Background: The associations between decreased kidney function and cardiovascular disease recently have been established. However, there is uncertainty about the consistency between the independent associations of proteinuria as a risk factor across all cardiovascular end points. We undertook a meta-analysis of published cohort studies to provide a reliable estimate of the strength of association between proteinuria and risk of stroke.
Study design: Meta-analysis of observational cohort studies.
Setting & population: General population of participants with diabetes. Studies were excluded if participants had known glomerular disease or had undergone dialysis or transplantation.
Selection criteria for studies: MEDLINE, EMBASE, and CINAHL databases were searched for studies that reported age- or multivariate-adjusted risk ratio with some estimate of the variance of the association between proteinuria and risk of stroke, without language restriction.
Factor: Proteinuria or albuminuria.
Outcomes: Fatal or nonfatal stroke.
Results: Data from 10 published studies involving 140,231 participants and 3,266 strokes were eligible for inclusion. Participants with proteinuria had a 71% greater risk of stroke compared with those without proteinuria (95% confidence interval, 1.39 to 2.10). There was evidence of significant quantitative heterogeneity in the magnitude of the association across studies (I(2) = 60%; P for heterogeneity = 0.008), which was partially explained by differences in methods for measuring proteinuria. The risk of stroke remained significant after adjustment for other vascular risk factors.
Limitations: Because individual patient data were unavailable, we were unable to fully examine the impact of adjustment for known cardiovascular risk factors on the strength of the association between proteinuria and stroke risk. It is possible that the pooled estimate was affected by regression dilution bias.
Conclusions: These findings support the independent relationship between proteinuria and stroke. Additional studies are warranted to determine whether interventions to reduce proteinuria are effective at reducing rates of stroke.