Heart failure (HF) is the ultimate stage of all heart diseases, affecting approximately 40 million individuals globally, with its incidence and mortality rates increasing annually [1, 2]. HF with preserved ejection fraction (HFpEF) is a subtype of HF, accounting for approximately half of all HF cases, with a 5-year survival rate of only approximately 50% [3, 4]. HFpEF has long been considered an independent subtype of HF. However, recent research has indicated that the pathophysiological mechanisms of HFpEF vary across different ranges of left ventricular ejection fraction (LVEF). Additionally, morphological and functional differences exist between them [5], which may result in distinct prognoses. Therefore, the purpose of this study was to evaluate the clinical characteristics and outcomes of HFpEF patients with different LVEF ranges to provide a basis for better individualized treatment approaches for HF patients.

While there has been a considerable amount of research investigating the clinical characteristics and outcomes of HFpEF, available data on HFpEF subtypes based on different LVEF ranges are limited. This is partly due to the categorization of HF patients with LVEF $>$50% as a single independent phenotype, which has, to some extent, restricted further observations of potential subtypes.

The focus of this study was to explore the clinical features and prognosis of HFpEF within different LVEF ranges. This study revealed significant differences in clinical characteristics and two-year survival rates among HFpEF subtypes with varying LVEF ranges. Compared to the high LVEF-HFpEF group, the low LVEF-HFpEF group exhibited higher diastolic blood pressure and faster HR, suggesting that there may be an underlying decrease in arterial elasticity and myocardial contractility in patients in the low LVEF-HFpEF group.

In the past, HFpEF was traditionally considered to be solely caused by diastolic dysfunction of the heart. However, recent research has revealed that HFpEF also encompasses significant impairment in systolic function and a restricted contractile reserve during periods of stress [7, 8]. When cardiac contractility declines, LVEF decreases, and the heart needs to beat faster to maintain adequate blood supply. Therefore, patients with lower LVEF may exhibit a higher HR as the disease progresses normally.

Furthermore, in comparison to the high LVEF-HFpEF group, patients in the low LVEF-HFpEF group had higher levels of serum creatinine and lower eGFR. Notably, there were no significant differences between the low LVEF-HFpEF group and the HFmrEF or HFrEF groups. This suggests that low LVEF-HFpEF may share similarities with HFmrEF and HFrEF. In this study, for each unit increase in eGFR, there was a 1.2% reduction in the risk of all-cause mortality and a 1.8% reduction in the risk of cardiovascular mortality among HF patients (Tables 2,3). The findings indicate a negative correlation between the eGFR and both all-cause and cardiovascular mortality in HF, which aligns with previous research [9, 10, 11]. This correlation may be related to improved renal perfusion at higher LVEF levels. The intricate relationship between the heart and kidneys has long been a subject of extensive research [12], involving mechanisms such as neurohumoral drive, autonomic reflexes, and fluid balance alterations, all of which collectively maintain circulatory and internal environmental homeostasis. However, the findings of this study underscore the differences in HFpEF subtypes within different LVEF ranges, possibly reflecting distinct pathological processes between these two HFpEF subtypes.

This study observed differences in survival rates among various HF phenotypes, with the high LVEF-HFpEF group having the lowest mortality rate and the HFrEF group having the highest. In contrast, the survival rates of the low LVEF-HFpEF and HFmrEF groups were relatively close. HFmrEF patients are typically situated on a dynamic trajectory transitioning from HFrEF to either improvement or deterioration [6, 13]. This suggests that low LVEF-HFpEF may also represent a clinically unstable phenotype. Therefore, dynamic monitoring of LVEF changes over time and actively searching for the causes of EF variation are crucial.

