Erschienen in:
01.06.2014 | Original Article
Lower serum albumin level is associated with higher fractional excretion of creatinine
verfasst von:
Masaru Horio, Enyu Imai, Yoshinari Yasuda, Tsuyoshi Watanabe, Seiichi Matsuo
Erschienen in:
Clinical and Experimental Nephrology
|
Ausgabe 3/2014
Einloggen, um Zugang zu erhalten
Abstract
Background
Creatinine clearance (Ccr) overestimates glomerular filtration rate (GFR) due to the tubular secretion of creatinine. It is known that fractional excretion of creatinine (FE-Cr) increases with decreasing GFR. Association of serum albumin level with the tubular secretion of creatinine was also reported previously. Alteration of FE-Cr may affect the performance of GFR estimating equations based on serum creatinine. Therefore, we analyzed the factors influencing FE-Cr and compared the performance of GFR equations in subjects stratified by serum albumin levels.
Methods
Seven hundred and fifty-seven Japanese subjects were included. GFR was measured by inulin renal clearance. GFR and Ccr were measured simultaneously. FE-Cr was calculated as the ratio of Ccr to GFR. Multivariate analysis was performed to evaluate the factors influencing FE-Cr. Age, gender, GFR, body mass index (BMI), body weight, height and serum albumin level were analyzed as the parameters. Estimated GFR was calculated by Japanese GFR equations based on serum creatinine (Eq-cr), serum cystatin C (Eq-cys) and 5 variables including serum albumin (Eq-5var).
Results
FE-Cr in subjects with serum albumin <3.0, 3.0–3.9 and ≥4.0 g/dl were 1.63 ± 0.48, 1.53 ± 0.55, and 1.40 ± 0.36, respectively. FE-Cr in subjects with serum albumin <3.0 or 3.0–3.9 g/dl were significantly higher than the value in subjects with serum albumin ≥4.0 g/dl. Multivariate analysis showed that GFR (p < 0.0001) and serum albumin level (p = 0.004) were independent parameters affecting FE-Cr. Biases of Eq-cr, Eq-cys and Eq-5var in subjects with serum albumin <3.0 g/dl were −9.5 ± 17.5, −0.7 ± 17.1 and −0.6 ± 14.8 ml/min/1.73 m2, respectively. Eq-cr significantly overestimated GFR compared with Eq-cys or Eq-5var. Biases in subjects with serum albumin ≥4.0 g/dl were 6.4 ± 18.8, 2.0 ± 18.1 and 3.0 ± 18.3 ml/min/1.73 m2, respectively. Eq-cr significantly underestimated GFR compared with Eq-cys or Eq-5var.
Conclusion
GFR and serum albumin level were independent parameters affecting FE-Cr. Alteration of FE-Cr according to the serum albumin levels may be one of the reasons of the bias of GFR equation based on serum creatinine.