Aims: Uncertainties about the role of cystatin C-based estimated glomerular filtration rate (eGFR) in the prediction of cardiovascular disease (CVD) beyond traditional CVD risk factors remain. We assessed contributions of eGFR to CVD and mortality in the general population.
Methods: Using 14 year follow-up data on 9353 adults without a reported history of CVD from the Australian Diabetes, Obesity and Lifestyle study, we assessed the contributions of eGFR (assessed by cystatin C (eGFRcysC) and serum creatinine (eGFRcr) and albuminuria (uACR) to total and CVD mortality.
Results: After adjusting for age, sex, CVD risk factors and uACR, compared with an eGFRcysC >90 mL/min per 1.73 m2, eGFRcysC <60 mL/min per 1.73 m2 was associated with 56% and 73% increases in the risks for all-cause and CVD mortality, respectively. The respective changes for the c-statistic when eGFRcysC was added to a risk prediction model were 0.003 (95% confidence interval: 0.001 to 0.005) and 0.002 (95% confidence interval: −0.001 to 0.006). The net proportion of non-events assigned a lower-risk category significantly improved with the addition of eGFR (non-event net reclassification index eGFRcr: 1.0% and eGFRcysC: 1.5%) for all-cause mortality, but for CVD mortality, improvements were only significant when eGFR was combined with uACR. The net proportion of events assigned a higher-risk category was not significantly improved.
Conclusion: In our community-based cohort, reduced eGFRcysC was associated with all-cause and CVD mortality. The addition of chronic kidney disease measures to risk prediction models improved overall risk stratification among those at low risk as opposed to those at high baseline risk of mortality.