| Abstract
| - Background. Full, as compared with partial, correction of anaemia did not reduce the mortality risk in patients with chronic kidney disease (CKD), although the underlying mechanisms are unknown. Since CKD is a high-risk population for cardiovascular disease (CVD), we tested a hypothesis that the presence of atherosclerosis affects the relationship between anaemia and mortality risk. Methods. We performed a single-centre 10-year follow-up study with an observational cohort of 505 haemodialysis patients to analyse the relationship between haematocrit and all-cause mortality. Baseline haematocrit levels did not differ between the 153 patients with CVD and the 352 patients without CVD. Results. During the follow-up, 268 patients died. Both Kaplan-Meier and univariate Cox analyses showed that higher haematocrit levels were a significant predictor of lower risk of death in the CVD (−) group, whereas haematocrit did not predict death in the CVD (+) group. In multivariate Cox analyses, the inverse relationship between haematocrit and mortality in the CVD (−) group remained significant and independent of 14 covariates including the use of erythropoietin. In contrast, using the same Cox models, the CVD (+) group did not show such a beneficial effect of higher haematocrit. Similar observations were made when the subjects were divided based on carotid artery intima-media thickness instead of the presence of CVD. Conclusions. These results support the hypothesis that the presence of atherosclerosis alters the relationship between anaemia and mortality risk in haemodialysis patients.
|
