Predictive value of malnutrition markers for mortality in peritoneal dialysis patients

Alterations in nutritional status have been described as important predictors of mortality in patients with chronic kidney disease (CKD). However, the association between multiple markers for nutritional status and the mortality rates of patients with CKD on peritoneal dialysis (PD) has not yet been illustrated in previously published data, particularly by using the new definition of protein energy wasting (PEW).

To evaluate the predictive value of malnutrition markers for mortality rates, on the basis of the PEW definition, of PD patients.

At the start of PD treatment, the nutritional status of 199 patients (mean age, 56 ± 13.3 years; 53% females) was evaluated. Body mass index (BMI), arm circumference, mid-arm muscle circumference, protein and caloric intake (by using a 3-day food record), and serum albumin were all recorded, as well as a subjective global assessment (SGA) and presence of PEW. Cut-off points were defined on the basis of the consensus of the International Society for Renal Nutrition and Metabolism (albumin, <3.8 g/dL; BMI, <23 kg/m(2); mid-arm muscle circumference, >10% in comparison with the 50th percentile for the reference population; protein intake, <0.8 g/kg/daily; caloric intake, <25 kcal/kg/daily). The data were obtained retrospectively between the years 2001 and 2008 on the basis of routine nutritional evaluation. Patients were monitored for fatal events from all possible causes.

The mean BMI for the population was 26.6 ± 5.0 kg/m(2). A median protein intake of 0.94 (0.18 to 4.57) g/kg/daily was reported and 60.3% of the patients reported a protein intake of <0.8 g/kg/daily. With respect to caloric intake, 38.7% of the patients consumed <25 kcal/kg/daily. A median of 3.5 (1.4 to 5.3) g/dL for serum albumin was observed and 29.3% of the patients presented values of <3.8 g/dL. PEW was diagnosed in 17.5% of patients. In the univariate model, being of age >65 years (P = .002), cardiovascular disease (P < .001), diabetes mellitus (P = .02), SGA (P = .02), and albumin (P = .002), were all significant markers for mortality. The presence of patients aged >65 years (P = .02), with diabetes mellitus (P = .057), cardiovascular disease (P = .005), and albumin were considered as independent factors for mortality in this study.

SGA, albumin, and PEW were the only nutritional markers found to be associated with mortality in this cohort of PD patients. In the multivariate analysis, after adjusting for classic mortality risk factors, only patients with hypoalbuminemia were found to be at a high risk for mortality at follow-up. These results may be limited by the number of observations and a necessity for confirmation in larger prospective studies.