Renal dysfunction in acute stroke: an independent predictor of long-term all combined vascular events and overall mortality

Background. Acute stroke is the third leading cause of death in western societies after ischemic heart disease and cancer. Although it is an emergency disease sharing the same atherosclerotic risk factors with ischemic heart disease, the association of renal function and stroke is poorly investigated. The present study aims at assessing renal function status in patients with acute stroke and investigate any prognostic significance on the outcome. Methods. This is a prospective study of hospitalized first-ever stroke patients over 10 years. The study population comprised 1350 patients admitted within 24 h from stroke onset and followed up for 1 to 120 months or until death. Patients were divided in 3 groups on the basis of the estimated Glomerular Filtration Rate (eGFR) that was calculated from the abbreviated equation of the Modification Diet for Renal Disease in ml/min/1.73m(2) of body surface area: Group-A comprised patients who had eGFR > 60, group-B those with 30 <= eGFR <= 60 and group-C patients with eGFR < 30. Patients with Acute Kidney Injury (AKI) were excluded from the study. The main outcome measures were overall mortality and the composite new cardiovascular events (myocardial infarction, recurrent stroke, vascular death) among the 3 groups during the follow-up period. Results. Almost 1/3 (28.08%) of our acute stroke patients presented with moderate (group B) or severe (group C) renal dysfunction as estimated by eGFR. After adjusting for basic demographic, stroke risk factors and stroke severity on admission, eGFR was an independent predictor of stroke mortality at 10 years. Patients in groups B and C had an increased probability of death during follow-up: Hazard ratio = 1.21 with 95% CI 1.01-1.46, p < 0.05 and Hazard ratio = 1.76 with 95% CI 1.14-2.73, p < 0.05 respectively, compared to patients belonging to group A. The probability of death from any cause was significantly different among groups (log rank test 55.4, p = 0.001) during the follow-up period: in group-A patients it was 62.8 (95% CI 57.6-68.1), in group-B 77.3 (95% CI 68.5-86.1) and in group-C 89.2 (95% CI 75.1-100). During the follow-up period 336 new cardiovascular events occurred. The probability to have a new composite cardiovascular event was also significantly different among the 3 groups (log rank test 21.1, p = 0.001): in group-A patients it was 45.2 (95% CI 38.7-51.7), in group-B 67.4 (95% CI 56.2-78.6) and in group-C 77.6 (95% CI 53.5-100). Conclusion. Renal function on admission appears to be a significant independent prognostic factor for long term mortality and new cardiovascular morbidity over a 10-year period.