Dyspnea in heart failure: pathophysiology, clinical assessment, and evidence from clinical trials Authors: Mateusz Garus, Mateusz Guzik, Rafał Tymków, Piotr Gajewski, Robert Zymliński, Jan Biegus Published: 2026 ABSTRACT Dyspnea is a key symptom of heart failure (HF), particularly in its acute form, where it represents the most common cause of hospitalization. Although traditionally attributed to pulmonary congestion, dyspnea in HF is a multidimensional phenomenon resulting from complex interactions among elevated cardiac filling pressures, hemodynamic alterations, fluid redistribution, impaired gas exchange, activation of mechano- and chemoreceptors, and maladaptive neurohormonal activation. Comorbidities, including chronic lung disease, obesity, cachexia, renal dysfunction, anemia, and chronic inflammation, further modulate symptom perception and lower the threshold for its occurrence. In this review, we synthesize current knowledge on the pathophysiological mechanisms underlying dyspnea in HF, emphasizing its heterogeneity and the complex interplay between hemodynamics, ventilatory regulation, and peripheral factors. We discuss available dyspnea assessment tools, highlighting their strengths and methodological limitations in clinical practice and randomized trials. Despite the use of standardized scales, dyspnea remains a subjective and dynamically evolving symptom. Data from large clinical trials in acute HF indicate that early improvement in dyspnea is common but does not translate into reduced mortality or rehospitalization, whereas residual dyspnea identifies patients at higher risk of adverse events. A better understanding of its multidimensional pathophysiology, measurement limitations, and prognostic significance has important implications for the interpretation of clinical trial results and for optimizing the assessment and management of this fundamental symptom. Introduction Dyspnea is one of the most common, universal, and characteristic symptoms in heart failure (HF). Different forms of dyspnea have been described in HF (Table 1) and are clinically useful tools for its assessment. Dyspnea represents the leading cause of presentation to hospital emergency departments in the acute heart failure (AHF) setting [1, 2]. The sudden onset or its rapid progression reflects complex, dynamic disturbances involving hemodynamic, respiratory, and neurohormonal pathways that accompany decompensation. Understanding the mechanisms underlying dyspnea is essential for accurate clinical assessment and for selecting appropriate and effective therapeutic strategies.Table 1Types of dyspnea in heart failureType of dyspneaExplanation of the conceptClinical Features Exertional dyspneaIt occurs during physical activity due to an insufficient increase in cardiac output, elevated pulmonary capillary wedge pressure, progressive pulmonary congestion, muscle hypoxia, and other complex mechanisms described below.Initially with heavy exertion, later with minimal activity, relieved by rest.Severity is assessed using the four-class New York Heart Association (NYHA) functional classification, described below. Resting dyspneaIt occurs at rest and reflects a marked exacerbation of the underlying pathophysiological mechanisms described below, resulting in dyspnea.Persistent dyspnea is present even at rest, regardless of physical activity. Paroxysmal nocturnal dyspneaSudden episodes of severe breathlessness at night caused by fluid redistribution.Awakens 1–3 h after falling asleep. Patient needs to sit or stand to relieve symptoms. OrthopneaWorsening dyspnea in the supine position due to increased venous return and left ventricular overload.Patient sleeps with multiple pillows. Symptoms improve when sitting upright. BendopneaDyspnea occurring when bending forward (e.g., when tying shoes) due to a sudden increase in preload and other contributing mechanisms.Seen in advanced heart failure, develops within seconds of bending. TrepopneaMay result from asymmetric pulmonary congestion or coexisting lung disease.Dyspnea that worsens in a specific lateral decubitus position. PlatypneaIt may occur in right-to-left shunting through a patent foramen ovale (PFO), pulmonary arteriovenous malformations, hepatopulmonary syndrome, pulmonary embolism, aortic dilation, constrictive pericarditis, pericardial effusion.Dyspnea worsens when standing or