Background and Purpose- Chronic kidney disease is strongly associated with stroke with various purported mechanisms proposed. Low glomerular filtration rate appears to be a risk factor for stroke independent of cardiovascular risk factors in epidemiological studies, but there has been no systematic assessment of the impact of more complete adjustment for blood pressure on the association. Methods- We did a systematic review to February 2018 (MEDLINE/EMBASE) for cohort studies or randomized controlled trials that reported stroke incidence in adults according to baseline estimated glomerular filtration rate. Study and participant characteristics and relative risks (RR) were extracted. Estimates were combined using a random-effects model. Heterogeneity was assessed by x2 statistics and I2 and by subgroup strata and meta-regression. Results- We identified 168 studies reporting data on 5 611 939 participants with 115 770 stroke outcomes. Eighty-five studies (3 417 098 participants; 72 996 strokes) provided adequate data for meta-analysis of estimated glomerular filtration rate and stroke risk. Incident stroke risk was increased among participants with estimated glomerular filtration rate <60 mL/min per 1.73 m2 (RR=1.73; 95% CI, 1.57-1.90; P<0.001), but there was substantial heterogeneity between studies (P<0.0001; I2, 78.5%). Moreover, the association was reduced after adjustment for cardiovascular risk factors, with progressive attenuation on more thorough adjustment for hypertension: single baseline blood pressure measure (RR=1.63; CI, 1.34-1.99; P<0.001); history or treated hypertension (RR=1.35; CI, 1.24-1.46; P<0.001); multiple blood pressure measurements over months to years (RR=1.10; CI, 1.02-1.18; P=0.01). Conclusions- The association between chronic kidney disease and stroke appears to be highly dependent on the method of adjustment for hypertension. The apparently independent relationship between chronic kidney disease and stroke may be confounded by their shared association with long-term prior blood pressure.
3085 - 3092
From the Center for Prevention of Stroke and Dementia, Nuffield Department of Clinical Neurosciences, John Radcliffe Hospital, University of Oxford, United Kingdom.