期刊
DIABETOLOGIA
卷 55, 期 2, 页码 499-508出版社
SPRINGER
DOI: 10.1007/s00125-011-2339-6
关键词
Low eGFR; Metabolomics; Non-proteinuric; Type 2 diabetes; Uraemic toxin
资金
- NUS
- NUS Life Sciences (LSI)
- NUS Environmental Research Institute (NERI)
We carried out a urinary metabolomic study to gain insight into low estimated GFR (eGFR) in patients with non-proteinuric type 2 diabetes. Patients were identified as being non-proteinuric using multiple urinalyses. Cases (n = 44) with low eGFR and controls (n = 46) had eGFR values < 60 and a parts per thousand yen60 ml min(-1) 1.73 m(-2), respectively, as calculated using the Modification of Diet in Renal Disease formula. Urine samples were analysed by liquid chromatography/mass spectrometry (LC/MS) and GC/MS. False discovery rates were used to adjust for multiple hypotheses testing, and selection of metabolites that best predicted low eGFR status was achieved using least absolute shrinkage and selection operator logistic regression. Eleven GC/MS metabolites were strongly associated with low eGFR after correction for multiple hypotheses testing (smallest adjusted p value = 2.62 x 10(-14), largest adjusted p value = 3.84 x 10(-2)). In regression analysis, octanol, oxalic acid, phosphoric acid, benzamide, creatinine, 3,5-dimethoxymandelic amide and N-acetylglutamine were selected as the best subset for prediction and allowed excellent classification of low eGFR (AUC = 0.996). In LC/MS, 19 metabolites remained significant after multiple hypotheses testing had been taken into account (smallest adjusted p value = 2.04 x 10(-4), largest adjusted p value = 4.48 x 10(-2)), and several metabolites showed stronger evidence of association relative to the uraemic toxin, indoxyl sulphate (adjusted p value = 3.03 x 10(-2)). The potential effect of confounding on the association between metabolites was excluded. Our study has yielded substantial new insight into low eGFR and provided a collection of potential urinary biomarkers for its detection.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据