Abstract
Several observational studies have reported an association between higher serum bicarbonate level and high mortality risk in dialysis patients. However, in such studies mere discovery of associations does not allow one to infer causal relationships. This association may be related to inadequate dietary protein intake that may lead to less acid generation and hence a higher serum bicarbonate level. Since undernutrition is a strong predictor of death in hemodialysis patients, the observed association may be an epiphenomenon and not a biologically plausible relationship. Higher protein and fluid intake between two subsequent hemodialysis treatments may lead to lower serum bicarbonate level. This low bicarbonate level may appear protective, as patients with higher food intake and better appetite generally exhibit greater survival. In the contemporary three-stream proportioning system of hemodialysis treatment, the bicarbonate concentrate is separate from the acid concentrate, and the contribution of the acid concentrate organic acid (acetate, citrate or diacetate) to the delivered bicarbonate pool of the patient is negligible. The concept of 'total buffer' that assumes that the combination of bicarbonate and acetate concentrations in the dialysate are added equally as bicarbonate equivalents is likely wrong and based on the misleading notion that the acetate of the acid concentrate is fully metabolized to bicarbonate in the dialysate. Given these uncertainties it is prudent to avoid excessively high or low bicarbonate levels in dialysis patients.