| Abstract
| - Background. Elevated levels of inflammatory markers have been shown to be associated with increased mortality in chronic kidney disease (CKD) patients. Comorbid indexes are also helpful clinical instruments for predicting mortality. At present, it is unknown whether inflammatory markers add predictive information of death beyond that provided by comorbid indexes. Methods. This observational single-centre study included 404 patients (mean age 63±16 years) with CKD stage 4 and 5 predialysis who were prospectively followed-up. Data obtained at baseline: demographics, grade of comorbidity by Davies index, serum albumin, creatinine clearance, total white blood cell (WBC), polymorphonuclear leukocyte (PMN) counts, and high-sensitivity C-reactive protein (CRP) were analysed as potential determinants of the subsequent all-cause mortality. Receiver-operating characteristic (ROC) curves were used to determine the values of CRP, WBC and PMN most closely related to mortality. These cut-off values were used to define subgroups with high or low inflammatory markers. Uni- and multivariate Cox regression models were performed. Results. Median follow-up time was 583 days, with a mortality of 26%, and overall survival rate of 47%. In unadjusted Cox models, inflammatory markers (CRP, WBC and PMN) were all significantly associated with all-cause mortality. Age (HR 1.05; 95% CI 1.03-1.07, P<0.0001) and comorbid index (HR 2.15; 95% CI 1.54-3.00, P<0.0001) were strongly associated with mortality. The introduction of inflammatory markers into the multivariate Cox regression model did not add significant predictive power. In a stepwise Cox model, the age, comorbid index, serum albumin levels and creatinine clearance were the best predictive variables of mortality. Conclusions. Although elevated inflammatory markers are associated with a worse outcome in CKD patients, they did not add predictive information of all-cause mortality beyond that provided by age and the comorbid index.
|
