Original Investigation
Dialysis Therapy
Serum Albumin, C-Reactive Protein, Interleukin 6, and Fetuin A as Predictors of Malnutrition, Cardiovascular Disease, and Mortality in Patients With ESRD

https://doi.org/10.1053/j.ajkd.2005.09.014Get rights and content

Background: Serum albumin (S-Alb), C-reactive protein (CRP), and interleukin 6 (IL-6) predict malnutrition, atherosclerotic cardiovascular disease (CVD), and mortality in patients with end-stage renal disease (ESRD). Fetuin A, an inhibitor of vascular calcification, also is associated strongly with clinical outcome in patients with ESRD. In this study, multivariate analyses were performed to assess these 4 biomarkers as predictors of malnutrition, CVD, and mortality in patients with ESRD. Methods: One hundred seventy-six patients with ESRD (54 ± 12 years) underwent measurements of S-Alb, high-sensitivity CRP (hs-CRP), plasma IL-6, and fetuin A close to the start of dialysis therapy and were followed up for a median of 26 months (range, 1 to 66 months). Nutritional status was evaluated by means of subjective global assessment. CVD was defined based on medical history. Associations between biomarker levels and malnutrition, CVD, and mortality were evaluated by means of receiver operating characteristic curve, logistic regression analysis, and Cox proportional hazards model. Results: All 4 biomarkers predicted malnutrition, CVD, and mortality. Multivariate analysis, according to receiver operating characteristic analysis, showed that malnutrition was predicted best by hs-CRP and IL-6 levels; CVD, by IL-6 level; and mortality, by albumin and IL-6 levels. When using the cutoff levels derived from receiver operating characteristics, logistic regression analysis showed that only hs-CRP level (odds ratio [OR], 3.6) was associated with malnutrition, and only IL-6 level (OR, 3.7) was associated with CVD. Levels of S-Alb, IL-6, and fetuin A, but not hs-CRP, were associated with increased relative risk for mortality as assessed by Cox in a model adjusting for age, sex, and diabetes mellitus. Conclusion: Multivariate analyses show that in patients with ESRD, malnutrition is predicted best by hs-CRP and IL-6 levels; CVD, by IL-6 level; and mortality, by S-Alb, IL-6, and fetuin A levels, but not by hs-CRP level. This comparative analysis indicates that of these biomarkers, IL-6 level may be the most reliable predictor of CVD and mortality in patients with ESRD.

Section snippets

Methods

One hundred seventy-six patients with ESRD (114 men; age, 54 ± 12 years; range, 23 to 70 years) were examined close to the start of renal replacement therapy. On average, blood samples were obtained approximately 1 month before the start of renal replacement therapy in 128 patients and 8 days after the start of renal replacement therapy in 48 patients; however, because levels of the 4 investigated biomarkers did not differ between these 2 groups, data were analyzed in all 176 patients. Patients

Patient Characteristics

Characteristics of enrolled patients are listed in Table 1. Sixty patients had a history of CVD (34%), and 50 patients (30%) were malnourished. As expected, mean age was older in patients with CVD (60 ± 8 versus 50 ± 12 years; P < 0.01) and PEM (60 ± 9 versus 50 ± 13 years; P < 0.01) than in patients without these complications. Moreover, patients with CVD, compared with patients without CVD, had lower S-Alb levels (P < 0.01) and greater median hs-CRP (1.35 versus 0.45 mg/dL; P < 0.01) and IL-6

Discussion

In the present comparative analysis of 4 biomarkers (S-Alb, hs-CRP, IL-6, and fetuin A), only IL-6 level among the 4 biomarkers predicted PEM, CVD, and mortality (Table 2). In the ROC analysis, an AUC between 0.7 and 0.8 is considered to represent a “fair predictor,”16 and this was seen for only IL-6 level. Previous studies showed that S-Alb level and inflammation predict PEM, CVD, and outcome,4, 17, 18, 19and fetuin A level is reported to predict CVD and outcome11 in patients with ESRD. In the

References (41)

Cited by (447)

View all citing articles on Scopus

Support and disclosure: This study was supported by grants from the Baxter Healthcare Corp, Deerfield, IL, and Söderbergh Foundation (P.S.). Bengt Lindholm is employed by Baxter Healthcare, Deerfield, IL.

Originally published online as doi:10.1053/j.ajkd.2005.09.014 on November 7, 2005.

View full text