影響慢性腎病進展之相關因子探討

Abstract

Although there are numerous studies addressing the clinical parameters governing the evolution of chronic kidney disease (CKD) into end-stage renal disease (ESRD), the results are inconsistent. We examined the relationship between clinical characteristics and chronic kidney disease progression with different perspectives using a cohort in Taiwan. We enrolled 801 patients with chronic kidney disease stage 3 to 5 in one single medical center. We examined the different characteristics between CKD stages with application of cumulative and multinomial logit models. For longitudinal followed data, generalized estimating equation and generalized linear mixed model were adopted. We investigated the associations between covariates and time to ESRD by Cox regression models and checked the interaction between these covariates and CKD stages. We also applied time dependent Cox regression model for longitudinal followed data. We found changing relationships between certain covariates and CKD stages in different models. Most biomarkers showed consistently progressive nature with advance of CKD stages, while some differences exist only between certain stages. There was significantly higher hazard for ESRD among those patients with more advanced CKD stages. Plasma levels of albumin, calcium, phosphorous, hematocrit and amount of proteinuria are independent and consistent risk factors of ESRD in multivariate regression model. However, there were significant interactions between several factors and CKD stages. Most of the associations tend to be stronger in CKD 3-4 and weaker or even lack of significance in CKD stage 5. The estimates of hazard ratios for ESRD differed between time-independent and time-dependent models. This study provides solid evidence from domestic data about clinically pertinent risk factors in progression of CKD and clinical significance concerning CKD staging. These may form a basis for early intervention of modifiable factors to delay the progression of CKD.