Background: Chronic heart failure (CHF) in elderly patients has become an emerging healthcare issue in the current super-aging society. Thus, it is important to elucidate differences in prognosis and prognostic risk factors between elderly and non-elderly patients with CHF.
Methods and Results: We addressed this issue in 4,876 consecutive patients with Stage C/D CHF (mean 69 years-old, female 32%) in our CHART-2 Study by dividing them into 3 groups (G1, <64 years, n=1,521; G2, 65-74 years, n=1,510; and G3, ≥75 years, n=1,845). From G1 to G3, the prevalence of female increased (23, 31, to 40%, respectively) along with increase in left ventricular ejection fraction (mean 55, 57, to 58%, respectively) and plasma BNP levels (median 63, 93, to 158pg/ml, respectively) (all P<0.001). From G1 to G3, the incidence of death increased (23, 45, to 113/1,000 person-years) (P<0.001), where females had significantly lower non-cardiovascular (CV) death (1,000 person-years) than males in G2 (14.9 vs. 21.8, P=0.027) and G3 (39.2 vs. 55.9, P<0.001), but not in G1 (5.0 vs. 8.0, P=0.142). Since the proportion of non-CV death significantly increased especially after the age 65 years in both sexes, we compared the prognostic factors between patients with age <65 years (n=1,521) and those ≥65 years (n=3,355). In both groups, the multivariable Cox proportional hazard models showed that BNP levels and history of stroke were associated with CV death in both groups, while BMI levels and history of cancer with non-CV death in both groups. Importantly, patients ≥65 years had additional specific prognostic factors with CV and non-CV deaths that were not the case in those <65 years, including hemoglobin levels (hazard ratio (HR) 0.89, P<0.001), history of hyperuricemia (HR 1.40, P=0.003) for CV death and hemoglobin levels (HR 0.90, P=0.001), albumin levels (HR 0.58, P<0.001), history of hyperuricemia (HR 1.35, P=0.006), and history of stroke (1.70, P<0.001) for non-CV death. Thus, although the incidence of death increased particularly after age of 75 years, mode of death and prognostic factors dramatically changed after age of 65 years.
Conclusions: These results indicate that CHF management to prevent both CV and non-CV deaths should be initiated at least as early as at age 65 years in the elderly.