sitting and decreases when lying down. Table 1Types of dyspnea in heart failureType of dyspneaExplanation of the conceptClinical Features Exertional dyspneaIt occurs during physical activity due to an insufficient increase in cardiac output, elevated pulmonary capillary wedge pressure, progressive pulmonary congestion, muscle hypoxia, and other complex mechanisms described below.Initially with heavy exertion, later with minimal activity, relieved by rest.Severity is assessed using the four-class New York Heart Association (NYHA) functional classification, described below. Resting dyspneaIt occurs at rest and reflects a marked exacerbation of the underlying pathophysiological mechanisms described below, resulting in dyspnea.Persistent dyspnea is present even at rest, regardless of physical activity. Paroxysmal nocturnal dyspneaSudden episodes of severe breathlessness at night caused by fluid redistribution.Awakens 1–3 h after falling asleep. Patient needs to sit or stand to relieve symptoms. OrthopneaWorsening dyspnea in the supine position due to increased venous return and left ventricular overload.Patient sleeps with multiple pillows. Symptoms improve when sitting upright. BendopneaDyspnea occurring when bending forward (e.g., when tying shoes) due to a sudden increase in preload and other contributing mechanisms.Seen in advanced heart failure, develops within seconds of bending. TrepopneaMay result from asymmetric pulmonary congestion or coexisting lung disease.Dyspnea that worsens in a specific lateral decubitus position. PlatypneaIt may occur in right-to-left shunting through a patent foramen ovale (PFO), pulmonary arteriovenous malformations, hepatopulmonary syndrome, pulmonary embolism, aortic dilation, constrictive pericarditis, pericardial effusion.Dyspnea worsens when standing or sitting and decreases when lying down. Types of dyspnea in heart failure Initially with heavy exertion, later with minimal activity, relieved by rest. Severity is assessed using the four-class New York Heart Association (NYHA) functional classification, described below. The pathophysiology of dyspnea in AHF is multifactorial and extends beyond the simplistic concept of pulmonary congestion alone. It encompasses interactions among elevated intracardiac filling pressures, transudation of fluid into the pulmonary interstitium and alveolar spaces, impaired gas exchange, and activation of pulmonary and central mechanoreceptors and chemoreceptors. Additional contributing factors include alterations in ventilatory control, heightened SNS activity, and the presence of comorbid conditions, such as chronic lung disease, obesity, and cachexia (Table 2). Together, these mechanisms give rise to dyspnea. The multifactorial nature of this symptom makes it difficult for clinicians to determine the primary cause that should be targeted for treatment. The more mechanisms contributing to dyspnea, the lower the threshold for its occurrence. However, it is important to remember that dyspnea is a very subjective symptom, further complicating assessment and diagnosis.Table 2Determinants of dyspnea severity in heart failureCategoryPotential Main MechanismClinical Impact Pulmonary congestion/fluid redistributionThe accumulation of fluid in the lungs, insufficient lymphatic drainageImpaired gas exchange, possible sudden onset of dyspnea Hemodynamic disturbancesReduced cardiac output,elevated left ventricular filling pressures, increased pulmonary capillary wedge pressure (PCWP)Dyspnea even without overt pulmonary congestion Neurohormonal activationSympathetic nervous system, renin–angiotensin–aldosterone system (RAAS) activationFluid retention,increased ventilatory drive Pulmonary and chest wall mechanoreceptorsReduced compliance of the lungs and chest wall, stimulation of receptorsDyspnea even with normal blood gas parameters and the absence of pulmonary congestion ChemoreceptorsHypersensitivity to changes in PaO₂, PaCO₂, and pHPathological breathing patterns, including excessive ventilatory drive and periodic breathing ComorbiditiesLung disease,Obesity,Anemia,Renal dysfunction,Cachexia and sarcopenia,Mental disorders (neurosis, depression),Smoldering inflammationLower threshold for dyspnea, worsened symptom perception Table 2Determinants of dyspnea severity in heart failureCategoryPotential Main MechanismClinical Impact Pulmonary congestion/fluid