Background: Low serum bicarbonate level has been reported to be an independent predictor of kidney function decline and mortality in patients with chronic kidney disease. Mechanisms underlying low serum bicarbonate levels may differ in patients with and without diabetes. We aimed to specifically investigate the association of serum bicarbonate level with kidney disease progression and cardiovascular outcome in a cohort of patients with type 2 diabetes and nephropathy.
Study Design: Post hoc analysis of 2 multicenter randomized controlled trials.
Setting & Participants: 2,628 adults with type 2 diabetes and nephropathy.
Factor: Serum bicarbonate level.
Outcomes: Incidence of: (1) end-stage renal disease (ESRD), (2) ESRD or doubling of serum creatinine level, (3) all-cause mortality, (4) cardiovascular events (fatal/nonfatal stroke/myocardial infarction), and (5) heart failure.
Measurements: Serum bicarbonate was measured at baseline as total carbon dioxide. Associations of baseline serum bicarbonate level with end points were investigated using Cox regression models. Serum bicarbonate levels were studied as a continuous variable and stratified in quartiles. Follow-up was 2.8 +/- 1.0 (SD) years.
Results: Cox regression analyses showed that serum bicarbonate level had inverse associations with incident ESRD (HR, 0.91; 95% CI, 0.89-0.93; P <0.001) and incidence of the combined end point of ESRD or serum creatinine doubling (HR, 0.94; 95% CI, 0.92-0.96; P <0.001). These associations were independent of age, sex, and cardiovascular risk factors, but disappeared after adjustment for baseline estimated glomerular filtration rate (all P > 0.05). Analysis of bicarbonate quartiles showed similar results for the quartile with the lowest bicarbonate (
Limitations: Post hoc analysis and single measurement of serum bicarbonate.
Conclusions: In this cohort of patients with type 2 diabetes with nephropathy, serum bicarbonate level associations with kidney disease end points were not retained after adjustment for estimated glomerular filtration rate, which is in contrast to results of earlier studies in nondiabetic populations. (C) 2015 by the National Kidney Foundation, Inc.