Differences in dietary intake assessed by dietary recall and that estimated by protein nitrogen appearance rate

Peritoneal dialysis (PD) patients are at risk of muscle wasting and clinical guidelines recommend assessing dietary intake, by calculating protein equivalent of nitrogen appearance (PNA) to assure adequate intake. As the PNA equations were developed some time ago based on continuous ambulatory PD (CAPD) treatments, we re-evaluated them by comparing dietary recall protein intake. Dietary histories were obtained from 42 PD patients, mean age 62.2 ± 15.8 years, 58.1% male, 41.9% white, 39.5% treated with CAPD, median dialysis vintage 13.2 (3.1-23.3) months when attending for routine outpatient peritoneal membrane testing. Dietary protein intake (DPI), median 50.8 (37.5-71.5) g/day was determined with the Nutrics software program^®, and using PNA equations, DPI varied from 32.6 (31-34.3) to 68.8 (58.7-86.3) g/day. Bland Altman bias ranged from -22.8 ± 23.4 to +12.0 ± 24.6 g/day, with the older equations tending to underestimate and the more recent equations overestimating dietary recall DPI. There was systematic bias with the older equations, particularly as DPI increased. Although there was a wide variation, the two equations exhibiting least variation with dietary DPI were: PNA=15.1 + (0.195 × 24-h combined urinary and peritoneal urea mmol) + (24-h combined urinary and peritoneal protein loss), and PNA = 20.1 + (0.209 × 24-h combined urinary and peritoneal urea mmol), with a mean bias of +6.3 and +7.8 g/day compared to dietary DPI, and negative nitrogen balance of -1.7 and 2.6 g/day, respectively. Our study would suggest that the most used PNA equations may overestimate DPI in many patients and potentially result in unrecognised protein malnutrition.