redistributionThe accumulation of fluid in the lungs, insufficient lymphatic drainageImpaired gas exchange, possible sudden onset of dyspnea Hemodynamic disturbancesReduced cardiac output,elevated left ventricular filling pressures, increased pulmonary capillary wedge pressure (PCWP)Dyspnea even without overt pulmonary congestion Neurohormonal activationSympathetic nervous system, renin–angiotensin–aldosterone system (RAAS) activationFluid retention,increased ventilatory drive Pulmonary and chest wall mechanoreceptorsReduced compliance of the lungs and chest wall, stimulation of receptorsDyspnea even with normal blood gas parameters and the absence of pulmonary congestion ChemoreceptorsHypersensitivity to changes in PaO₂, PaCO₂, and pHPathological breathing patterns, including excessive ventilatory drive and periodic breathing ComorbiditiesLung disease,Obesity,Anemia,Renal dysfunction,Cachexia and sarcopenia,Mental disorders (neurosis, depression),Smoldering inflammationLower threshold for dyspnea, worsened symptom perception Determinants of dyspnea severity in heart failure Reduced cardiac output, elevated left ventricular filling pressures, increased pulmonary capillary wedge pressure (PCWP) Fluid retention, increased ventilatory drive Lung disease, Obesity, Anemia, Renal dysfunction, Cachexia and sarcopenia, Mental disorders (neurosis, depression), Smoldering inflammation Dyspnea in HF is not merely a marker of pulmonary congestion but a multidimensional neurocardiorespiratory phenomenon whose trajectory may reflect decongestion, but not necessarily prognosis. The key pathophysiological mechanisms contributing to dyspnea in HF, along with commonly used tools for assessing its severity, are summarized in Fig. 1.Fig. 1Schematic representation of the main pathophysiological mechanisms underlying dyspnea in heart failure and commonly used tools for its assessment Schematic representation of the main pathophysiological mechanisms underlying dyspnea in heart failure and commonly used tools for its assessment This article reviews the principal pathophysiological mechanisms underlying dyspnea in AHF and highlights their clinical relevance. Furthermore, we provide a comprehensive overview of tools for assessing dyspnea severity and discuss their applicability in routine clinical practice. Finally, we summarize evidence from large-scale clinical trials in AHF, with particular emphasis on the effects of therapeutic interventions on the trajectory of dyspnea perception and on prognosis in relation to symptom severity. Pathophysiology of dyspnea in (acute) heart failure Dyspnea is a complex and subjective clinical symptom. Its perception may result from various pathophysiological mechanisms reflecting disorders of the respiratory, cardiovascular, nervous, and muscular systems, underscoring the heterogeneous nature of this condition (Table 2). A comprehensive understanding of the mechanisms underlying dyspnea is important for accurate clinical interpretation, differential diagnosis, and the development of effective diagnostic and therapeutic strategies [3]. Pulmonary congestion One mechanism underlying dyspnea in AHF is pulmonary congestion. The objective assessment of pulmonary congestion is difficult [3–7]. Pulmonary congestion is the accumulation of fluid in the lungs, leading to impaired gas exchange. In general, congestion results from two fundamental mechanisms: volume overload and fluid redistribution [8–11]. The process leading to volume overload is typically slow, developing over days or weeks and often results in chronic congestion. This phenomenon is driven by neurohormonal compensatory responses to reduced cardiac output, including activation of the renin–angiotensin–aldosterone system (RAAS), the sympathetic nervous system (SNS), and the arginine-vasopressin system. These mechanisms promote vasoconstriction and impair renal sodium excretion, ultimately leading to water retention [12–14]. The resulting imbalance between hydrostatic and oncotic pressures in the capillaries and interstitium facilitates fluid leakage across the vascular endothelium, leading to edema. In chronic heart failure (CHF), persistent neurohormonal activation leads to sustained sodium and water retention, reducing the pulmonary interstitium compliance to accommodate additional fluid [13–15]. It is also hypothesized that permanent volume/pressure