adjusted for age, sex, ethnicity, C-peptide level, diabetes duration. A total of 13, 2, 7 and biomarkers were independent determinants of DM subtypes 2 to 5 respectively (all P<4.3x10-5) . A combination of 5 biomarkers that were distinctively associated with clusters (fig) , showed a diagnosis performance, through AUC-ROC curves, of 0.71, 0.86, 0.88, 0.82 to respectively distinguish cluster 2 to 5 from the others. No biomarkers other than GAD antibodies were determinants of cluster 1. We identified 5 serum biomarkers, as independent determinants of DM subtypes, that could be used as a diagnosis test for DM subtyping. Although this requires further validation in an independent population.


M.Pigeyre: n/a. H.C.Gerstein: Advisory Panel; Abbott, Eli Lilly and Company, Hanmi Pharm. Co., Ltd., Novo Nordisk, Pfizer Inc., Sanofi, Viatris Inc., Consultant; Kowa Company, Ltd., Other Relationship; DKSH, Eli Lilly and Company, Sanofi, Zuellig Pharma Holdings Pte. Ltd., Research Support; AstraZeneca, Eli Lilly and Company, Merck & Co., Inc., Novo Nordisk, Sanofi. L.Groop: None. S.Hess: Employee; Sanofi. G.Pare: Advisory Panel; Amgen Inc., Bayer AG, Sanofi, Research Support; Bayer AG.


NCT00069784 Canadian Institute of Health Research Sanofi

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.