Variations in C-reactive protein during a single haemodialysis session do not associate with mortality

Abstract

An increase in C-reactive protein (CRP) levels during a single haemodialysis (HD) session has been associated with mortality. These associations, however, are difficult to understand from the current understanding of CRP metabolism. In 190 Swedish haemodialysis (HD) patients from the Mapping of Inflammatory Markers in Chronic Kidney Disease (MIMICK) cohort, CRP was measured before and after a HD session. During follow-up, events of death and censoring were recorded, and hazard ratios were calculated and analysed as a function of CRP variation. Results were replicated in 94 Dutch HD patients from the Netherlands Cooperative Study on the Adequacy of Dialysis (NECOSAD). In this cohort, also correlation and kappa statistics were calculated to assess concordance in CRP changes amid multiple dialysis sessions from the same individuals. In both cohorts, mean CRP values did not increase during a single HD session. In the MIMICK, median (interquartile range) dialysis vintage was 29.0 (14.8-57.0) months. In both crude [hazard ratio (95% confidence interval): 1.008 (0.971-1.047)] and multivariate Cox models [0.996 (0.949-1.046)], no association was observed with mortality. In the NECOSAD, individuals endured 6.0 (6.0-12.0) months on dialysis. No association was found with mortality neither in a crude [0.961 (0.908-1.018)] nor in an adjusted analysis [0.978 (0.923-1.037)]. Finally, the concordance between changes in different sessions was poor. CRP changes during a single HD session do not associate with mortality, thereby adding to the biological uncertainty concerning the ability of CRP to rise in such a short period.