Anemia is a highly prevalent comorbidity in patients with chronic heart failure (HF) and is associated with an increased risk of adverse outcomes (1). The reported prevalences of anemia in patients with HF were widely varying among different studies, ranging from 10% to 70%, mainly depending on the severity of HF and the definition of anemia (1–3). The etiologies of anemia in HF patients are multifactorial and heterogeneous, including inadequate erythropoietin levels due to renal dysfunction, hemodilution, hematinic deficiencies such as iron deficiency, bone marrow dysfunction, inflammation, and medications (4). Compared with chronic HF, progressive volume overload was more common in patients hospitalized with acute HF, which indicated that hemodilution might play a relatively larger role in anemia (5). Approximately half of the patients with acute HF had anemia (6); however, available data regarding the prognostic value of anemia in acute HF were relatively limited and conflicting. Amicis et al. and Rodríguez et al. reported that anemia was an independent predictor of mortality in patients with acute HF (7, 8). In contrast, Tymińska et al. found that the association of anemia with increased risk of mortality was significant in univariate analysis but no longer significant after adjustment for potential confounders (9). In the present study, we sought to assess the association of on-admission anemia with 1-year mortality in patients hospitalized with acute HF and evaluate whether anemia was an independent predictor of mortality in the acute setting of HF.

In this large-scale prospective study of patients hospitalized with acute HF, on-admission anemia was present in almost half of the patients. The association of anemia with increased 1-year all-cause and cardiovascular mortality was significant in the unadjusted model and attenuated after adjustment for most potential confounders but remained statistically significant. Only after additional adjustment for natriuretic peptides, on-admission anemia was no longer a significant predictor of all-cause or cardiovascular mortality. When considering the degree of anemia, moderate-to-severe anemia was an independent predictor of mortality after full adjustment, whereas mild anemia was not.

Most previous studies regarding the prognostic value of anemia in HF were conducted in patients with chronic HF (1, 14). A large-scale meta-analysis including 1,53,180 patients with chronic HF demonstrated that the crude mortality rate of anemic patients was twice that of non-anemic patients, and this association remained significant after adjustment (adjusted HR: 1.46, 95%CI: 1.26–1.69) (1). The anemic rate tends to be higher in patients with acute HF than in patients with chronic HF; however, the impact of anemia on mortality in the acute setting is controversial (7–9). In the present study, the anemic patients with acute HF had a 54% increase in the risk of all-cause mortality and a 58% increase in the risk of cardiovascular mortality before the adjustment; however, after the additional adjustment for natriuretic peptides, the association was no longer significant. Interestingly, the two previous studies (7, 8), which concluded that anemia was an independent predictor of mortality, did not include natriuretic peptides values in the multivariate Cox model.

The discrepancies in the prognostic value of anemia between acute and chronic HF may partly be due to the differences in the etiology of anemia (4). Several small studies in patients with chronic HF have established the role of hemodilution caused by volume overload in anemia. Androne et al. measured the red blood cell and plasma volume in 37 anemic patients with chronic HF using a radiolabeled albumin technique and found that anemia in 17 patients was due to hemodilution, and only 20 patients had true anemia (15). Previous studies also demonstrated that the increases in plasma volume and extracellular volume were independent predictors of low hemoglobin levels in patients with chronic HF (16, 17). Progressive volume overload and congestion are the main reasons for hospital admission in patients with acute HF (18). Compared to the chronic setting, greater volume expansion and hemodilution during HF decompensation may be partly responsible for the higher anemic rate in patients hospitalized with acute HF (16), thus leading to a relatively higher proportion of pseudo-anemia. This might be the reason why natriuretic peptides, the markers of elevated filling pressure and volume overload (19), were the critical confounding factor that influenced the impact of anemia on all-cause and cardiovascular mortality in the present study. This is to say that hemoglobin levels are probably more like a marker of volume overload in the acute setting of HF. Diuretics are the cornerstone of acute HF treatment (20). Some pseudo-anemia caused by transient hemodilution might be corrected after effective hemoconcentration during hospitalization (15). Meanwhile, effective hemoconcentration during acute HF hospitalization is a predictor of improved prognosis (21, 22) and thus may mask the impact of anemia on mortality. However, there were no available data on the dynamic changes of hemoglobin levels before and after discharge in the HERO study; therefore, further studies are needed to clarify the detailed association of anemia with hemodilution and hemoconcentration in the acute setting of HF.

In addition, anemic patients were older and were more likely to have impaired renal function, lower BMI, diabetes, hypertension, and higher NYHA class. In the present study, the rate of impaired renal function (eGFR <60 mL/min/1.73 m²) in patients with moderate-to-severe anemia was about twice that of those without anemia and about 1.5 folds that of those with mild anemia. The significant association of anemia with adverse outcomes in unadjusted model and in previous studies may also be the results secondary to a higher comorbidity burden. Our study found that the usage rates of renin-angiotensin system inhibitors and β-blockers in anemic patients were significantly lower than those without anemia. This may be explained by possible intolerance of the medications due to worse HF severity or more renal dysfunction in anemic patients.

The present study showed that moderate-to-severe anemia was an independent predictor of all-cause mortality. This may be because a higher proportion of patients with moderate-to-severe anemia have real anemia, and anemia in these patients may still exist even after effective hemoconcentration. From this perspective, the cut-off point of WHO criteria may be relatively high to define anemia in patients with acute HF. Whether anemia in HF was a treatable target or merely a marker of HF severity or other comorbidities had been debated (4). In patients with chronic HF and anemia, extraneous erythropoietin did increase the hemoglobin concentration but failed to improve the prognosis of HF (23). However, a meta-analysis of randomized controlled trials demonstrated that intravenous iron therapy might improve the symptoms and the prognosis of systolic HF (24). Our findings suggest that mild anemia in acute HF may be more a marker of volume overload, whereas moderate-to-severe anemia should draw the attention of clinicians. Further studies are needed to explore whether specific treatment such as intravenous iron therapy will improve prognosis in patients hospitalized with acute HF and moderate-to-severe anemia.

On-admission anemia, defined by the WHO criteria, is not an independent predictor of mortality in patients hospitalized with acute HF. Moderate-to-severe anemia in patients with acute HF is independently associated with increased mortality.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.

The studies involving human participants were reviewed and approved by The First Affiliated Hospital of Zhengzhou University. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual (s) for the publication of any potentially identifiable images or data included in this article.

JD, XD, and JL contributed to conception and design of the study. XW, XZ, LingL, and LiL participated in the collection of samples and data. JL and YL performed statistical analysis and all the results were checked by CJ and LiL. JL wrote the first draft of the manuscript. LiL, YL, XZ, XW, and CJ wrote sections of the manuscript. The manuscript was critically revised by CM, XD, and JD. All authors contributed to the article and approved the submitted version.

The study has been funded by the National Thirteenth 5-year Science and Technology Support Projects by the Ministry of Science and Technology of China (Grant No. 2016YFC1301000).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.