Albuminuria, estimated glomerular filtration rate, and traditional predictors for composite cardiovascular and kidney outcome: a population-based cohort study in Korea.

Abstract

BackgroundCertain pharmacotherapies have shown to be effective for both cardiac and kidney outcomes. Although risk prediction is important in treatment decision-making, few studies have evaluated prediction models for composite cardiovascular and kidney outcomes.MethodsThis study included 2,195,341 Korean adults from a nationwide cohort for chronic kidney disease and a representative sample of the general population, with a 9-year follow-up. This study evaluated prediction models for a composite of major cardiovascular events or kidney disease progression that included albuminuria and estimated glomerular filtration rate (eGFR) and/or traditional cardiovascular disease predictors.ResultsThe addition of albuminuria and eGFR to a model for the composite outcome that included age, sex, and traditional predictors increased a C statistic by 0.0459, while the addition of traditional predictors to age, sex, albuminuria, and eGFR increased a C statistic by 0.0157. When age and sex-adjusted incidence rates were calculated across the combined Pooled-Cohort-Equations (PCEs) and Kidney Disease: Improving Global Outcomes (KDIGO) risk categories in diabetic or hypertensive participants, the incidence of ≥10 per 1,000 person-years was observed among all categories with high or very high KDIGO risk and among categories with moderate (or low) KDIGO risk and a PCEs 10-year risk of ≥10% (or ≥20%), accounting for 36% of diabetic and 18% of hypertensive populations.ConclusionThis study strongly supports the utility of the KDIGO risk matrix combined with a conventional cardiovascular risk score for the prediction of composite cardiovascular and kidney outcome and provides epidemiologic data relevant to the development of efficient treatment strategies.