Clinical evaluation of serum cystatin C and creatinine in patients with chronic kidney disease: A meta-analysis

Abstract

Objective

Glomerular filtration rate (GFR) is an important indicator of renal function. This meta-analysis aimed to evaluate the diagnostic value of serum cystatin C (CysC) and serum creatinine (SCr) for estimating GFR in patients with chronic kidney disease.

Methods

Google Scholar, PubMed®, Cochrane Library and China National Knowledge Infrastructure databases were searched, to identify randomized controlled trials that determined the diagnostic value of CysC and SCr, for estimating GFR in patients with chronic kidney disease.

Results

The inclusion criteria were met by 17 studies (total number of patients with chronic kidney disease, 2521). Meta-analysis showed that when the diagnostic cut-off value of GFR was 80–90 ml/min/1.73 m², the heterogeneity was modest for CysC (I²= 48%, summary sensitivity [SEN] = 0.803, summary specificity [SPE] = 0.821), but there was no heterogeneity for SCr (I²= 0.0%, SEN = 0.697, SPE = 0.787). Meta-analysis of the studies demonstrated a significant difference between patients with chronic kidney disease and controls, for CysC and SCr.

Conclusions

This meta-analysis demonstrated significant correlations between CysC, SCr and GFR. CysC was more sensitive, but less specific, than SCr for the estimation of GFR.

Keywords

Cystatin C creatinine chronic kidney disease diagnostic value meta-analysis

Introduction

Glomerular filtration rate (GFR) is an important indicator that is used to evaluate kidney function. The gold-standard method for the determination of GFR is inulin clearance.¹ However, inulin clearance is generally only used for scientific research because the method is cumbersome. In the clinic, the most commonly used method for GFR assessment is the determination of the serum creatinine (SCr) clearance rate, but this method is affected by a variety of factors,² including the production of kidney tubule-derived creatinine.³ Research has demonstrated that serum cystatin C (CysC) is a relatively ideal endogenous marker of GFR.^4–6 Research has shown that CysC is a more sensitive serum marker of GFR than SCr,⁷ but there are differing views.^8,9 Although several meta-analyses have evaluated the use of CysC to estimate GFR,^10,11 research undertaken in China were not included and no analyses of heterogeneity among the studies were undertaken. The present meta-analysis investigated the diagnostic value of CysC and SCr in the estimation of GFR, in patients with chronic kidney disease.

Materials and methods

Search strategy

A systematic search of publications listed in electronic databases (Google Scholar, PubMed®, Cochrane Library and China National Knowledge Infrastructure) between February 1990 and October 2012 was conducted, using the following keywords: cystatin C, creatinine, GFR, diagnosis test, sensitivity and specificity. Language restrictions were not applied, but the search was limited to human studies. The list of articles was reviewed by two authors (M. Z. and X-Y. C.), who read the full text of all studies that had been identified; they decided on their suitability for inclusion in the review, based on whether they met the prespeciﬁed inclusion criteria. Any disagreement between the two authors was resolved using a consensus procedure and, if necessary, with the input of a third author (G-Y. C.).

Study selection

Studies were eligible for inclusion if they met the following criteria: (i) they investigated the use of CysC and SCr in the estimation of GFR in patients with chronic kidney disease; (ii) they provided computed data including sensitivity and specificity. Cross-over trials were not included in this meta-analysis. The search attempted to identify cluster-randomized trials; they were included and analysed in accordance with section 16.3 of the Cochrane Handbook for Systematic Reviews of Interventions.¹² All trials that did not use intention-to-treat analysis were recorded, and every attempt was made to analyse the current data by this principle.

Data extraction and management

Two reviewers (M. Z. and X-Y. C.) extracted the data independently to a self-developed data extraction form. Studies reported in non-English language journals were translated before assessment. Where more than one publication of one trial existed, only the publication with the most complete data was included. If publications did not contain the full information necessary for a meta-analysis, the missing information was obtained directly from the study authors. Any disagreement between the two reviewers was resolved using a consensus procedure, and if necessary, with the input of a third reviewer (G-Y. C.).

