Aims: To evaluate the relationship between systolic blood pressure (SBP) and long-term mortality in patients with acute heart failure (AHF) stratified by ejection fraction (LVEF): reduced (< or =40%) vs. preserved (> or =50%).
Methods and results: We studied 1049 consecutive patients admitted with AHF. Systolic blood pressure was determined in the emergency department. Left-ventricular ejection fraction was categorized as < or =40% (n = 288), 41-49% (n = 174), or > or =50% (n = 587). Cox regression analysis was used for multivariable analysis. Mean age and SBP were 73 +/- 11 years and 150 +/- 36 mmHg, respectively. During a median follow-up of 18 months, 290 deaths (33.1%) were identified. Higher SBP was associated with lower mortality. In multivariable analysis, a differential effect of SBP across LVEF status was documented (P-value for interaction = 0.036). In linear models, SBP was shown to be inversely related with mortality in both groups (per 10 mmHg decrease): HR((LVEF > or = 50%)): 1.06, CI 95% = 1.01-1.11; P = 0.016, and HR((LVEF < or = 40%)): 1.16, 95% CI = 1.08-1.25; P < 0.001). When SBP was modelled with restrictive cubic splines, an inverse and almost linear relationship with mortality was shown in patients with LVEF < or =40% (P < 0.001), whereas in patients with LVEF > or =50%, SBP followed a J-shape curve.
Conclusion: In patients with AHF, SBP showed a differential prognostic effect on mortality according to LVEF status; when LVEF was < or =40%, SBP was linearly and inversely associated with mortality. Conversely, in patients with LVEF > or =50% this relationship showed a J-shape pattern.