Abstract
BACKGROUND: Hemoglobin variability (Hb-var) has been associated with increased mortality both in non-dialysis dependent chronic kidney disease (NDD-CKD) and end-stage renal disease (ESRD) patients. However, the impact of Hb-var in advanced NDD-CKD on outcomes after dialysis initiation remains unknown. METHODS: Among 11,872 US veterans with advanced NDD-CKD transitioning to dialysis between October 2007 through September 2011, we assessed Hb-var calculated from the residual SD of at least 3 Hb values during the last 6 months before dialysis initiation (prelude period) using within-subject linear regression models, and stratified into quartiles. Outcomes included post-transition all-cause, cardiovascular, and infection-related mortality, assessed in Cox proportional hazards models and adjusted for demographics, comorbidities, length of hospitalization, medications, estimated glomerular filtration rate (eGFR), type of vascular access, Hb parameters (baseline Hb [i.e., intercept] and change in Hb [i.e., slope]), and number of Hb measurements. RESULTS: Higher prelude Hb-var was associated with use of iron and antiplatelet agents, tunneled dialysis catheter use, higher levels of baseline Hb, change in Hb, eGFR, and serum ferritin. After multivariable adjustment, higher prelude Hb-var was associated with higher post-ESRD all-cause and infection-related mortality, but not cardiovascular mortality (adjusted hazard ratios [95% CI] for the highest [vs. lowest] quartile of Hb-var, 1.10 [1.02-1.19], 1.28 [0.93-1.75], and 0.93 [0.79-1.10], respectively). CONCLUSIONS: High pre-ESRD Hb-var is associated with higher mortality, particularly from infectious causes rather than cardiovascular causes. Further research is required to clarify the underlying mechanisms and true causal nature of the observed association.