Differential metabolomic signatures of declining renal function in Types 1 and 2 diabetes. Issue 10 (30th September 2020)

• Record Type: Journal Article
• Authors: Manca, Maria Laura; Solini, Anna; Haukka, Jani K; Sandholm, Niina; Forsblom, Carol; Groop, Per-Henrik; Ferrannini, Ele
• Is Part Of: Nephrology dialysis transplantation. Volume 36:Issue 10(2021)
• Page Start: 1859
• Page End: 1866
• Publication Date: 2020-09-30
• Subjects: GFR -- metabolomics -- nephropathy -- Type 1 diabetes -- Type 2 diabetes; Nephrology -- Periodicals; Hemodialysis -- Periodicals; Kidneys -- Transplantation -- Periodicals; Hemodialysis; Kidneys -- Transplantation; Nephrology; Periodicals; 616.61
• DOI: 10.1093/ndt/gfaa175
• Languages: English
• ISSNs: 0931-0509

Abstract

Abstract: Background: Chronic kidney disease (CKD) shows different clinical features in Types1 (T1D) and 2 diabetes (T2D). Metabolomics have recently provided useful contribution to the identification of biomarkers of CKD progression in either form of the disease. However, no studies have so far compared plasma metabolomics between T1D and T2D in order to identify differential signatures of progression of estimated glomerular filtration rate (eGFR) decline. Methods: We used two large cohorts of T1D (from Finland) and T2D (from Italy) patients followed up to 7 and 3 years, respectively. In both groups, progression was defined as the top quartile of yearly decline in eGFR. Pooled data from the two groups were analysed by univariate and bivariate random forest (RF), and confirmed by bivariate partial least squares (PLS) analysis, the response variables being type of diabetes and eGFR progression. Results: In progressors, yearly eGFR loss was significantly larger in T2D [−5.3 (3.0), median (interquartile range)mL/min/1.73 m 2 /year] than T1D [−3.7 (3.1) mL/min/1.73 m 2 /year ; P = 0.018]. Out of several hundreds, bivariate RF extracted 22 metabolites associated with diabetes type (all higher in T1D than T2D except for 5-methylthioadenosine, pyruvate and β-hydroxypyruvate) and 13 molecules associated with eGFR progression (all higher in progressors than non-progressors except for sphyngomyelin). Three of the selected metabolites (histidylphenylalanine, leucylphenylalanine, tryptophylasparagine) showed a significant interaction between disease type and progression. Only eight metabolites were common to both bivariate RF and PLS. Conclusions: Identification of metabolomic signatures of CKD progression is partially dependent on the statistical model. Dual analysis identified molecules specifically associated with progressive renal impairment in both T1D and T2D.