Mitochondrial metabolites predict adverse cardiovascular events in individuals with diabetes.

Metabolic mechanisms underlying the heterogeneity of major adverse cardiovascular events (MACE) risk in individuals with type 2 diabetes mellitus (T2D) remain unclear. We hypothesized that circulating metabolites reflecting mitochondrial dysfunction predict incident MACE in T2D. Targeted mass-spectrometry profiling of 60 metabolites was performed on baseline plasma from TECOS (discovery) and EXSCEL (validation) trial biomarker substudy cohorts. A principal components analysis metabolite factor comprised of medium-chain acylcarnitines was associated with MACE in TECOS and validated in EXSCEL, with higher levels associated with higher MACE risk. Meta-analysis showed that long-chain acylcarnitines and dicarboxylacylcarnitines were also associated with MACE. Metabolites remained associated with MACE in multivariate models and favorably changed with exenatide therapy. A third cohort (CATHGEN) with T2D assessed whether these metabolites improved discriminative capability multivariate for MACE; nine metabolites (medium- and long-chain acylcarnitines and one dicarboxylacylcarnitine) were associated with time-to-MACE in CATHGEN. Addition of these metabolites to clinical models minimally improved the discriminative capability for MACE but did significantly down reclassify risk. Thus, metabolites reporting on dysregulated mitochondrial fatty acid oxidation are higher in individuals with T2D who experience subsequent MACE. These biomarkers may improve CV risk prediction models, be therapy responsive, and highlight emerging risk mechanisms.