After PSM, the survival rate of the low LVEF-HFpEF group remained lower than that of the high LVEF-HFpEF group (Fig. 4A,B). This difference may be attributed to variations in cardiac morphology and pathophysiology between the two subtypes. Sebastian Rosch and colleagues analyzed patients with HFpEF whose LVEF fell within the ranges of 50%–60% and LVEF $>$60% using various diagnostic methods, including imaging (including echocardiography and cardiac magnetic resonance imaging), histology (myocardial biopsy), and hemodynamic catheterization (conduction catheter). Their findings indicated that HFpEF patients with LVEF in the range of 50%–60% exhibited reduced contractility, impaired ventricular-arterial coupling, and increased myocardial fibrosis. In contrast, patients with LVEF $>$60% exhibited a high-systolic state characterized by left ventricular afterload excess and reduced preload reserve, suggesting fundamental differences in cardiac morphology and hemodynamic responses between high LVEF-HFpEF and low LVEF-HFpEF [5]. This, to some extent, explains the varying prognoses observed in HFpEF patients with different LVEF ranges. Furthermore, even after accounting for competing risks, there still existed a difference in cardiovascular mortality between the two groups (Fig. 5A), consistent with the findings of Gu et al. [14]. However, it should be noted that with regard to noncardiovascular death, there was no significant difference between the high LVEF-HFpEF and low LVEF-HFpEF groups (Fig. 5B). This suggests that the two groups share some degree of homogeneity in certain aspects.

This study showed that HF patients with hypoalbuminemia were more likely to experience adverse outcomes. Hypoalbuminemia leads to decreased plasma oncotic pressure and effective circulating blood volume, exacerbating microcirculation and multiorgan dysfunction. Additionally, reduced serum albumin levels diminish antibody production and immune function while increasing the risk of various infections, further deteriorating the condition of HF patients [15]. Furthermore, elevated potassium (K+) levels were associated with an increased cardiovascular disease (CVD) risk in high LVEF-HFpEF patients. Blood potassium levels were reported to be independently associated with hospitalization and long-term mortality events in HF patients, exhibiting a “U”-shaped curve pattern [16]. Moreover, HF patients possess multiple risk factors for hyperkalemia, including advanced age, diabetes, chronic kidney disease, or metabolic acidosis, all of which are associated with an increased risk of hospitalization and mortality [17]. Therefore, the importance of monitoring blood potassium levels should not be underestimated in the management of HF patients.

This study did not reveal any significant impact of ACEI, ARB, or ARNI on the prognosis of HFpEF. Additionally, due to the timeframe of the study, sodium-glucose cotransporter-2 inhibitors (SGLT2is) were not included in this research study. Initially, recommended for treating HFrEF, recent studies have suggested that SGLT2is exhibit significant therapeutic potential for HF patients across all LVEF categories, in part due to their capacity to reduce blood volume in HF patients [18, 19]. Reducing blood volume can alleviate cardiac load, improve both systolic and diastolic function, decrease myocardial oxygen consumption, and potentially enhance cardiac structure and function. Hence, it is a valuable strategy for HF patients. However, HFpEF patients exhibit notable differences in hemodynamics and myocardial fibrosis between resting and exercise states, with patients having LVEF $>$60% showing reduced ventricular size and increased diastolic and systolic stiffness, potentially limiting their response to volume regulation [5]. This observation nicely explains the varying effects of diuretics on the prognosis of different HFpEF subtypes in this study, i.e., diuretics can reduce the risk of endpoint events in HFpEF patients with lower LVEF, while this effect is not observed in patients with higher LVEF values. Hence, reducing blood volume may not be the optimal approach when treating HFpEF patients with higher LVEF.

