Temporal Trends and Predictors of Mortality in Heart Failure with Reduced Ejection Fraction: The Role of Renal Function and Left Ventricular Improvement

Background/Objectives: Traditionally, risk stratification in heart failure with reduced ejection fraction (HFrEF) patients has been a static process. This study aims to identify clinical predictors and causes of mortality at different stages of the disease, distinguishing between early mortality (≤1 year) and late mortality (>1 year), as well as cardiovascular and non-cardiovascular causes. Methods: This is an observational, longitudinal study of a cohort of 1510 patients with HFrEF followed up in a specialist heart failure unit between 2011 and 2023. Patients were categorized into three groups: alive (n = 1093), early mortality (n = 55) and late mortality (n = 362). Multivariate Cox regression models were used to identify mortality independent predictors in each period. Results: During the first year, cardiovascular and non-cardiovascular mortality showed a relatively balanced distribution, with a slight predominance of non-cardiovascular causes (52.8%). Progression of heart failure remained the leading individual cause of early death (38.2%). Independent predictors of early cardiovascular death included ischemic etiology (HR 2.78; 95% CI: 1.20–6.40; p = 0.01), NYHA functional class (HR 4.38; 95% CI: 2.18–8.81; p < 0.001) and renal function (HR 0.60; 95% CI: 0.44–0.84; p = 0.002). In contrast, late mortality (>1 year) was predominantly due to non-cardiovascular causes (59.7%), notably cancer (24.3%) and infections (7.5%). Late all-cause mortality was independently associated with age (HR 1.35; 95% CI: 1.17–1.56; p < 0.001), hypertension (HR 1.38; 95% CI: 1.06–1.81; p = 0.015) and ischemic etiology (HR 1.45; 95% CI: 1.16–1.80; p < 0.001). In the follow-up, the improvement of left ventricular ejection fraction (LVEF) emerged as a protective factor against long-term all-cause mortality (HR 0.77; 95% CI: 0.63–0.94; p = 0.01) and, even more markedly, against late cardiovascular mortality (HR 0.47; 95% CI: 0.33–0.66; p < 0.001), whereas baseline LVEF at diagnosis lacked significant predictive value in either model. Conclusions: Mortality in HFrEF is a dynamic process. Early survival is determined by hemodynamic stability, renal function and functional severity, whilst late survival is predominantly driven by non-cardiovascular comorbidities, biological aging, and ventricular remodeling trajectory. The improvement of LVEF during follow-up is a more reliable prognostic marker than the baseline measurement, suggesting that clinical management should shift from early pharmacological optimization towards a multidisciplinary approach focused on long-term non-cardiovascular risk.