LAUSR

INTRODUCTION Red cell distribution width (RDW) has been linked to cardiovascular disease. We sought to determine whether addition of RDW improved the Framingham risk score (FRS) model to predict cardiovascular mortality in a healthy US cohort. METHODS We performed a post-hoc analysis of the National Health and Nutritional Examination Survey-III (1988-94) cohort, including non-anemic subjects aged 30-79years. Primary endpoint was death from coronary heart disease (CHD). We divided the cohort into three risk categories: <6%, 6-20% and >20%. RDW>14.5 was considered high. Kaplan-Meier survival curves and Cox proportional hazards models were created. Discrimination, calibration and reclassification were used to assess the value of addition of RDW to the FRS model. RESULTS We included 7005 subjects with a mean follow up of 14.1years. Overall, there were 233 (3.3%) CHD deaths; 27 (8.2%) in subjects with RDW>14.5 compared to 206 (3.1%) in subjects with RDW≤14.5 (p<0.001). Adjusted hazard ratio of RDW in predicting CHD mortality was 2.02 (1.04-3.94, p=0.039). Addition of RDW to FRS model showed significant improvement in C-statistic (0.8784 vs. 0.8751, p=0.032) and area under curve (0.8565 vs. 0.8544, p=0.05). There was significant reclassification of FRS with a net reclassification index (NRI) of 5.6% (p=0.017), and an intermediate-risk NRI of 9.6% (p=0.011). Absolute integrated discrimination index (IDI) was 0.004 (p=0.02), with relative IDI of 10.4%. CONCLUSIONS Our study demonstrates that RDW is a promising biomarker which improves prediction of cardiovascular mortality over and above traditional cardiovascular risk factors.