overload of the pulmonary leads to its remodeling that may facilitate lung fluid accumulation and lower the threshold for dyspnea perception. The pulmonary lymphatic system plays a significant, though often underappreciated, role in the pathogenesis of lung fluid homeostasis and, therefore, congestion development. Under physiological conditions, it ensures efficient drainage of excess fluid from the pulmonary interstitium, thereby maintaining normal lung compliance. In AHF, a rapid increase in hydrostatic pressure within the pulmonary circulation may exceed the compensatory capacity of the lymphatic system. Concurrent venous congestion further impairs lymphatic outflow into the venous system, leading to interstitial fluid accumulation. This process results in reduced lung compliance, increased work of breathing, and consequently contributes to the development of dyspnea [16–18]. The second, and in many cases likely predominant, mechanism of pulmonary congestion in AHF is fluid redistribution. Rapid shifts of fluid between body compartments, rather than an increase in total blood volume, may represent the primary mechanism underlying congestion-driven dyspnea (i.e., pulmonary oedema) [19]. This concept is supported by observations that a substantial proportion of patients with AHF develop signs and symptoms of congestion without significant weight gain [20]. Fluid redistribution may result from vasoconstrictor neurohormonal effects on the arterial compartment and reduced compliance of the venous reservoir, particularly within the splanchnic circulation [11, 21, 22]. Rapid transfer of fluid between these compartments can lead to abrupt increases in afterload, venous return, and central venous pressure, thereby increasing cardiac preload. In HF, this mismatch between the arterial and venous systems promotes fluid accumulation in the pulmonary vasculature, leading to varying degrees of pulmonary congestion and dyspnea. In many patients with CHF, hemodynamic disturbances alone—characterized by elevated left ventricular filling pressures and increased pulmonary capillary wedge pressure—may cause dyspnea that precedes the overt clinical manifestation of pulmonary congestion [23–25]. This may also be the major mechanism of dyspnea perception during exercise in CHF (see below). The central hemodynamics and contribution of mechanoreceptors to the perception of dyspnea Another factor underlying the development of dyspnea is hemodynamic disturbances in both the systemic and pulmonary circulations. Dynamic hemodynamic changes, often initially occurring without overt signs of fluid overload, may ultimately lead to pulmonary congestion. Depending on the predominant HF phenotype, either an increase in left ventricular filling pressure, with a secondary rise in left atrial pressure and pulmonary capillary wedge pressure, or a reduction in cardiac output may be observed. In the first scenario, rising pressures within the pulmonary vasculature lead not only to activation of mechanoreceptors located in the airway walls and chest but also to the initiation of a neurohormonal response. This response triggers a cascade of multiorgan reactions, with predominant activation of the SNS and the kidneys. In current clinical practice, assessment of natriuretic peptide levels (NT-proBNP, BNP, MR-proANP) is widely used [26]. However, these biomarkers do not show a direct relationship with the degree of alveolar–capillary membrane injury. Increasing attention has therefore been directed toward biochemical markers associated with hemodynamic changes and increased tension within the alveolar capillaries, referred to as alveolo-capillary membrane stress failure (ACMSF). The release of specific markers of barrier injury, such as surfactant proteins—particularly the immature form of surfactant protein B (SP-B)—provides a mechanistic link between elevated pulmonary capillary pressures and the perception of dyspnea, extending beyond pathways limited to mechanoreceptor activation alone. This distinction is particularly important because natriuretic peptides are released in response to ventricular wall stress and primarily reflect cardiac pressure and volume overload. In contrast, immature SP-B is secreted by type II pneumocytes, and its release results from mechanical stretching and structural injury of the alveolar–capillary membrane. Consequently, it