Phenotype-specific prognostic implications of pulse pressure in heart failure

The prognostic significance of pulse pressure (PP) differs across HF phenotypes, but its relevance in HFmrEF remains uncertain. We evaluated associations between PP and clinical outcomes across HF phenotypes and assessed whether PP provides prognostic information beyond SBP.

We performed a patient-level pooled analysis of 13 randomised HF trials including 35,550 patients with HFrEF, 6,484 with HFmrEF, and 19,173 with HFpEF. PP-outcome associations were assessed using Cox models for quartiles and Poisson regression with restricted cubic splines for continuous PP. Across HF phenotypes, patients with higher PP were older, more frequently female, and had higher LVEF and greater adiposity. In HFrEF, PP was inversely associated with first HF hospitalization or cardiovascular death, with the lowest event rate in the highest PP quartile (>56 mmHg; 13.3 per 100 person-years [95% CI, 12.8-13.8]) and the highest event rate in the lowest quartile (<40 mmHg; 18.2 [17.5-18.9]). In contrast, HFpEF demonstrated a J-shaped relationship, with the lowest event rate in the second quartile (48-56 mmHg; 6.8 [6.4-7.3]) and the highest in the fourth quartile (> 66 mmHg; 9.7 [9.1-10.3]). HFmrEF exhibited a similar J-shaped association, closely paralleling HFpEF. PP improved model fit beyond SBP, but the incremental improvement in discrimination was small and largely limited to HF hospitalization-related outcomes in HFpEF.

PP has phenotype-specific prognostic implications in HF. Low PP identifies higher risk in HFrEF, whereas high PP identifies higher risk in HFpEF and HFmrEF. PP may support phenotype-specific risk stratification.