Is a reduced estimated glomerular filtration rate a risk factor for stroke in patients with type 2 diabetes?

Although chronic kidney disease is a risk factor for cardiovascular disease it is unclear whether diabetic patients with a reduced glomerular filtration rate (GFR), independent of (micro)albuminuria, carry an increased risk of stroke. We therefore investigated the independent effect of estimated GFR (eGFR) on stroke events in patients with type 2 diabetes mellitus (T2DM). We studied T2DM patients with an eGFR >or=15 ml min(-1) per 1.73 m(2), who had no history of stroke. Patients were divided into four categories by the eGFR at baseline for comparison: >or=90, 60-89, 30-59 and 15-29 ml min(-1) per 1.73 m(2). The end point was an incident stroke event. The Cox proportional hazard model was used to calculate the hazard ratio (HR) and 95% confidence interval (CI). The study included a total of 1300 T2DM patients (546 women and 754 men) with a mean (+/-s.d.) age of 63+/-13 years. During a mean follow-up period of 3.7+/-1.4 years, 91 patients experienced an incident stroke event. Although a lower eGFR was associated with an increased stroke risk using a univariate model, statistical significance disappeared after adjusting for other risk factors including albuminuria. The HR (95% CI) was 0.75 (0.40-1.41, P=0.373), 0.99 (0.50-1.95, P=0.964) and 0.91 (0.36-2.28, P=0.844) for patients with eGFRs of 60-89, 30-59 and 15-29 ml min(-1) per 1.73 m(2), respectively, compared with patients with an eGFR >or=90. Clinical albuminuria remained a significant risk factor for stroke, and the adjusted HR compared with normoalbuminuria was 2.40 (1.46-3.95, P=0.001). In conclusion, the association between reduced GFR and stroke events in patients with T2DM is likely to be mediated by albuminuria.