One author (M. Z.) entered the data into RevMan software version 5.1 (Cochrane Collaboration, Oxford, UK) and Meta-DiSc (version 1.4; Unit of Clinical Biostatistics, Ramón y Cajal Hospital, Madrid, Spain);^13,14 and a second author (G-Y. C.) independently checked the data entry. For each study, the following data were extracted: summary receiver operating characteristic (SROC) curves, summary sensitivity (SEN), summary specificity (SPE), area under the receiver operating characteristic curves (SAUC), positive likelihood ratio (+LR) and negative likelihood ratio (−LR).

Statistical analyses

RevMan 5.1 and Meta-DiSc 1.4 software were used to pool data from all included studies. For dichotomous data, the results were summarized as odds ratios (OR) with 95% conﬁdence intervals (CI). For continuous outcomes, weighted mean difference (WMD) (when measures were in the same unit) or standardized mean difference (SMD) (when different scales were used to evaluate the same outcome), with 95% CI, were used.

Two authors (D. W. and R-B. W.) independently used the Grading of Recommendations Assessment, Development, and Evaluation criteria to assess the risk of bias for all included studies.¹⁵ Heterogeneity among studies was assessed using the I² method and an I²> 50% was considered to indicate significant heterogeneity; I²values of 25–50% were deemed to show modest heterogeneity and I² values <25% were deemed to represent low heterogeneity. Subgroup analyses were used to explore possible sources of heterogeneity. When there was significant heterogeneity, a random-effects model was used. A sensitivity analysis was performed if there was signiﬁcant heterogeneity (I²> 50%). A fixed-effects model was used in the absence of heterogeneity. Reporting and publication bias was estimated using funnel plots; although reporting and publication bias was minimized by searching multiple databases, contacting authors for missing data, and by utilizing clinical practice guidelines and systematic reviews to make the search as complete as possible. A P-value < 0.05 was considered statistically significant.

Results

The initial database search identified 256 studies. Of these, 239 were excluded after reading the titles and abstracts, leaving 17 double-blind, randomized controlled trials that met the inclusion criteria.^16–32 All 17 studies evaluated diagnostic tests for the assessment of GFR using CysC and SCr and included a total of 2521 patients with chronic kidney disease (Table 1).^16–32 According to the GFR staging for chronic kidney disease, patients with stage 1 were set as the control group and the diagnostic cut-off point of GFR was 80–90 ml/min/1.73 m². The gold-standard methods of estimating GFR were as follows: exogenous inulin, iohexol or ¹²⁵I-iothalamate, ^99mTc-diethylene triamine pentacaetic acid (^99mTc-DTPA) and ⁵¹Cr-ethylenediamine tetra-acetic acid (⁵¹Cr-EDTA) clearance rate. The serum CysC and SCr were also measured. The detection methods of serum CysC were particle-enhanced turbidimetric immunoassay (PETIA) and particle-enhanced nephelometric immunoassay (PENIA).^33,34 The detection methods for SCr were the Jaffe method and an enzymatic method.³⁵ None of the 17 studies described the randomization methods that they used or how patient allocation was concealed.Data for CysC were available from all 17 studies.^16–32 There was a total of 3567 study participants: 2521 patients with chronic kidney disease and 1046 control patients. Meta-analysis of the 17 studies demonstrated a significant difference between patients with chronic kidney disease and controls (P < 0.00001) (Figure 1).

Figure 1.

Meta-analysis of 17 studies that investigated the diagnostic value of serum cystatin C in the evaluation of glomerular filtration rate in patients with chronic kidney disease.^16–32 M-H, Mantel–Haenszel; CI, confidence interval.

Table 1.

