Diabetes is a public health problem, with an estimated prevalence of 29 M people in the U.S. Over 26 M people in the U.S. suffer from chronic kidney diseases (CKD) (3,4), and diabetic nephropathy (DN) is the leading cause of CKD, accounting for 45% of incident end stage renal disease (ESRD). Among subjects with diabetes and/or CKD, morbidity and mortality is significant, with the most common cause of death due to cardiovascular disease (CVD). The ability to predict which diabetic patients will develop DN or CVD is poor. Current biomarker standards for DN progression, serum creatinine and albuminuria, are suboptimal. Serum creatinine is insensitive; microalbuminuria frequently regresses spontaneously, and diabetics often develop renal dysfunction without albuminuria. In DN diseased glomeruli leak non-esterified fatty acids, which are then exposed to the proximal tubule luminal surface. Tubular atrophy, a strong predictor of DN progression, is linked to proximal tubule reabsorption of palmitate, which leads to lipotoxicity and apoptosis. Data from a longitudinal diabetic cohort revealed that Upalm:creatinine ratio better predicts eGFR decline compared to Ualb:creatinine. The causes and risk predictors for CVD in CKD differ from those in the non-CKD population. The serum concentrations of 5 solutes (trimethylamine oxide (TMAO), symmetric and asymmetric dimethylarginine (SDMA and ADMA), p-cresol sulfate and indoxyl sulfate) rise with CKD and have been proposed as CVD predictors in the general and CKD populations. We developed targeted LC/MS assays for all 5 solutes, and found 2-40 fold greater serum levels in ESRD subjects compared to normal controls. Hypotheses: (1) Urine palmitate is an accurate predictor of eGFR decline in DN, and is a superior biomarker compared to albuminuria. (2) A panel of uremic solutes (TMAO, SDMA, ADMA, p-cresol sulfate and indoxyl sulfate) will have an additive effect to traditional Framingham risk factors for predicting cardiovascular events in a diabetic population. The hypotheses will be tested with the following specific aims:
Aim 1 : Test urine palmitate:creatinine ratio as biomarker for DN progression. The correlation between baseline Upalm:creatinine vs. Ualb:creatinine and rate of eGFR decline, will be tested in a diabetic cohort with baseline eGFR >60 using a mixed effects model and AUC analysis of ROC curves. A secondary composite outcome (serum creatinine doubling, eGFR decrease by >50%, incident ESRD), will be analyzed using a multivariate Cox proportional hazards model. Analyses will be validated in matched ACCORD trial subjects.
Aim 2 : Test a panel of 5 retained solutes as a predictor of CVD in CKD. The discrimination ability of the 5 markers (in addition to the traditional Framingham factors) will be evaluated in an ACCORD cohort with stages 1-3 CKD by computing the increase in the estimated AUC from ROC curve analysis. Computation of the integrated discrimination index will be used to assess improvement due to inclusion of the 5 markers.
Over 500,000 people in the U.S. have kidney failure that requires dialysis or transplantation to stay alive, and approximately half of these people have diabetes as a cause of renal failure. The leading cause of death in diabetic patients is cardiovascular disease. The purpose of this project is to develop more precise urine and blood tests to identify diabetic patients at risk for developing kidney or cardiovascular disease.
