Risk of recurrent stroke and dementia following acute stroke by changes in kidney function: results from the Perindopril Protection Against Recurrent Stroke Study

Abstract

Background: Limited data exist on the relationship between declining kidney function and cardiovascular events, dementia, and mortality in patients with a history of stroke. Thus the aims of the study were to investigate functional relationships between dynamic kidney function change and cardiovascular outcomes, and clarify whether adding kidney parameters to conventional cardiovascular risk factors improves model discrimination. Methods: Post hoc analysis of the Perindopril Protection Against Recurrent Stroke Study (PROGRESS) clinical trial of blood pressure lowering for the secondary prevention of stroke. We examined the association between dynamic kidney function defined as percentage change (declines of >30%, and >0 to ≤30%, and increases of ≥0 to <30%, and ≥30%) in estimated glomerular filtration rate (eGFR) over 2 years and recurrent stroke, major cardiovascular events, dementia and all-cause death over the next 2 years using Cox proportional hazard models controlling for eGFR at registration and potential confounders. Restricted cubic splines were used to assess the functional relationships. C-statistics and Net Reclassification Improvement (NRI) at 2 years were used to assess model discrimination. Results: In 4591 patients followed for a mean of approximately 2 years, 254 (5.5%) developed recurrent stroke, 391 (8.5%) had a major cardiovascular event, 221 (4.8%) developed dementia, and 271 (5.9%) died. Reverse J-like or U-like relationships were observed for percent declines in eGFR and outcomes. Using declines in eGFR of >0 to ≤30% as a reference, increased risks were evident for a greater decline (>30%) in relation to recurrent stroke [adjusted hazard ratio 1.85, 95% confidence interval (CI) 1.20–2.85], major cardiovascular event (2.24, 1.62–3.10) and all-cause death (2.09, 1.39–3.15). A larger increase (≥30%) in eGFR was also associated with a greater risk of all-cause death (1.96, 1.14–3.37). Improvements in the C-statistic were found by adding baseline eGFR and percent change compared with a model with conventional cardiovascular risk factors alone, for major cardiovascular events, dementia, and all-cause mortality. Conclusion: Declining kidney function following an incident cerebrovascular event is associated with additional risk of a major cardiovascular events, dementia, and 2-year mortality. However, a large increase in kidney function was also found to be associated with a higher risk of mortality.