부산대학교 도서관

Introduction: Impaired fatty acid oxidation is common to both kidney and CV complications of diabetes mellitus (DM). We compared metabolic profiles of different cardiorenal trajectories to identify common pathways and unique signatures which may inform disease mechanisms.Methods The EXSCEL trial randomized patients with type 2 DM to exenatide vs placebo. For this study, we performed targeted metabolomics of 60 metabolites (45 acylcarnitines [ACs], 15 amino acids, ketone bodies, and nonesterified fatty acids) in 978 EXSCEL participants. Principal components analysis (PCA) was used for dimensionality reduction and multivariable logistic regression to examine the association of each PCA factor with four cardiorenal outcomes: CV (CV death, MI, or stroke), renal (≥30% eGFR decline, ESKD, or renal death), cardiorenal (both outcomes), or neither. We also evaluated the factors associations with hospitalization for heart failure (HHF) and new-onset albuminuria, adjusting for baseline CKD, medication, and treatment allocation.Results PCA metabolite factors and cardiorenal outcome associations are listed in Table 1. Considering the directionality of factor loadings, baseline levels of small-chain dicarboxylated ACs and medium-chain ACs were associated with worse cardiorenal outcomes. Long-chain ACs were associated with worse renal outcomes. In contrast, 12-month change in medium-chain saturated ACs and long-chain unsaturated ACs were associated with better CV and renal outcomes, respectively. No factors associated with albuminuria. Baseline nonesterified fatty acids (OR 0.64, 95% CI 0.42-0.99) and 12-month change in medium-chain ACs (OR 1.57, 95% CI 1.04-2.37) associated with HHF.Conclusions Baseline associations between ACs and adverse cardiorenal outcomes further link mitochondrial dysfunction with both disorders. However, the association between changes in some ACs with improved outcomes may suggest an adaptive role for these metabolites.