Context: Insulin resistance reflects the inadequate insulin-mediated use of metabolites and predicts type 2 diabetes (T2D) but is also frequently seen in long-standing type 1 diabetes (T1D) and represents a major cardiovascular risk factor.
Objective: We hypothesized that plasma metabolome profiles allow the identification of unique and common early biomarkers of insulin resistance in both diabetes types.
Design, setting, and patients: Two hundred ninety-five plasma metabolites were analyzed by mass spectrometry from patients of the prospective observational German Diabetes Study with T2D (n = 244) or T1D (n = 127) and known diabetes duration of less than 1 year and glucose-tolerant persons (CON; n = 129). Abundance of metabolites was tested for association with insulin sensitivity as assessed by hyperinsulinemic-euglycemic clamps and related metabolic phenotypes.
Main outcomes measures: Sixty-two metabolites with phenotype-specific patterns were identified using age, sex, and body mass index as covariates.
Results: Compared with CON, the metabolome of T2D and T1D showed similar alterations in various phosphatidylcholine species and amino acids. Only T2D exhibited differences in free fatty acids compared with CON. Pairwise comparison of metabolites revealed alterations of 28 and 49 metabolites in T1D and T2D, respectively, when compared with CON. Eleven metabolites allowed differentiation between both diabetes types and alanine, α-amino-adipic acid, isoleucin, and stearic acid showed an inverse association with insulin sensitivity in both T2D and T1D combined.
Conclusion: Metabolome analyses from recent-onset T2D and T1D patients enables identification of defined diabetes type-specific differences and detection of biomarkers of insulin sensitivity. These analyses may help to identify novel clinical subphenotypes diabetes.
Trial registration: ClinicalTrials.gov NCT01055093.