The independent association between the indices of iron stores or administered intravenous iron, both of which vary over time, and survival in patients who are on maintenance hemodialysis (MHD) is not clear. It was hypothesized that the observed associations between moderately high levels of three iron markers (serum ferritin, iron, and iron saturation ratio) or administered intravenous iron and all-cause and cardiovascular death is due to the time-varying confounding effect of malnutrition-inflammation-cachexia syndrome (MICS). Time-dependent Cox regression models were examined using prospectively collected data of the 2-yr (July 2001 to June 2003) historical cohort of 58,058 MHD patients from virtually all DaVita dialysis clinics in the United States. After time-dependent and multivariate adjustment for case mix, administered intravenous iron and erythropoietin doses, and available surrogates of MICS, serum ferritin levels between 200 and 1200 ng/ml (reference 100 to 199 ng/ml), serum iron levels between 60 and 120 microg/ml (reference 50 to 59 microg/ml), and iron saturation ratio between 30 and 50% (reference 45 to 50%) were associated with the lowest all-cause and cardiovascular death risks. Compared with those who did not receive intravenous iron, administered intravenous iron up to 400 mg/mo was associated with improved survival, whereas doses >400 mg/mo tended to be associated with higher death rates. The association between serum ferritin levels >800 ng/ml and mortality in MHD patients seems to be due mostly to the confounding effects of MICS. For ascertaining whether the observed associations between moderate doses of administered intravenous iron and improved survival are causal or due to selection bias by indication, clinical trials are warranted.