Major characteristics of 17 studies selected for a meta-analysis of the diagnostic value of serum cystatin C (CysC) and serum creatinine (SCr) in the evaluation of glomerular filtration rate, in chronic kidney disease.^16–32

Author	Year	Study participants		Methodological quality
Author	Year	Patients, n	Controls, n	Study design	Randomized method	Allocation concealment	Double- blind	Baseline^a	Detection method
Perlemoine et al.²¹	2003	53	36	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Jaffe (SCr)
Cho et al.¹⁷	2011	286	34	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Jaffe (SCr)
Hermida-Cadahia et al.¹⁸	2012	794	50	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Jaffe (SCr)
Hoek et al.¹⁹	2003	95	51	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Enzymatic (SCr)
Filler et al.²⁰	2002	75	150	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Enzymatic (SCr)
Chantrel et al.¹⁶	2000	77	84	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Jaffe (SCr)
Macisaac et al.²²	2007	90	107	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Jaffe (SCr)
Spanaus et al.²³	2010	155	72	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Jaffe (SCr)
Pucci et al.²⁴	2007	159	129	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Jaffe (SCr)
Mussap et al.²⁵	2002	28	24	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Enzymatic (SCr)
Pan et al.²⁶	2011	184	50	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Jaffe (SCr)
Zhou²⁷	2008	123	63	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Jaffe (SCr)
Li et al.²⁸	2005	14	37	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Jaffe (SCr)
Li et al.²⁹	2005	51	20	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Jaffe (SCr)
Zhang³⁰	2008	100	35	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Jaffe (SCr)
Wang et al.³¹	2010	62	14	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Enzymatic (SCr)
Li et al.³²	2006	175	90	RCT	Unclear	Unclear	Yes	Similar	PENIA (CysC) Jaffe (SCr)

Baseline clinical characteristics such as age, sex ratio and body weight of the control and patient groups.

RCT, randomized controlled trial; PENIA, particle-enhanced nephelometric immunoassay; PETIA, particle-enhanced turbidimetric immunoassay.

Data for SCr was available from 13 studies.^{16,18,20–29,32} There was a total of 2890 study participants: 1978 patients with chronic kidney disease and 912 control patients. The meta-analysis of the 13 studies demonstrated a significant difference between patients with chronic kidney disease and controls (P < 0.00001) (Figure 2).

Figure 2.

Meta-analysis of 13 studies that investigated the diagnostic value of serum creatinine in the evaluation of glomerular filtration rate, in patients with chronic kidney disease.^{16,18,20–29,32} M-H, Mantel–Haenszel; CI, confidence interval.

Analyses demonstrated that there was modest heterogeneity among the 17 studies used for the meta-analysis of CysC (I²= 48%, SEN = 0.803, SPE = 0.821) (Figures 1 and 3), but no significant heterogeneity among the 13 studies used for the meta-analysis of SCr (I²= 0.0%, SEN = 0.697, SPE = 0.787) (Figures 2 and 4).

Figure 3.

Begg’s funnel plot for heterogeneity, for 17 studies that evaluated the diagnostic value of serum cystatin C in the evaluation of glomerular filtration rate, in patients with chronic kidney disease.^16–32 SE, standard error; OR, odds ratio.

Figure 4.

Begg’s funnel plot for heterogeneity, for the 13 studies that evaluated the diagnostic value of serum creatinine in the evaluation of glomerular filtration rate in patients with chronic kidney disease.^{16,18,20–29,32} SE, standard error; OR, odds ratio.

The SROC curves of the use of CysC and SCr to evaluate GFR in patients with chronic kidney disease are shown in Figures 5 and 6, respectively. The SROC curve for CysC in the evaluation of GFR in patients with chronic kidney disease gave the following results: SAUC = 0.9332; standard error (SE) < 0.01; Q-test* = 0.8832; and SE (Q-test*) < 0.01. For SCr, the results from the SROC curve were as follows: SAUC = 0.8435; SE < 0.05; Q-test* = 0.8372; and SE (Q-test*) < 0.05.

