The ratio of CRP to prealbumin levels predict mortality in patients with hospital-acquired acute kidney injury

Abstract

Background: Animal and human studies suggest that inflammation and malnutrition are common in acute kidney injury (AKI) patients. However, only a few studies reported CRP, a marker of inflammation, albumin, prealbumin and cholesterol, markers of nutritional status were associated with the prognosis of AKI patients. No study examined whether the combination of inflammatory and nutritional markers could predict the mortality of AKI patients.

Methods: 155 patients with hospital-acquired AKI were recruited to this prospective cohort study according to RIFLE (Risk, Injury, Failure, Lost or End Stage Kidney) criteria. C-reactive protein (CRP), and the nutritional markers (albumin, prealbumin and cholesterol) measured at nephrology consultation were analyzed in relation to all cause mortality of these patients. In addition, CRP and prealbumin were also measured in healthy controls (n = 45), maintenance hemodialysis (n = 70) and peritoneal dialysis patients (n = 50) and then compared with AKI patients.

Results: Compared with healthy controls and end-stage renal disease patients on maintenance hemodialysis or peritoneal dialysis, patients with AKI had significantly higher levels of CRP/prealbumin (p < 0.001). Higher level of serum CRP and lower levels of albumin, prealbumin and cholesterol were found to be significant in the patients with AKI who died within 28 days than those who survived >28 days. Similarly, the combined factors including the ratio of CRP to albumin (CRP/albumin), CRP/prealbumin and CRP/cholesterol were also significantly higher in the former group (p < 0.001 for all). Multivariate analysis (Cox regression) revealed that CRP/prealbumin was independently associated with mortality after adjustment for age, gender, sepsis and sequential organ failure assessment (SOFA, p = 0.027) while the others (CRP, albumin, prealbumin, cholesterol, CRP/albumin and CRP/cholesterol) became non-significantly associated. The hazard ratio was 1.00 (reference), 1.85, 2.25 and 3.89 for CRP/prealbumin increasing according to quartiles (p = 0.01 for the trend).

Conclusions: Inflammation and malnutrition were common in patients with AKI. Higher level of the ratio of CRP to prealbumin was associated with mortality of AKI patients independent of the severity of illness and it may be a valuable addition to SOFA score to independent of the severity of illness and it may be a valuable addition to SOFA score to predict the prognosis of AKI patients.

Figure 1

(A) Risk profiles for CRP according to quartiles, p = 0.029 for the trend; (B) Risk profiles for prealbumin, p = 0.013 for the trend; (C) Risk profile for CRP: prealbumin, p = 0.01 for the trend. Values along horizontal axis represent the means of CRP, prealbumin and CRP/prealbumin of the respective quartiles. Hazard ratios of death for CRP rising, prealbumin decreasing and CRP/prealbumin rising according to quartiles were adjusted by age and gender.

Figure 2

Kaplan-Meier analysis for the cumulative percentage of surviving patients at 90 days according to different CRP/prealbumin levels. The median of CRP/prealbumin was used as the cut-off point to divide the patients with AKI into 2 groups. Patients with higher levels of CRP/prealbumin (n = 78) had a significantly lower survival rate in these patients (log rank test, p < 0.01).

Figure 3

Levels of the ratio of CRP to prealbumin in different populations. Healthy (n = 45); MHD: maintenance hemodialysis (n = 70); PD: peritoneal dialysis (n = 50); AKI: acute kidney injury (n = 155). The units of CRP and prealbumin were both mg/dL. Mann-Whitney U test was used for comparison between the populations, p < 0.001. Horizontal bars indicate 10th, 25th, 50th (median), 75th, and 90th percentile levels.