Abstract
Objective. We examined the hypothesis that cystatin C, a novel marker of renal function, is elevated in rheumatoid arthritis (RA) and is associated with inflammation and coronary atherosclerosis.
Methods. We measured serum cystatin C, creatinine, tumor necrosis factor-α and interleukin 6 concentrations, coronary artery calcium score (CACS), and Modified Diet in Renal Disease estimated glomerular filtration rate in 167 patients with RA and 91 controls.
Results. Cystatin C was higher in RA patients [median (IQR) 1.16 (0.99–1.35) mg/l] than controls [1.01 (0.90–1.19) mg/l; p < 0.001] and correlated positively with erythrocyte sedimentation rate (p < 0.001), C-reactive protein (p = 0.01), 28-joint Disease Activity Score (p = 0.006), and Framingham risk score (FRS; p = 0.02). Cystatin C was correlated with CACS (p < 0.001) in RA, but this was not significant after adjustment for age, race, sex, and FRS (p = 0.44).
Conclusion. Cystatin C concentrations are higher in RA than controls and may reflect inflammation and undetected subclinical renal dysfunction. Cystatin C provides information regarding the risk of atherosclerosis in RA, but this is not independent of the information provided by conventional cardiovascular risk factors.
Cystatin C is a novel marker of renal function. Unlike creatinine, it is not substantially affected by muscle mass or diet1 and thus is a more accurate measure of glomerular filtration rate (GFR). Cystatin C concentrations also predict cardiovascular events2,3. Rheumatoid arthritis (RA) is frequently associated with chronic inflammation, subclinical renal impairment4, and increased cardiovascular risk5. We examined the hypotheses that cystatin C concentrations (1) are higher in patients with RA than in controls, independent of renal function; (2) are associated with measures of inflammation in RA; and (3) are associated with coronary atherosclerosis.
MATERIALS AND METHODS
As described, we studied 167 patients with RA and 91 controls frequency matched for age, sex, and race5,6. Our study was approved by the Institutional Review Board of Vanderbilt University Hospital and all subjects gave written informed consent. Clinical and demographic variables were measured: Agatston coronary artery calcium score (CACS), Framingham risk score (FRS), C-reactive protein (CRP), tumor necrosis factor-α (TNF-α), and interleukin 6 (IL-6), as described5,6. Cystatin C concentrations in serum were measured by ELISA. The Modification of Diet in Renal Disease (MDRD) formula was used to estimate GFR (eGFR)7.
Statistical analysis
Descriptive statistics are presented as the mean ± SD or as the median with interquartile range (IQR), as appropriate. Concentrations of cystatin C, creatinine, and MDRD-eGFR were compared in patients with RA and controls using Wilcoxon signed-rank tests and multivariable linear regression with adjustment for age, race, and sex. In RA, correlation between cystatin C and measures of renal function, traditional cardiovascular risk factors, inflammation, other clinical variables, and CACS was examined using Spearman’s rank correlation coefficient. The association between clinical and laboratory variables and renal function measures (cystatin C, creatinine, and MDRD-eGFR) was assessed using separate multivariable linear regression adjusting for age, race, and sex.
The effect of the renal function measures on the CACS was examined by applying a proportional odds logistic regression model with adjustment for age, race, sex, and FRS. Age was allowed to have nonlinear terms because it has the strongest predictive value for atherosclerosis. Concentrations of cystatin C, creatinine, homocysteine, CRP, TNF-α, and IL-6 were natural logarithm-transformed to improve normality. Statistical analyses were performed using R version 2.10.0 (http://www.r-project.org), and 2-sided p values < 0.05 were considered statistically significant.
RESULTS
Characteristics of patients with RA (n = 167) and controls (n = 91) are shown in Table 1. Cystatin C was significantly higher in patients with RA (p < 0.001); however, creatinine and MDRD-eGFR did not differ significantly. Cystatin C remained significantly higher in RA after adjustment for age, race, and sex (p < 0.001), and after additional adjustment for MDRD-eGFR (p < 0.001).
