Evidence for an independent role of metabolic acidosis on nutritional status in haemodialysis patients

Malnutrition in haemodialysis (HD) patients has been referred to underdialysis with low protein intake, and to acidosis. However, the separate effects of underdialysis and acidosis on nutrition have not been clearly demonstrated. To evaluate the role of the dialysis dose and of metabolic acidosis on nutrition, we measured the predialysis serum HCO3, pH, serum albumin, PCRn, Kt/V, and BMI in 81 uraemic patients on maintenance bicarbonate HD for 93+/-80 months. Patients with chronic liver diseases, malignancies, and cachexia were excluded. Mean age was 59+/-17 years, Kt/V was 1.29+/-0.21, PCRn 1.06+/-0.22 g/kg/day, serum albumin 4.07+/-0.28 g/dl, BMI 23+/-4 kg/m2, HCO3 21.1+/-1.9 mmol/l, pH 7.36+/-0.04. Serum albumin showed a significant direct correlation with: PCRn (P=0.001), HCO3 (P=0.001), pH (P=0.002), but no correlation with Kt/V and BMI. Serum HCO3 correlated inversely with PCRn (P=0.027). Multiple regression analysis confirmed the significant role of serum bicarbonate and age, but not of Kt/V, on serum albumin concentrations. The role of PCRn appeared to be marginal compared to serum bicarbonate in determining serum albumin levels. Dividing patients into two groups, serum albumin was 3.96+/-0.22 g/dl with HCO3 < or = 20 mmol/l and 4.18+/-0.31 g/dl in those with serum HCO3 > or = 23 mmol/l (P=0.002). PCRn in the same groups was respectively 1.14+/-0.24 g/kg/day and 1.01+/-0.23 g/kg/day (P=0.03). Most importantly, serum albumin levels did not appear to be affected by the dialysis dose, with Kt/V ranging from 0.90 to 1.88. In HD patients with adequate Kt/V, metabolic acidosis exerts a detrimental effect on serum albumin concentrations partially independently of the protein intake, as evaluated by PCRn. In the presence of moderate to severe metabolic acidosis, PCRn does not reflect the real dietary protein intake of the patients, probably as a result of increased catabolism of endogenous proteins. For this reason PCRn should be considered with caution as an estimate of the dietary protein intake in HD patients in the presence of metabolic acidosis.