Introduction

Optimal conventional therapy of heart failure (HF) remains insufficient to improve the quality of life of many patients with comorbidities. As consequence, they are susceptible to repeated hospitalisations [1]. Iron deficiency (ID) is one such comorbidity and affects approximately 55% of patients with HF irrespective of the range of left ventricular ejection fraction (LVEF) or the coexisting anaemia [2]. It is well known that ID contributes to worsening mitochondrial dysfunction, already present in HF, and promotes the metabolic shift from fatty acids to glucose use. Consequently, myocardial ID can cause progressive left ventricular remodelling and can lead to a decrease in exercise capacity [3]. Furthermore, there is a hypothesis that ID may play a key pathophysiological role in the progression of chronic HF as it enhances the activity of the sympathetic nervous system [4]. According to European Society of Cardiology (ESC) 2021 guidelines [2], all HF patients should be periodically screened for the presence of anaemia and ID by using full blood count, serum ferritin and transferrin saturation (TSAT) measurement. It was found out that i.v. iron decreases the risk of hospitalisation in patients with HF with reduced ejection fraction (HFrEF) [2]. Although the impact of ID on HF with preserved ejection fraction (HFpEF) was also confirmed, further investigation is still needed [2, 5]. This study aimed to assess the significance of ID for the population with HF, especially with HFpEF.

Material and methods

The study was approved by the Bioethical Commission of Medical University (MU) of Łódź, the resolution number RNN/25/22/KE dated 8 February 2022. The retrospective analysis included 150 patients (69 women and 81 men) with heart failure (HFpEF = 44.67%; n = 67) treated between 2018 and 2021 in the Department of Non-invasive Cardiology, MU of Łódź. The average patient’s age was 72 ± 10.55 years. Collected data included a standard set of diagnostic tests, such as electrocardiogram, echocardiography, 6 minute walking test (6MWT) and blood tests. The information about patients’ body mass index, HF aetiology, number of recurrent HF hospitalizations and concomitant diseases or conditions were also scored. The diagnosis of HFrEF, HFpEF and ID were established using the ESC 2021 guidelines criteria [2].

Patients were analysed related to ID and EF < 50 vs. ≥ 50% using appropriate statistical tests. The normality of distribution for continuous variables was tested using the Shapiro-Wilk test. To compare continuous variables, the Mann-Whitney U test or Student’s t-test were used, as applicable. Continuous variables were presented as mean ± standard deviation or together with the interquartile range as applicable. Nominal variables were tested using the Chi-square test, with Yates correction where needed. Statistically significant nominal variables were expressed as percentages.

To assess which variables were associated with hospitalizations due to HF in one year, univariable analysis using Spearman’s correlation coefficient and multivariable
logistic regression using the backward stepwise method was performed. The correlation was evaluated using the R coefficient and the degree of change in the dependent variable was assessed with beta coefficients. A p-value of < 0.05 was statistically significant. Statistical analysis was performed using Statistica 13.1PL (StatSoft, Tulsa, USA).

Results

Iron deficiency was diagnosed in 86 (57.33%) patients, including 40 (59.7%) HFpEF patients. Therefore, numerous comorbidities were noticed — among others arterial hypertension (86%; n = 129), type 2 diabetes (38%; n = 57), gastroduodenal ulcer disease (11.33%; n = 17), chronic obstructive pulmonary disease (11.33%; n = 17) and cancer history (13.33%; n = 20).

Clinical variables in relation to iron status are summa­rized in Table 1. Significant differences between HF patients with and without ID were observed not only in red blood cell parameters or other laboratory test results (urine acid, sodium ions) but also in 6MWT and stroke/transient ischemic attack (TIA) history (Table 1).

However no significant differences in iron status between patients with EF < 50% and EF > 50% were found (serum iron: 12.40 μg/dL vs. 11.46 μg/dL; p = 0.224; ferritin: 161.96 µg/L vs. 135.47 µg/L; p = 0.288; TSAT: 21.94% vs. 20.37%; p = 0.523).

Spearman’s rank correlation analysis revealed a ne­gative correlation between the number of hospitalizations due to HF in the last 12 months and red blood cell count, Hb, Ht and 6MWT (Table 2), and a positive correlation between the number of hospitalisations due to HF in the last 12 months and red cell distribution width (RDW) (Table 2). Associations between HF stage (lower echocardiographic parameters such as EF and tricuspid annular plane systolic excursion or elevated troponin and NT-proBNP levels) and repeated HF hospitalizations were also found (Table 2).