In RA, cystatin C was positively correlated with creatinine (p < 0.001) and negatively correlated with MDRD-eGFR (p < 0.001). In controls, cystatin C was negatively correlated with MDRD-eGFR (p = 0.01) but not with creatinine (p = 0.14). In RA, serum creatinine was significantly higher in men than women (p < 0.001); however, cystatin C (p = 0.12) and MDRD-eGFR (p = 0.39) were not (Table 2). Cystatin C, creatinine, and MDRD-eGFR were not significantly associated with smoking, diabetes mellitus, or current use of corticosteroids, nonsteroidal antiinflammatory drugs, and cyclooxygenase-2 inhibitors after adjustment for age, race, and sex. However, methotrexate (MTX) use was associated with lower cystatin C concentrations (p < 0.001; Table 2). Anti-TNF therapy did not affect cystatin C concentrations.
Cystatin C and MDRD-eGFR were correlated significantly with the FRS, but after adjustment for age, race, and sex, only cystatin C remained significant (p = 0.02). After statistical adjustment for age, race, and sex, cystatin C was significantly associated with erythrocyte sedimentation rate (p < 0.001), CRP (p = 0.01), and 28-joint Disease Activity Score (p = 0.006), but not with IL-6 (p = 0.32) or TNF-α (p = 0.11; Table 3). MDRD-eGFR and serum creatinine were not significantly associated with measures of inflammation (Table 3).
Cystatin C (p < 0.001), creatinine (p = 0.01), and MDRD-eGFR (p = 0.03) all correlated significantly with CACS in univariate analyses, but not after adjustment for age, race, sex, and FRS (Table 3).
DISCUSSION
Cystatin C is superior to creatinine in providing an accurate measure of renal function1. However, cystatin C also predicts cardiovascular risk2,3,8, perhaps through association with inflammation9,10,11, as well as renal impairment1.
The concentrations of cystatin C in RA [median (IQR) 1.16 (0.99–1.35) mg/l] and controls [1.01 (0.90–1.19) mg/l] were similar to those found in other populations without known renal impairment. For example, in blood donors the reference range for cystatin C concentrations was 0.53–0.92 mg/l in those < 50 years of age and 0.58–1.02 mg/l in those > 50 years of age12, and in the Framingham Offspring Study (n = 3214, mean age 61 yrs) the mean cystatin C concentration was 0.93 ± 0.18 mg/l13. Cystatin C was significantly higher in patients with RA than controls, even after adjustment for age, race, sex, and a creatinine-based measure of GFR. Therefore, it seems likely that inflammation rather than subclinical renal dysfunction is the major explanation for the increased concentrations of cystatin C in RA. MTX, but not other drugs commonly used to treat RA, was associated with lower cystatin C concentrations. Because serum creatinine did not differ significantly in patients using MTX, the lower cystatin C concentrations in MTX users are unlikely to be explained by the preferential prescription of MTX to patients with better renal function. MTX possibly affects the inflammatory pathways involved in the production or elimination of cystatin C.
In the Framingham Offspring Study, cystatin C was associated with several cardiovascular risk factors even in the absence of chronic kidney disease13. Our study confirmed the association between cardiovascular risk, measured as FRS, and cystatin C in RA. Generally, cystatin C has been associated with CACS in univariate but not multivariable analyses that adjusted for cardiovascular risk factors14,15. This is similar to our findings in RA.
Cystatin C concentrations are higher in patients with RA even after adjustment for renal function. Cystatin C provides information regarding the risk of atherosclerosis in RA, but this is not independent of the information provided by conventional cardiovascular risk factors.
Footnotes
-
Supported by National Institutes of Health grants HL65082, HL67964, GM07569, UL1 RR024975 from NCRR/NIH, P60 AR056116, and the Dan May Chair in Medicine.
- Accepted for publication June 6, 2011.