Objective: Patients with diabetes who manifest proteinuria are at increased risk for cardiovascular events. Some studies suggest that proteinuria exerts its cardiovascular effects at least partly through a positive association with total plasma homocysteine (tHcy). Modestly sized but better designed contrary studies find no such link through a limited range of serum creatinine and proteinuria. We tested the hypothesis that proteinuria independently predicts tHcy levels in a larger cohort of type 2 diabetic patients with nephropathy throughout a much broader range of kidney disease and proteinuria.
Research design and methods: Baseline data for the cross-sectional study were obtained from 717 patients enrolled in the multicenter Irbesartan Diabetic Nephropathy Trial. All subjects had type 2 diabetes, hypertension, and proteinuria and were between 29 and 78 years of age. Data included age, sex, BMI, serum creatinine and albumin, LDL and HDL cholesterol, triglyceride, proteinuria and albuminuria, plasma folate, B12, and pyridoxal 5'-phosphate (PLP) (the active form of B6), HbA(1c), and tHcy levels. Unadjusted and multivariable models were used in the analysis.
Results: Crude analyses revealed significant associations between tHcy and age (r = 0.074; P = 0.008), creatinine (r = 0.414; P < 0.001), PLP (r = -0.105; P = 0.021), B12 (r = -0.216; P < 0.001), folate (r = -0.241; P < 0.001), and HbA(1c) (r = -0.119; P = 0.003), with serum albumin approaching significance (r = 0.055; P = 0.072). Only serum creatinine, plasma folate, B12, serum albumin, sex, HbA(1c), and age were independent predictors of tHcy after controlling for all other variables.
Conclusions: By finding no independent correlation between proteinuria (or albuminuria) and tHcy levels, this study improves the external validity of previous negative findings. Therefore, it is unlikely that the observed positive association between proteinuria and cardiovascular disease is directly related to hyperhomocysteinemia.