Furthermore, this study revealed that $\beta$-blockers can lower cardiovascular mortality in low LVEF-HFpEF patients, which differs partially from previous research conclusions. In most clinical trials targeting HFrEF patients, $\beta$-blockers have shown positive effects, extending patient survival and reducing overall mortality, cardiovascular mortality, and the incidence of HF readmissions [20, 21, 22, 23]. However, in HFpEF, $\beta$-blockers have not shown significant benefits and may even be harmful [24, 25]. This disparity may be due to the distinct pathological characteristics of HFpEF patients compared to HFrEF, including various factors such as ventricular diastolic and systolic reserve function, HR reserve and rhythm, atrial dysfunction, ventricular and vascular stiffness, impaired vasodilation, pulmonary artery hypertension, endothelial dysfunction, and various other complex interactions involving peripheral tissues such as skeletal muscle. These mechanisms vary in their degree of involvement in the development of HFpEF, making the treatment of HFpEF more challenging [8]. One reason for these differences might be the exclusion of patients who did not undergo follow-up echocardiography in this study, which introduced some selection bias. Moreover, the partial similarity between low LVEF-HFpEF and HFmrEF/HFrEF could also contribute to these findings. Nevertheless, the results of this study suggest that there are potential subtypes within HFpEF, and investigating the efficacy of existing drugs for different HFpEF subtypes might be a future avenue of research. Therefore, a deeper exploration of the characteristics of HFpEF subtypes and tailored treatment strategies for different subtypes may be a focus of future research.

In summary, recognizing the clinical significance of observed differences among HFpEF subgroups, our study underscores the heterogeneity of the HFpEF population and the necessity for tailored therapeutic strategies. The distinct clinical characteristics and prognoses of low LVEF-HFpEF versus high LVEF-HFpEF subgroups suggest that a one-size-fits-all approach is insufficient for managing HFpEF patients. The low LVEF-HFpEF subgroup may benefit from more aggressive management of risk factors. Optimal management of comorbidities such as hyperlipidemia, coronary artery disease, and cerebrovascular disease, alongside vigilant monitoring of renal function and cardiac structural changes (i.e., LVEDD and LVESD), could be key to improving prognosis in this subgroup. Moreover, attention to nutritional status to ensure appropriate body mass index and improvement of hypoalbuminemia would be beneficial. Conversely, the high LVEF-HFpEF subgroup may benefit more from primary prevention strategies and lifestyle modifications. This includes smoking cessation programs and targeted dietary interventions such as appropriate potassium intake.

(1) This is a single-center retrospective study. Although patients were rigorously selected based on inclusion and exclusion criteria, larger sample sizes and longer-term follow-up results are needed in the future to validate these findings. (2) A significant portion of HF patients in this study had received interventions at other health care institutions before their visits to the study center. For these patients, we did not have information on their EF before the intervention, and we did not obtain other information about them at that time, which could result in missing data bias. (3) This study primarily utilized noninvasive echocardiography to assess the LVEF in heart failure patients. Due to technical and resource limitations, we encountered challenges in collecting global longitudinal strain (GLS) data. Future research will aim to address these limitations to provide more comprehensive insights.

In conclusion, the results of this study support the notion that HFpEF with LVEF $>$60% and HFpEF with LVEF between 50% and 60% represent two distinct phenotypes. Patients with 50% $\le$ LVEF $\le$ 60% had a mortality rate similar to that of HFmrEF and better than that of HFrEF patients, suggesting that aggressive treatment of valvular heart disease and improvement in renal function may improve short-term prognosis and reduce mortality. On the other hand, for patients with LVEF $>$60%, diuretic therapy may not be the optimal treatment approach. In the future, practicing clinicians need to pay more attention to the differences between subtypes of HFpEF patients, especially in the selection of treatment strategies. We advocate for a stratified therapeutic approach to HFpEF patients tailored to the distinct clinical characteristics and risk profiles of different subtypes. A deeper understanding of the diversity within HFpEF subgroups is essential for developing more informed, precise, and effective management strategies.

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

JJS: collected data, interpreted, analyzed data, wrote and authored the manuscript; critical revision of important elements. KKS and YPS: patient follow-up management, data collection; methodology; critical revision of important content. LHH, YTW: Data analysis and statistics and textual embellishment of the manuscript. SXC, JG: supervision; study design and critical revision of the manuscript based on comments from other authors and feedback from reviewers. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

This study was reviewed and approved by the Ethics Committee of Hebei General Hospital, with informed consent waived (NO.2023142).

Not applicable.

This research received no external funding.

The authors declare no conflict of interest.