Figure 5.

Summary receiver operating characteristic curves for the 17 studies that evaluated the diagnostic value of serum cystatin C in the evaluation of glomerular filtration rate, in patients with chronic kidney disease.^16–32 Area under the receiver operating characteristic curve = 0.9332, standard error (SE) <0.01, Q-test* = 0.8832, SE (Q-test*)<0.01.

Figure 6.

Summary receiver operating characteristic curves for the 13 studies that evaluated the diagnostic value of serum creatinine in the evaluation of glomerular filtration rate, in patients with chronic kidney.^{16,18,20–29,32} Area under the receiver operating characteristic curve = 0.8435, standard error (SE) <0.05, Q-test* = 0.8372, SE (Q-test*) <0.05.

In the current meta-analysis, 13 studies used the PENIA method for measuring serum CysC,^{16–18,20–28,32} and the heterogeneity analysis results were as follows: P = 0.083, I²= 38.6%, diagnostic SEN = 0.882, diagnostic SPE = 0.712, OR = 33.4 (95% CI 22.6, 48.9) and SAUC = 0.9388. The remaining four studies used the PETIA method for measuring CysC,^19,29–31 and the heterogeneity analysis results were as follows: P = 0.269, I²= 30.3%, diagnostic SEN = 0.845, diagnostic SPE = 0.832, OR = 25.8 (95% CI 12.6, 55.1) and SAUC = 0.9062. In terms of the methods used to measure SCr, 13 studies used the Jaffe method,^{16–18,21–24,26–30,32} and the heterogeneity analysis results were as follows: P = 0.794, I²= 0.01%, diagnostic SEN = 0.803, diagnostic SPE = 0.883, OR = 23.2 (95% CI 18.6, 34.2) and SAUC = 0.8938. The remaining four studies used an enzymatic method^19,20,25,31 and the heterogeneity analysis results were as follows: P = 0.104, I²= 60.8%, diagnostic SEN = 0.572, diagnostic SPE = 0.975, OR = 16.9 (95% CI 7.8, 43.9) and SAUC = 0.9367.

Eight studies investigated the use of CysC and SCr to measure GFR in patients with different GFR cut-off levels within the decreased range of 60–90 ml/min/1.73 m².^{18–21,29–32} The results for CysC and SCr ,when there was a decrease in GFR in the range of 60–90 ml/min/1.73 m² from the combined studies, were as follows: CysC P = 0.754, I²= 0.04%, OR = 8.14 (95% CI 5.59, 13.21), diagnostic SEN = 0.854, diagnostic SPE = 0.636 and SAUC = 0.8312; SCr P = 0.998, I²= 0.03%, OR = 10.13 (95% CI 7.72, 17.32), diagnostic SEN = 0.696, diagnostic SPE = 0.883 and SAUC = 0.8365.

Discussion

Serum creatinine is the most commonly used indicator for the evaluation of renal function, so it plays an important clinical role in the assessment of GFR in patients with chronic kidney disease. CysC is secreted by nucleated cells, its concentration is not affected by age, sex, diet, inflammation and other factors, which make it an ideal endogenous marker of GFR changes and renal function.^4–6 Many studies have reported the diagnostic value of CysC and SCr for estimating GFR.^16–32 In this study, all of the 17 studies that were identified and included in the meta-analysis were comparative studies of the diagnostic value of CysC and SCr for estimating GFR.^16–32 In the studies in included in this meta-analysis,^16–32 GFR = 80–90 ml/min/1.73 m² was set at the diagnostic cut-off point, so that patients with a GFR > 80–90 ml/min/1.73 m² were classified as the control group (stage 1), while patients with a GFR < 80–90 ml/min/1.73 m² were classified as the experimental group. The meta-analysis results showed that there was a good correlation between CysC, SCr and the gold-standard measures of GFR. The diagnosis of kidney disease using CysC demonstrated modest heterogeneity, but there was no significant heterogeneity in the diagnosis of kidney disease using SCr.

