Relevance of C-reactive protein levels in peritoneal dialysis patients

Kidney Int. 2002 Feb;61(2):615-20. doi: 10.1046/j.1523-1755.2002.00145.x.


Background: C-reactive protein (CRP) levels are increased in 30 to 50% of dialysis patients and predict cardiovascular morbidity and mortality. It is usually considered that raised CRP levels reflect underlying atherosclerosis. However, many patients may have clinically apparent cardiovascular disease without raised CRP levels. This study was designed to assess both the risk factors for high CRP levels and the usefulness of the test as a marker of clinically apparent coronary artery disease (CAD), peripheral vascular disease (PVD) and the presence of ongoing infections/inflammatory disorders (INF-INFL) in peritoneal dialysis patients.

Methods: A chart review of 190 prevalent peritoneal dialysis patients was performed. CRP, albumin, ferritin, erythropoietin (EPO) dose and resistance, Kt/V, and residual renal function values were obtained and a history or presence of cardiovascular disease (CAD, PVD) and presence of INF-INFL recorded. Data were analyzed by Chi-square, Spearman correlation and logistic regression.

Results: A total of 31% of patients had a raised CRP. INF-INFL was highly predictive of raised CRP levels (OR 16.97; 95% CI 5.41 to 53.14, P=0.000), whereas CAD and PVD either singly or in combination had no such association. The sensitivity/specificity for CRP as a test for INF-INFL was 83/77%. For CAD and PVD, the sensitivities were less than 40% and specificities 70%. Increased CRP values were more common in females but not in diabetics. Weak linear correlations existed between CRP levels and albumin, ferritin and residual renal function (r=-0.212, 0.228 and -0.163 respectively, P < 0.02). By regression analysis, INF-INFL predicted high CRP levels, but CAD and PVD did not. The majority of patients (57%) with high CRP had no identifiable cause; 40% of these patients had subsequent or previous normal CRP values. High transport status predicted high CRP levels (OR 7.28; 95% CI 1.417 to 37.36, P=0.006).

Conclusions: The majority of elevated CRP levels in peritoneal dialysis patients occur without an obvious cause. Clinically apparent cardiovascular disease does not predict high CRP levels. CRP levels vary over time in the same patient, from normal to high or vice versa, for no obvious reason. Sources of inflammation other than CAD, PVD and clinically obvious INF-INFL in peritoneal dialysis patients remain to be identified.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Aged
  • Arteriosclerosis / metabolism
  • C-Reactive Protein / metabolism*
  • Coronary Artery Disease / metabolism
  • Female
  • Humans
  • Kidney Failure, Chronic / metabolism*
  • Kidney Failure, Chronic / therapy
  • Male
  • Middle Aged
  • Peritoneal Dialysis*
  • Predictive Value of Tests
  • Retrospective Studies


  • C-Reactive Protein