Relating protein intake to nutritional status in haemodialysis patients: how to normalize the protein equivalent of total nitrogen appearance (PNA)?

Abstract
Background. The protein equivalent of total nitrogen appearance (PNA) is assumed to be a reliable estimate of dietary protein intake in haemodialysis patients. Protein requirements are related to body size. In order to standardize PNA to individual differences in body size, PNA is normalized to various terms related to the patient's body weight. It is not clear which is the most appropriate method to normalize PNA. Methods. We calculated five commonly used variants of normalized PNA and related them to indices of nutritional status in 57 stable chronic haemodialysis patients, 57±15 (mean±SD) years of age. PNA, determined by direct dialysate quantification, was normalized to actual post-dialysis dry body weight (DBW), normal body weight (DBWnormal), lean body mass (LBM), normal lean body mass (LBMnormal), and `normalized' body weight (N). Nutritional status was assessed using an index of nutrition composed of anthropometry derived parameters and plasma albumin concentration. Results. PNADBW (0.85±0.14 g/kg/d) tended to be higher than PNADBWnormal (0.81±0.14 g/kg/d). PNALBM (1.17±0.19 g/kg/d) did not differ from PNALBMnormal (1.19±0.21 g/kg/d). PNAN (1.06± 0.14 g/kg/d) was significantly higher than PNADBW and PNADBWnormal, but lower than PNALBM and PNALBMnormal. Actual PNA (61±13 g/d) correlated significantly with DBW (r=0.52) and LBM (r=0.63) indicating that large patients eat more protein. Interestingly, actual PNA correlated with plasma albumin (r=0.33) and with the overall index of nutrition (r=0.27) as well. PNADBW correlated negatively with relative DBW (r=−0.32), expressed as a percentage of normal values, indicating that PNADBW is relatively high in underweight patients. In contrast, PNADBWnormal correlated positively with all nutritional parameters as well as with the overall index of nutrition (r=0.33). PNAN and PNALBM did not correlate with the nutritional status, but PNALBMnormal correlated positively with relative DBW (r=0.50) and with overall nutritional status (r=0.34). PNADBWnormal and PNALBMnormal in well-nourished patients showed overlap with the values in patients with evident malnutrition, despite the positive correlation of the normalized PNA values with nutritional status. Conclusions. Normalizing PNA by DBWnormal and LBMnormal appeared to be the most appropriate method to standardize protein intake in haemodialysis patients. Since actual PNA is the purest estimate of protein intake that correlated with nutritional status, we recommend to evaluate actual PNA as well in studies that relate protein intake to patient outcome.