Association of Uremic Solutes With Cardiovascular Death in Diabetic Kidney Disease.

RATIONALE AND OBJECTIVE Cardiovascular disease (CVD) is a major cause of mortality among people with diabetic kidney disease (DKD). The pathophysiology is inadequately explained by traditional CVD risk factors. Three uremic solutes, trimethylamine-N-oxide (TMAO), asymmetric and symmetric dimethylarginine (ADMA, SDMA), have been linked to CVD in kidney failure with renal replacement therapy (KFRT), but data are limited in populations with diabetes and less severe kidney disease. STUDY DESIGN Observational cohort. SETTINGS AND PARTICIPANTS Random subcohort of 555 REGARDS Study participants with diabetes and eGFR <60 ml/min/1.73m2 at study entry. EXPOSURES ADMA, SDMA and TMAO were assayed by liquid chromatography-mass spectrometry in plasma and urine. OUTCOMES CV mortality (primary outcome), all-cause mortality and incident KFRT (secondary outcomes). ANALYTIC APPROACH Plasma and urine:plasma (U/P) ratios of ADMA, SDMA and TMAO were tested for association with outcomes. Adjusted Cox regression models were fitted and hazard ratios (HR) of outcomes presented per standard deviation, log2 increments and interquartile comparisons. RESULTS Mean baseline eGFR was 44 ml/min/1.73 m2. CV death, overall mortality and KFRT occurred in 120, 285 and 89 participants during mean 6.2 years follow-up. Higher plasma ADMA, and lower U/P ratios of ADMA, SDMA and TMAO were associated with CV mortality by interquartile (HR 1.9-3.9), log2 concentration comparisons (HR 1.4-2.1), and per standard deviation (HR 1.2). Higher plasma concentrations (HR 1.1-2.8) and lower U/P ratios (HR 1.3-2.3) of all three solutes were associated with all-cause mortality. Higher plasma SDMA was associated with incident KFRT (HR 1.2-3.0). LIMITATIONS Single cohort, restricted to diabetic patients with eGFR <60, potential residual confounding by GFR, no dietary information. CONCLUSIONS Higher plasma concentrations and lower U/P ratios of uremic solutes were independently associated with CV and all-cause mortality in DKD. Associations of U/P ratios with mortality suggest a connection between renal uremic solute clearance and CVD pathogenesis.