国家神经系统疾病临床医学研究中心数据管理与共享系统

发表杂志：Journal of the American Heart Association

发表类型: Article

影响因子(IF): 5

PMID: 39817528 Pubmed原文链接 全文下载

Renal dysfunction is associated with systemic inflammation, endothelial dysfunction, accelerated atherosclerosis, and enhanced thrombogenicity, thus increasing the risk for cardiovascular disease (CVD) and ischemic stroke. Previous studies also suggested that stroke patients with low estimated glomerular filtration rate (eGFR) are at increased risk of all-cause mortality, stroke recurrence and disability. However, most of the previous studies assessed baseline eGFR at a single time point. Although some studies focused on the impact of acute kidney injury (AKI) or acute kidney disease (AKD) on stroke outcomes, the definition of AKI or AKD cannot differentiate the decrease or increase of eGFR. Two studies have investigated the fluctuation in eGFR during hospital stay in stroke patients. The study by Vart, P et al reported that >5% decline or increase in eGFR were at substantially higher risk of poor stroke outcomes during a mean follow up of 2.9 years. While Tutuncu S et al reported only decreasing eGFR have increased risk for recurrent vascular events and death.But the gap between 2 eGFR measures in these two studies were 2 days and 8 days respectively, which is close to the definition of AKI duration. There has been a growing interest in the association between the annual change in kidney function and risk of adverse outcomes. Previous large scale studies, including the Cardiovascular Health Study (CHS) and Atherosclerosis Risk In Communities (ARIC) cohort, have reported that the annual declining and increasing in eGFR were both independently associated with CVD, stroke incidence and mortality. However, whether the long-term change of eGFR after ischemic stroke could be a prognostic factor is not investigated On the other hand, until now, the research focus on brain and kidney cross talk has focused on the role of kidney impairment in inducing brain dysfunction. However, the effects of brain injury on kidney function remain unexplored. Besides the use of imaging contrast agents and mannitol or the withdraw of antihypertensive agents, cerebrovascular diseases per se can stimulate kidney dysfunction as kidney activity is regulated by the brain through neural pathways and peripheral humoral pathways. Considering the limited data in this area, studies investigating the post-stroke change of eGFR and its related factors are required to verify the brain-kidney interaction and to identify potential mechanisms which may enable development of novel therapeutic interventions. Compared with the eGFR estimated by creatinine, cystatin C captures earlier and more subtle changes of kidney function, and possibly reflects a preclinical status for kidney disease, which was reported to strengthen the association between the eGFR and adverse outcomes. Moreover, due to loss of muscle mass in chronically illness, especially in disability after stroke, serum creatinine based eGFR may overestimate their kidney function. The Third China National Stroke Registry (CNSR-III) was a cohort study in patients with acute ischemic stroke or transient ischemic attack (TIA). Two renal function markers, serum creatinine and serum cystatin C, were centrally tested on admission, at 3-month and at 1-year follow-up. Therefore, by thorough evaluation of eGFR based on serum creatinine (eGFRsCr), serum cystatin C (eGFRsCysC), and combination of serum creatinine and cystatin C (eGFRsCr+sCysC), we attempt to investigate whether 3-month and 1-year change of eGFR would be associated with 5-year stroke outcome in this cohort. Also, we intend to describe the change of eGFR in the acute phase (3-month) and chronic status (1-year) after stroke to investigate the effect of stroke on renal dysfunction and the potential mediators in this process.

Baseline characteristics were compared among the eGFR quintiles. Cochran–Armitage test was used to test for trend for categorical variables and general linear regression continuous variables. Using the Q3 group as a reference, the associations between eGFR changes and all-cause mortality or stroke recurrence were evaluated using multivariable Cox proportional hazard regression models, with hazard ratios (HRs) and 95% CIs calculated. We also used zero-hypothesis test to verify proportional hazards assumption in all models for all-cause mortality. Multivariate logistic regression was used to determine stroke disability outcomes with odds ratios (ORs) and 95% CIs. Model 1 was adjusted for age and sex. Model 2 was further adjusted for medical history, including hypertension, coronary artery disease, diabetes, atrial fibrillation, ischemic stroke, current smoking status, heavy drinking, body mass index, high-density lipoprotein cholesterol level, low-density lipoprotein cholesterol level, triglyceride level, baseline National Institutes of Health Stroke Scale score, TOAST subtypes, and baseline eGFR. Model 3 was further adjusted for the baseline log ACR (log-transformed). Finally, C-statistics, integrated discrimination improvement, and the net reclassification index were used to assess the improvement in model performance by adding eGFR changes to the conventional model (risk factors in Model 3) for prognosis prediction.