The regression analysis revealed the independent variables for HF hospitalization in the 12 last months were NT-proBNP and RDW but not ID (Table 3). Both parameters were positively correlated with the number of hospitalizations (b = 0.263 and b = 0.346, respectively). The generated model explained 67% of the variance in the outcome variable (r = 0.670; p < 0.005).

Discussion

The present study showed that in patients hospitalized for HF those with ID had more comorbidities than those without ID. Furthermore, ID diagnosis was associated with a higher prevalence of a stroke/TIA in the study population. According to the authors’ knowledge, this crucial result
in research on ID in HF was observed for the first time.

According to the ESC 2021 guidelines [2], ID in patients in HF is defined as either serum ferritin concentration
< 100 ng/mL or 100–299 ng/mL with TSAT < 20%. The most common reasons for ID in HF patients are reduced iron intake, reduced iron absorption, increased iron loss, which might be an effect of malignancy or gastritis/ulceration, and finally impaired iron release [14]. The diagnosis of ID does not require any specialist equipment and can be performed in any clinical centre. This is a significant feature that makes ID diagnosis widespread. However, in Poland similar to other countries, ID is recognized too rarely.

As in Beale et al. [5], the prevalence of ID in HFpEF is higher than in patients with EF < 50%. It may suggest a higher overall incidence of ID in HFpEF. Hence, the high prevalence of ID emphasizes the importance of performing iron measurements in patients with HFpEF and HFrEF alike.

Previously, the association between ID and cerebrovascular diseases was mostly reported in children and in relation to co-existing anaemia [6]. Nowadays, ID anaemia (IDA) is thought to be an etiological factor for stroke in the paediatric population [7]. However, Tang et al. [8] identified the connection between certain factors such as ID or IDA and ischemic stroke (IS) in the general population. This conclusion was based on transferrin’s ability to enhance thrombin and restrain the inactivation of coagulation proteases by antithrombin. Analogically, Shovlin et al. [9] noticed probable stroke pathogenesis as increased platelet aggregation responses due to ID. Thus, low serum iron levels were proposed to double age-adjusted stroke risk. Potaczek et al. [10] highlighted that ID, independently of anaemia, may be a risk factor for recurrent venous thrombo­embolism. Moreover, Dubyk et al. [7] suggested that IDA and possibly sub-clinical ID may cause TIA and IS including in patients with cardiac diseases.

Subsequently, ID in the present population was associated with impaired exercise tolerance, which was indicated by a lower 6MWT result. It is consistent with previous research and was proved by Jankowska et al. [11] and confirmed in patients with HFpEF by Bekfani et al. [12].

The increased RDW level as a significant marker of long-term mortality in patients admitted to the hospital with acute HF is also worth emphasizing [15]. Therefore, the positive correlation between the number of hospitalizations due to HF in the last 12 months and RDW from the present study strengthens the abovementioned conclusion. This biomarker is an effective indicator of the worsening and prognosis of HF [13].

The positive correlation between NT-proBNP level and the number of hospitalizations in the last 12 months is consistent with data which emphasize the role and high accuracy of natriuretic peptides in stratifying risk of the hospitalization due to HF [16].

Study limitations

There are several limitations of this study.

First, most studies regarding stroke history in HF patients were focused on IDA impact and not on sub-clinical ID. In consequence, it was not possible to precisely compare the results of this study with previous research.

Subsequently, the size of this study group was rather small, and the diagnosis of ID preceding cerebrovascular accident was not always confirmed. Therefore, further research in this field is needed.

Moreover, the design is that of a single clinical department, retrospective study, with potential selection bias; however, each patient admitted to the authors’ department was screened regarding ID. This leads to the conclusion that results from other clinical centres are necessary.

Conclusions

The present study shows the importance of ID as a comorbidity in HF.

The most important finding from this study is a higher incidence of TIA/stroke in patients with ID. Therefore, ID screening and effective treatment of this condition are crucial for patients with HF.

Conflict of interest

None declared.

Funding

The authors have no funding to declare.