LVEF-based phenotyping in very old adults hospitalised for heart failure: cardiological discrimination, shared geriatric burden, and treatment gaps.

METHODS: We conducted a retrospective observational cohort study in a cardiogeriatric unit to compare cardiological, geriatric, biological, and therapeutic profiles across LVEF phenotypes in very old adults hospitalised for heart failure (HF). All patients with HF and classifiable LVEF were included and categorised according to 2021 European Society of Cardiology (ESC) criteria as heart failure with reduced ejection fraction (HFrEF, ≤40%), heart failure with mildly reduced ejection fraction (HFmrEF, 41%-49%), and heart failure with preserved ejection fraction (HFpEF, ≥50%). The primary objective was to describe and compare cardiological, biological, geriatric, and therapeutic profiles across LVEF phenotypes. The comprehensive geriatric assessment (CGA)-derived geriatric indicator count was used as a descriptive measure of multidimensional geriatric burden, and not as a validated frailty instrument. RESULTS: Among 1,299 patients [median age 88.5 years (IQR 82.5-92.2)], HFpEF predominated (58.8%). Despite markedly different cardiological profiles, CGA-derived geriatric burden was similar across phenotypes: the study-defined geriatric indicator count was identical in all groups [median 2 (1-3); p = 0.915; η²=0.00014]. Iron deficiency affected 52.8% of patients, with no significant difference across phenotypes (p = 0.782). Clinically significant anaemia with concomitant iron deficiency was more frequent in HFpEF than in HFrEF (16.1% vs. 9.5%; p = 0.001), whereas clinically significant anaemia without iron deficiency was similar across groups (p = 0.986). In HFrEF, 52.6% received ≥3/4 guideline-directed therapeutic pillars (defined as ACEi/ARB/ARNi, beta-blocker, MRA, and SGLT2i). Overall ACEi/ARB/ARNi pillar coverage was 53.2% in HFrEF, but ARNi uptake and MRA prescription declined with increasing age, with ARNi uptake falling from 44% before age 85 to 21% after age 90. CONCLUSIONS: LVEF category remains central to pharmacological treatment selection but does not discriminate CGA-derived geriatric burden or iron deficiency burden in very old HF patients. These findings support systematic comprehensive geriatric assessment and iron status assessment regardless of LVEF phenotype, alongside phenotype-informed pharmacological optimisation.