Figure 4.

Distribution of ECV% and Relationship With Clinical Events

ECV% is normally distributed (A). When divided into tertiles, both the all-cause mortality rate (red squares) and cardiovascular mortality (purple squares) progressively increased across the tertiles. On Kaplan-Meier analysis, there was a progressive increase in all-cause mortality across tertiles of ECV% (B) (p = 0.009). ECV% remained an independent predictor of all-cause mortality on multivariable analysis (C) (hazard ratio: 1.10; p = 0.013). Abbreviations as in Figure 3.