Prognostic impact of symptom relief under low and normal physical performance in patients hospitalized for worsening heart failure

Abstract Background Subjective symptoms, represented by the New York Heart Association (NYHA) class, are well-established prognostic indicators and treatment goals of heart failure (HF). However, older patients may be less likely to develop subjective symptoms due to an impaired physical performance. Purpose To evaluate the prognostic impact of symptom relief under low and normal physical performance in patients hospitalized for worsening HF. Methods We carried out a sub-analysis of FRAGILE-HF, a prospective multicenter study, including 1272 consecutive older (>=65 years) hospitalized patients with HF (median age 80; 57.2% male; left ventricular ejection fraction 46±17%). NYHA class and gait speed were documented before discharge. The patients were prospectively followed-up for two years after discharge. Results At the time of discharge, 715 patients (56%) remained symptomatic (NYHA2-4) whereas 557 patients (44%) were asymptomatic (NYHA1). Patients were classified into 4 groups according to symptom relief and physical performance represented by slow (<0.8 m/s) or normal (>=0.8 m/s) gait speed: 279 slow-asymptomatic, 278 normal-asymptomatic, 402 slow-symptomatic, and 313 normal-symptomatic patients, respectively. During the follow-up period, 266 patients (20.9%) died. Slow-asymptomatic (hazard ratio (HR) 1.72[1.09-2.72], p=0.019 in comparison with normal-asymptomatic patients as a reference) and slow-symptomatic (HR 1.85[1.21-2.82], p=0.004) status was associated with high mortality in a Cox regression analysis adjusted for the Meta-analysis Global Group in Chronic Heart Failure score and B-type natriuretic peptide level (Figure). Furthermore, the combined endpoint of HF rehospitalization and all-cause death within 90 days were more frequently observed in slow-asymptomatic (HR 1.94[1.05-3.60], p=0.034), normal-symptomatic (HR 1.89[1.06-3.36], p=0.031), and slow-symptomatic (HR 1.97[1.10-3.52], p=0.022) patients than normal-asymptomatic patients (Figure). Conclusions Symptom relief achieved at the time of discharge was not associated with a lower mortality in patients with low physical performance, for whom careful follow-up may be needed especially in the early period after discharge because the symptoms may be masked due to low physical performance.