Abstract 18867: Multiple Biomarkers Predict Risk of Incident Heart Failure in Population Free of Cardiovascular Disease: The Multi-Ethnic Study of Atherosclerosis

Abstract

Background: Heart Failure is a complex syndrome with numerous patho-physiological mechanisms and altered biomarker profiles. Biomarkers that indicate risk for incident HF are less well established. Objective: To evaluate the predictive role of biomarkers reflecting different patho-physiological pathways for prediction of HF in the Multi-Ethnic Study of Atherosclerosis (MESA), a large multi-ethnic community based cohort.

Methods: We evaluated 6,814 Caucasian, African-American, Hispanic, and Chinese participants with mean age 62.3 ± 10.3 (51.4% female). We examined individual and joint associations of CRP, fibrinogen, NT-proBNP, homocysteine, and urine albumin to creatinine ratio (UACR) with incident HF using Cox proportional-hazard models adjusting for age, sex, ethnicity, systolic BP, antihypertensive therapy, smoking, diabetes, total cholesterol, BMI, left ventricular hypertrophy by ECG, and estimated glomerular filtration rate. Hazard ratios (HR) were calculated per 1-SD change in log (biomarker). A multivariable-adjusted stepwise procedure (p for retention < 0.05) was used to determine the independent contributions of the biomarkers for prediction of HF. We examined the increased model discriminative value with calculation of the C-statistic. Net reclassification index (NRI) was calculated with participants stratified in 3 risk categories; low (10%).

Results: During median follow-up of 7.6 years, 176 participants developed HF. In multivariable-adjusted models, 3 of 5 biomarkers were individually associated with HF; NT-proBNP, CRP, and UACR [HRs: 2.46, (95% CI 2.11-2.87), 1.28 (1.09-1.50), and 1.29 (1.14-1.45), respectively]. After stepwise selection only NT-proBNP [HR: 2.56 (95% CI 2.17-3.03)] and CRP [HR: 1.20 (95% CI 1.03-1.41)] remained statistically significant. The C-statistic significantly increased when NT-proBNP and CRP were added to the model (0.864 vs. 0.808). The addition of biomarkers improved NRI (NRI 0.29, p <0.0001).

Conclusions: CRP and NT-proBNP are significant independent predictors and substantially improve risk classification for incident heart failure. More research is warranted to determine the clinical utility of measuring these biomarkers in subjects at risk for incident heart failure.