NT-proBNP Linking Low-Moderately Impaired Renal Function and Cardiovascular Mortality in Diabetic Patients: The Population-Based Casale Monferrato Study

BACKGROUND: Few data are available to assess whether a low-moderate reduction in estimated glomerular filtration rates (eGFR) has a role per se on cardiovascular (CV) mortality or other biomarkers such as NT-proBNP allow to explain such association. METHODS AND FINDINGS: In a prospective study including 1,645 type 2 diabetic subjects of the population-based Casale Monferrato Study, who had no clinical evidence of heart failure and eGFR >45 ml/min/1.73 m(2), we examined 6 years CV mortality. Multivariate Cox proportional hazards modeling were used to estimate the effect of NT-proBNP on the association between eGFR and mortality, independently of baseline CV risk factors, albumin excretion rate (AER) and C-reactive protein (CRP). During follow-up, 327 people died (149 of CV diseases) out of 8334.5 person-years. Compared to eGFR≥90 ml/min/1.73 m(2), values of 60–89 and 45–59 ml/min/1.73 m(2) conferred a fully adjusted hazard ratios (HRs) of CV mortality of 1.74 (1.08–2.82) and 1.95 (1.03–3.68), respectively. After further adjustment for NT-proBNP, however, HRs were no longer significant (HRs 1.42, 0.83–2.42 and 1.22, 0.59–2.51). In this model, HR for logNT-proBNP was 1.84 (1.52–2.22). Adding NT-proBNP to the model improved the C-statistic of CV mortality from 0.79 (0.76–0.83) to 0.84 (0.81–0.87), yielded an IDI of 0.03 (p = 0.02), and a NRI of 0.44 (p = 0.016). CONCLUSIONS: In diabetic people a modest reduction in renal function increased 6-year CV mortality independently of albuminuria. This association, however, was mainly explained by the effect of NT-proBNP, that remained the strongest prognostic marker for a worse CV outcome, even after adjustment for other CV risk factors and pre-existing CVD.