Protein-energy wasting, as well as overweight and obesity, is a long-term risk factor for mortality in chronic hemodialysis patients

Abstract

In patients with end-stage renal disease on chronic hemodialysis (HD), protein-energy wasting (PEW) is very common and is associated with increased morbidity and mortality. Evaluation of nutritional status should be performed regularly in all such patients, using multiple methods. In this study, we analyzed the influence of several nutritional markers on long-term (5 years) survival of HD patients in one center. This is the first study on the long-term influence of nutritional status on mortality in dialysis patients ever conducted in Romania. We included all prevalent HD patients in our center. Those with ongoing acute illnesses and with inflammation (C-reactive protein ≥ 6.0 mg/l) were excluded. In the remaining subjects (N = 149, 82 males, mean age 55 years old), we performed the following measurements of nutritional status: estimation of dietary protein intake by normalized protein equivalent of nitrogen appearance (nPNA), subjective global assessment (SGA), body mass index (BMI), tricipital skinfold thickness, mid-arm circumference, assessment of body composition by bioelectrical impedance analysis (BIA), and pre-dialysis serum creatinine, albumin, and total cholesterol. We used receiver operating characteristic curves to determine the cutoff points for most of the variables, and we applied the Kaplan-Meier estimator and the Cox's proportional hazards model (stepwise method) to analyze the influence of these variables on survival. In univariate analysis, general factors including age ≥ 65 years, male gender, dialysis vintage ≥ 2 years, and the presence of diabetes and heart failure were all significant predictors of mortality. Among nutritional parameters, BMI ≥ 25 kg/m(2), SGA-B (mild PEW), nPNA < 1.15 g/kg per day, and the BIA-derived phase angle (PhA) <5.58° were also significantly associated with reduced survival. All of these factors maintained statistical significance in multivariate analysis, except for male gender and heart failure. We showed that low values of SGA, nPNA, and PhA independently predict mortality in HD patients. In conjunction with an earlier study, we demonstrated that the relative risk of death associated with these markers is highest during the first year of monitoring and it decreases in the following 4 years, although it still remains significantly increased. On the other hand, overweight and obesity were also associated with lower survival after 5 years, whereas this association was not apparent after 1 year.