Abstract

A novel clustering approach identified five subgroups of diabetes with distinct progression trajectories of complications. We hypothesized that these subgroups differ in multiple biomarkers of inflammation. Serum levels of 74 biomarkers of inflammation were measured in 414 individuals with recent adult-onset diabetes from the German Diabetes Study (GDS) allocated to five subgroups based on data-driven analysis. Pairwise differences between subgroups for biomarkers were assessed with generalized linear mixed models before (model 1) and after adjustment (model 2) for the clustering variables. Participants were assigned to five subgroups: severe autoimmune diabetes (SAID, 21%), severe insulin-deficient diabetes (SIDD, 3%), severe insulin-resistant diabetes (SIRD, 9%), mild obesity-related diabetes (MOD, 32%) and mild age-related diabetes (MARD, 35%). In model 1, 23 biomarkers showed ≥1 pairwise difference between subgroups (Bonferroni-corrected p<0.0007). Biomarker levels were generally highest in SIRD and lowest in SIDD. All 23 biomarkers correlated with ≥1 of the clustering variables. In model 2, three biomarkers (CASP-8, EN-RAGE, IL-6) showed at least one pairwise difference between subgroups (e.g. lower CASP8, EN-RAGE and IL-6 in SIDD vs. all other subgroups, all p<0.0007). Thus, novel diabetes subgroups show multiple differences in biomarkers of inflammation, underlining a prominent role of inflammatory pathways in particular in SIRD.