Longitudinal and cross-sectional effects of C-reactive protein, equilibrated normalized protein catabolic rate, and serum bicarbonate on creatinine and albumin levels in dialysis patients

Am J Kidney Dis. 2003 Dec;42(6):1200-11. doi: 10.1053/j.ajkd.2003.08.021.


Background: Loss of muscle mass and hypoalbuminemia each may result in part from either malnutrition, inflammation, or a combination of both. Short-term acidosis increases muscle protein catabolism and inhibits albumin synthesis.

Methods: We analyzed albumin and creatinine levels as outcome variables and their association with C-reactive protein (CRP) level, equilibrated normalized protein catabolic rate (enPCR), and serum bicarbonate level as independent variables from laboratory data obtained from patients in the Hemodialysis Study. Analyses controlled for race, sex, age, body mass index, and randomized treatment group.

Results: Albumin level correlated with both enPCR and CRP level, but not serum bicarbonate level, in both cross-sectional and longitudinal analyses. Effects of CRP level and enPCR were not linear. Albumin level correlated positively with enPCR for an enPCR less than 1.0 g/kg/d, but not for a greater enPCR, and correlated inversely with CRP level for a CRP level greater than 13 mg/L. Similarly, creatinine level correlated with both enPCR and CRP level. As in the case of albumin level, effects were not linear. Creatinine level correlated positively with enPCR for values less than 1.0 g/kg/d, but not for greater enPCR values. In contrast to albumin level, creatinine level correlated negatively with serum bicarbonate level, even when adjusted for enPCR.

Conclusion: Albumin and creatinine levels are independently associated with nutrition (enPCR) and inflammation (CRP level). The cross-sectional relationship with enPCR is apparent only at values less than 1.0 g/kg/d. CRP level is associated with reduced albumin and creatinine values when increased to values greater than 5.6 mg/dL. CRP may be increased to levels associated with increased cardiovascular risk with little or no effect on either serum albumin or creatinine level. Thus, a normal albumin level does not exclude elevated CRP levels.

Publication types

  • Comparative Study

MeSH terms

  • Acidosis / blood
  • Acidosis / epidemiology
  • Acidosis / etiology
  • Adult
  • Aged
  • Bicarbonates / blood*
  • Biomarkers
  • C-Reactive Protein / analysis*
  • Creatinine / blood*
  • Cross-Sectional Studies
  • Female
  • Humans
  • Hypoalbuminemia / blood
  • Hypoalbuminemia / epidemiology
  • Hypoalbuminemia / etiology
  • Inflammation / blood
  • Inflammation / complications
  • Inflammation / epidemiology
  • Kidney Failure, Chronic / blood
  • Kidney Failure, Chronic / complications
  • Kidney Failure, Chronic / epidemiology
  • Kidney Failure, Chronic / therapy
  • Longitudinal Studies
  • Male
  • Malnutrition / blood
  • Malnutrition / complications
  • Malnutrition / epidemiology
  • Middle Aged
  • Multicenter Studies as Topic
  • Proteins / metabolism*
  • Randomized Controlled Trials as Topic
  • Renal Dialysis* / statistics & numerical data
  • Serum Albumin / analysis*


  • Bicarbonates
  • Biomarkers
  • Proteins
  • Serum Albumin
  • C-Reactive Protein
  • Creatinine