In terms of the source of heterogeneity, there were five gold-standard methods of estimating GFR: exogenous inulin, iohexol or ¹²⁵I-iothalamate, ^99mTc-DTPA and ⁵¹Cr-EDTA clearance rate. Inulin is the gold-standard method for the determination of GFR, which was reported in one study.³³ Determining GFR using radionuclide substances is associated with problems of accuracy and reproducibility, as a consequence of the use of different labelling substances and different methods of calculating GFR.

Serum CysC was detected using either PENIA or PETIA methods. Meta-analysis showed that the heterogeneity was related to the two detection methods and the differences were possibly attributable to methodological differences, such as the use of different antibodies, use of two substantially different assay methods, variability in the proneness to matrix effects, and use of different detection methodologies. The systematic difference in concentrations may be attributable to the lack of a commonly approved international standard. PETIA used reagents purchased from DAKO (Glostrup, Denmark), and PENIA used reagents purchased from Dade Behring (Hong Kong, China), so the differences between studies were possibly caused by different instruments, reagents and calibrators. The heterogeneity might also have been caused by differences in the methods used the select the critical CysC cut-off value that was used by individual studies. The SCr was detected using either the Jaffe method or an enzymatic method. Meta-analysis showed that the heterogeneity was related to the two detection methods; heterogeneity existed with the enzymatic method (I²= 60.8%), but not with the Jaffe method (I²= 0.01%), which might have been caused by there being fewer studies that used an enzymatic method (n = 4), or the reference range of SCr. For example, SCr was less sensitive (63%) but more specific (95%) in the heterogeneous Jaffe method, with a setting range of 44133 µmol/l compared with the enzymatic method, with a setting range of 40–97µmol/l.^36,37

The inclusion of GFR staging significantly affected the heterogeneity of the studies, which suggests that GFR staging research should be undertaken in patients with chronic kidney disease, focusing on the diagnostic value when the GFR = 60–90 ml/min/1.73 m². Eight studies conducted GFR staging.^{18–21,29–32} When there was a decrease in GFR in the range of 60–90 ml/min/1.73 m², the heterogeneity of CysC (I²= 0.04%) and SCr (I²= 0.03%) was significantly reduced, suggesting that heterogeneity was greatly increased when patients with mild, moderate and severe GFR were included in one study. The diagnostic sensitivity of CysC was higher than that with SCr, but the diagnostic specificity was lower with CysC.

The inappropriate selection of control groups might have affected the heterogeneity. In all of the included studies, patients with chronic kidney disease with a GFR > 80–90 ml/min/1.73 m² were defined as the control group (stage 1). Thereby, the specificity and false-positive rate were inaccurate. In further studies, the control group should consist of subjects that do not have kidney disease and they should be compared with patients with chronic kidney disease, defined as a GFR < 80–90 ml/min/1.73 m². Whether future studies investigate CysC or SCr, they should focus on the diagnostic value in subjects with a high filtration rate (GFR > 90 ml/min/1.73 m²) and in patients with early stage chronic kidney disease (GFR = 60–90 ml/min/1.73 m²), because there was no significant difference between the diagnostic value of CysC and SCr in patients with moderate-to-severe stage kidney disease (GFR < 60 ml/min/1.73 m²). The CysC level was increased when the GFR was higher (GFR > 90 ml/min/1.73 m²),³⁸ which might be related to the inappropriate control group settings, and the specific reason needs further investigation.

In conclusion, the diagnostic sensitivity of CysC in patients with chronic kidney disease was higher than that of SCr, but the diagnostic specificity of CysC was lower than that of SCr. There was modest heterogeneity for CysC, however no significant heterogeneity for SCr. This research area still requires further high-quality studies to be undertaken. Based on these current findings and in our opinion, future studies should focus on GFR staging by standardizing the definition of control groups and the cut-off values for CysC and SCr, in order to provide the diagnostic basis for the clinical application of CysC and SCr for estimating GFR.

