Cystatin C, creatinine, estimated glomerular filtration, and long-term mortality in stroke patients

Abstract

Renal dysfunction is associated with mortality in patients after ischemic stroke. Cystatin C is a potentially superior marker of renal function compared to creatinine and estimated glomerular filtration rate (GFR). In our observational cohort study, 390 Caucasian patients suffered from acute ischemic stroke (mean age 70.9 years; 183 women and 207 men) were included and prospectively followed up to maximal 56 months. Serum creatinine and cystatin C were measured at admission to the hospital; GFR was estimated according to CKD-EPI creatinine and CKD-EPI creatinine/cystatin equations. According to values of serum creatinine, estimated GFR and serum cystatin C patients were divided into quintiles. In the follow-up period, 191 (49%) patients died. For serum cystatin C and estimated GFR based on creatinine and cystatin C, the mortality and the hazard ratios for long-term mortality increased from the first to the fifth quintile nearly linearly. The associations of serum creatinine and estimated GFR categories based on creatinine with long-term mortality were J-shaped. As compared with lowest quintile of serum cystatin C, the fifth quintile was associated with long-term mortality significantly also after multivariate adjustment (age, gender, initial stroke severity, known risk factors for stroke mortality). In contrast, in adjusted analysis serum creatinine and estimated GFR (CKD-EPI creatinine and CKD-EPI creatinine/cystatin) were not associated with long-term mortality. In summary, serum cystatin C was independently and better associated with the risk of long-term mortality in patients suffering from ischemic stroke than were creatinine and estimated GFR using both CKD-EPI equations.

The Association between Glomerular Filtration Rate Estimated on Admission and Acute Stroke Outcome: The Shiga Stroke Registry

“…A more extensive investigation in the future on this particular issue will be of importance. In addition, we realize that our finding is inconsistent with other studies describing the “U”- or “J”-shaped association 12 , 14 ) . This may be due to inevitable differences in study population, eGFR equations, or study design.…”

“…Despite growing interest in similar subject, divulging the association either in overall stroke 7 – 9 ) , ischemic 10 – 14 ) or hemorrhagic subtypes 15 – 17 ) , in-between study findings have been contradictory, for example, while Yahalom et al 9 ) and Tsagalis et al 7 ) showed significant and positive result between low eGFR and mortality and poor outcome after stroke, the opposite was reported by Yang et al 8 ) . Result interpretations were complicated by small sample size 7 – 9 , 12 – 15 , 17 ) , non-population-based setting 7 – 9 , 11 , 12 , 14 ) , narrow-range eGFR category 7 , 11 , 13 – 15 ) , or lack of control for stroke severity 12 , 15 ) . The present study deals with these issues by employing data from a large-scale population-based setting registry, implementing wide-range eGFR category, and separating analyses between non-RRT and RRT patients.…”

Derivation of a Prediction Rule for Unfavorable Outcome after Ischemic Stroke in the Chinese Population

“…Many biomarkers have been shown to be associated with the prognosis of cerebral infarction, such as hemoglobin, albumin, total bilirubin, fasting blood glucose, creatinine, uric acid，prealbumin etc. [12,13,[33][34][35][36][37][38]. These biomarkers were analyzed comprehensively by logistic regression in our study.…”

Characterising the relationships between physiological indicators and all-cause mortality (NHANES): a population-based cohort study

“…Smoking status was assessed with serum cotinine levels. Socioeconomic status was defined by using the poverty income ratio as a categorical variable with five categories: [0,1], (1,2], (2,3], (3,4], and (4,5]. The reference group is poverty income ratio of one or less (ie, below the poverty line).…”

“…Most association studies consider linearity or non-linearity separately without quantitatively assessing which of the models better describes the relationship between a given physiological indicator and all-cause mortality. 5 , 6 A few studies have compared the prediction performance of various types of model such as linear, quadratic, cubic, and logarithmic but have not validated their findings, leading to model overfitting and lack of generalisation to other populations or cohorts. 7 Model flexibility has also led to the so-called vibration of effects, 8 whereby the direction of associations is a function of how variables are modelled, which is a threat to both generalisability and reproducibility.…”