| 초록 |
Both serum creatinine (Cr) and cystatin C (CysC) can be used to estimate glomerular filtration rate (eGFRCr and eGFRCysC). Nevertheless, the usefulness of eGFRCysC has not been fully determined. In this respect, it is an important issue when the discrepancy between eGFR trends from both parameters happens in patients with chronic kidney disease (CKD). A total of 1,069 patients whose Cr and CysC had been followed more than 4 years were included from the KoreaN Cohort Study for Outcome in Patients With Chronic Kidney Disease (KNOW-CKD) which enrolled predialytic patients with CKD. Trajectory analyses were performed based on the latent class mixed modeling. We defined the discrepancy as having decreasing eGFRCr but stable eGFRCysC. Multivariate logistic plus Firth’s penalized likelihood or generalized linear regression models were established to identify the condition related with the discrepancy. Trajectory pattern of eGFRCr was classified into three groups: two groups with stable eGFRCr (stable with high eGFRCr and stable with low eGFRCr) and one group with decreasing eGFRCr (Figure 1). Trajectory analysis of eGFRCysC also showed similar patterns such as two groups with stable eGFRCysC and one group with decreasing eGFRCysC. When the discrepancy trend (i.e., stable eGFRCysC but decreasing eGFRCr) was set up as the dependent variable, younger age, large proteinuria, and certain type of CKD (e.g, diabetic nephropathy and autosomal polycystic kidney disease) were selected as the predictors of the discrepancy trend. These results remained consistent while other regression models were additionally applied. The present study addresses the conditions related with discrepancy trend between eGFRCr and eGFRCysC, which should be notified in the clinics tracing both Cr and CysC. |