Footnotes

Declaration conflicting interest

The authors declare that there are no conflicts of interest.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Sterner

Frennby

Mansson

. Determining ‘true’ glomerular filtration rate in healthy adults using infusion of inulin and comparing it with values obtained using other clearance techniques or prediction equations. Scand J Urol Nephrol 2008; 42: 278–285.

Thorp

Eastman

. Potential application of the National Kidney Foundation's chronic kidney disease guidelines in a managed care setting. Am J Manag Care 2004; 10: 417–422.

Olsen

Ladefoged

Feldt-Rasmussen

. The effects of cimetidine on creatinine excretion, glomerular filtration rate and tubular function in renal transplant recipients. Scand J Clin Lab Invest 1989; 49: 155–159.

Zhang

Zhan

Xie

. Evaluation of glomerular filtration rate using cystatin C in diabetic patients analysed by multiple factors including tubular function. J Int Med Res 2010; 38: 473–483.

Mojiminiyi

Abdella

. Evaluation of cystatin C and beta-2 microglobulin as markers of renal function in patients with type 2 diabetes mellitus. J Diabetes Complications 2003; 17: 160–168.

Aksun

Ozmen

. Beta2-microglobulin and cystatin C in type 2 diabetes: assessment of diabetic nephropathy. Exp Clin Endocrinol Diabetes 2004; 112: 195–200.

Liu

. Is the Chronic Kidney Disease Epidemiology Collaboration four-level race equation better than the cystatin C equation? Nephrology (Carlton) 2012; 17: 407–414.

Knight

Verhave

Spiegelman

. Factors influencing serum cystatin C levels other than renal function and the impact on renal function measurement. Kidney Int 2004; 65: 1416–1421.

Stevens

Coresh

Greene

. Assessing kidney function – measured and estimated glomerular filtration rate. N Engl J Med 2006; 354: 2473–2483.

10.

Dharnidharka

Kwon

Stevens

. Serum cystatin C is superior to serum creatinine as a marker of kidney function: a meta-analysis. Am J Kidney Dis 2002; 40: 221–226.

11.

Roos

Dounst

Tett

. Diagnostic accuracy of cystatin C compared to serum creatinine for the estimation of renal dysfunction in adults and children – a meta-analysis. Clin Biochem 2007; 40: 383–391.

12.

Yoshii

Plaut

McGraw

. Analysis of the reporting of search strategies in Cochrane systematic reviews. J Med Libr Assoc 2009; 97: 21–29.

13.

Walter

. Properties of the summary receiver operating characteristic (SROC) curve for diagnostic test data. Stat Med 2002; 21: 1237–1256.

14.

Zamora

Abraira

Muriel

. Meta-DiSc: a software for meta-analysis of test accuracy data. BMC Med Res Methodol 2006; 6: 31–31.

15.

Balshem

Helfand

Schünemann

. GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol 2011; 64: 401–406.

16.

Chantrel

Agin

Offner

. Comparison of cystatin C versus creatinine for detection of mild renal failure. Clin Nephrol 2000; 54: 374–381.

17.

Cho

Lee

Suh

. The significance of serum cystatin C accompanied by a normal serum creatinine level in pediatric patients with chronic kidney disease. Lab Medline 2011; 42: 549–552.

18.

Hermida-Cadahia

Lampon

Tutor

. Impact of creatinine production on the agreement between glomerular filtration rate estimates using cystatin C-derived, and 4- and 6-variable Modification of Diet in Renal Disease (MDRD), and Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations. Ups J Med Sci 2012; 117: 402–410.

19.

Hoek

Kemperman

Krediet

. A comparison between cystatin C, plasma creatinine and the Cockcroft and Gault formula for the estimation of glomerular filtration rate. Nephrol Dial Transplant 2003; 18: 2024–2031.

20.

Filler

Priem

Lepage

. Beta-trace protein, cystatin C, beta(2)-microglobulin, and creatinine compared for detecting impaired glomerular filtration rates in children. Clin Chem 2002; 48: 729–736.

21.

Perlemoine

Beauvieux

Rigalleau

. Interest of cystatin C in screening diabetic patients for early impairment of renal function. Metabolism 2003; 52: 1258–1264.

22.

Macisaac

Tsalamandris

Thomas

. The accuracy of cystatin C and commonly used creatinine-based methods for detecting moderate and mild chronic kidney disease in diabetes. Diabet Med 2007; 24: 443–448.

23.

Spanaus

Kollerits

Ritz

. Serum creatinine, cystatin C, and beta-trace protein in diagnostic staging and predicting progression of primary nondiabetic chronic kidney disease. Clin Chem 2010; 56: 740–749.

24.

Pucci

Triscornia

Lucchesi

. Cystatin C and estimates of renal function: searching for a better measure of kidney function in diabetic patients. Clin Chem 2007; 53: 480–488.

25.

Mussap

Dalla Vestra

Fioretto

. Cystatin C is a more sensitive marker than creatinine for the estimation of GFR in type 2 diabetic patients. Kidney Int 2002; 61: 1453–1461.

26.

Pan

Liu

. Study on serum cystatin C evaluating glomerular filtration function. Modern J Integr Chin Trad West Med 2011; 20: 2088–2090. [in Chinese].

27.

Zhou

. Distribution and evaluation value of serum cystatin C for glomerular filtration rate. Zhejiang J Clin Med 2008; 10: 694–695. [in Chinese].

28.

. Comparison of evaluation value between serum cystatin C and creatinine for glomerular filtration function in patients with diabetes. Chin J Lab Med 2005; 28: 602–605. [in Chinese].

29.

Zhao

Huang

. Evaluation value of serum cystatin C for glomerular filtration function. Chin J Lab Med 2005; 28: 515–516. [in Chinese].

30.

Zhang

. Evaluation of serum cystatin C for glomerular filtration function in patients with Type 2 diabetes nephropathy kidney. Clin J Med Off 2008; 36: 999–1000. [in Chinese].

31.

Wang

Bao

. Evaluation of serum cystatin C for glomerular filtration function in patients with chronic kidney diseases. Chin J Integr Trad West Nephr 2010; 5: 421–423. [in Chinese].

32.

Zhang

. Reference values of serum cystatin C and creatinine in healthy volunteers, and comparison of their evaluation of glomerular filtration function in chronic renal patients. Chin J Lab Med 2006; 29: 970–974. [in Chinese].

33.

Satirapoj

Supasyndh

Boonyavarakul

. The correlation of insulin resistance and renal function in non diabetic chronic kidney disease patients. J Med Assoc Thai 2005; 88(Suppl 3): S97–S104.

34.

Voskoboev

Larson

Rule

. Analytic and clinical validation of a standardized cystatin C particle enhanced turbidimetric assay (PETIA) to estimate glomerular filtration rate. Clin Chem Lab Med 2012; 50: 1591–1596.

35.

Allegaert

Kuppens

Mekahli

. Creatinine reference values in ELBW infants: impact of quantification by Jaffe or enzymatic method. J Matern Fetal Neonatal Med 2012; 25: 1678–1681.

36.

Wong

Cobb

Umehara

. Heterogeneity of serum creatine kinase activity among racial and gender groups of the population. Am J Clin Pathol 1983; 79: 582–586.

37.

Rehman

Fought

Solomon

. N-acetylcysteine effect on serum creatinine and cystatin C levels in CKD patients. Clin J Am Soc Nephrol 2008; 3: 1610–1614.

38.

Waheed

Matsushita

Sang

. Combined association of albuminuria and cystatin C-based estimated GFR with mortality, coronary heart disease, and heart failure outcomes: The Atherosclerosis Risk in Communities (ARIC) Study. Am J Kidney Dis 2012; 60: 207–216.