reflects a distinct pathophysiological pathway—direct injury to the alveolar–capillary membrane—rather than the preceding hemodynamic overload. Available data also suggest that the release kinetics of immature SP-B differ from those of natriuretic peptides, more closely reflecting persistent alveolar injury and, consequently, sustained pulmonary congestion and dyspnea. From a prognostic perspective, circulating immature SP-B may correlate with the severity of pulmonary congestion assessed by chest imaging and lung ultrasound and may provide incremental prognostic information beyond that offered by natriuretic peptides alone in patients with AHF. However, the available evidence remains limited and is derived primarily from small, single-center clinical studies. Therefore, large multicenter studies are needed before immature SP-B can be incorporated into clinical practice as a biomarker reflecting alveolar–capillary membrane injury and pulmonary congestion [6, 7]. Another important hemodynamic mechanism contributing to dyspnea is reduced cardiac output, which may indirectly increase filling pressures and further exacerbate pulmonary congestion. Impaired cardiac output also results in inadequate oxygen delivery to peripheral tissues. Peripheral hypoperfusion, including hypoxia of the respiratory muscles, lowers the fatigue threshold and thereby intensifies dyspnea. In addition, the stress response to hypoperfusion of vital organs activates the SNS and the RAAS, leading to sodium and water retention and vasoconstriction, which further increase volume overload and secondarily contribute to dyspnea. Hemodynamic disturbances associated with elevated pulmonary vascular pressures lead to decreased lung compliance and increased inspiratory airflow resistance. Increased stiffness of the chest wall, airways, alveoli, and respiratory muscles modulates afferent signals generated by mechanoreceptors. These receptors, located in the airways, lungs, chest wall, and respiratory muscles, transmit information to the respiratory centers in the medulla oblongata and pons [27, 28]. Three main types of mechanoreceptors can be distinguished by location and response characteristics: slowly adapting stretch receptors (SARs), rapidly adapting receptors (RARs), and juxtacapillary (J) receptors. Depending on the site of the stimulus induced by excessive stretch, impulses are conducted via the vagus nerve to the brainstem, modulating the rate and depth of breathing and the cough reflex [29]. Increased airway and parenchymal stiffness result in excessive afferent signaling, which directly alters the basic respiratory rhythm, producing rapid, shallow breathing and the subjective sensation of lack of air. This discrepancy between respiratory drive and mechanoreceptor feedback is one of the many mechanisms of dyspnea [30, 31]. Mechanoreceptor activation explains why dyspnea can occur even with normal blood gas parameters and the absence of pulmonary congestion. This may have significant clinical implications, as treatment aimed solely at improving blood gas parameters will not bring the expected improvement. In many cases, effective alleviation of perceived dyspnea requires interventions specifically aimed at reducing mechanoreceptor stimulation, such as non-invasive ventilation or bronchodilator therapy [31]. A schematic summary of the relationship between the severity of dyspnea and hemodynamic changes is presented in Fig. 2.Fig. 2Schematic representation of the relationships between hemodynamic changes, mechanoreceptors, and the severity of dyspnea Schematic representation of the relationships between hemodynamic changes, mechanoreceptors, and the severity of dyspnea Chemoreceptors in the regulation and perception of dyspnea Dyspnea in AHF cannot be explained solely by mechanical respiratory disturbances, such as airflow limitation, impaired chest wall mechanics, or alterations in pulmonary vascular hemodynamics. Neurohormonal mechanisms—particularly excessive activation of peripheral and central chemoreceptors—play a central role in generating dyspnea. Dysregulation of gas exchange, resulting in changes in arterial oxygen (PaO₂) and carbon dioxide (PaCO₂) tensions as well as acid–base imbalance, represents an important but often underappreciated contributor to the pathogenesis of dyspnea in AHF. Chemoreceptors are specialized sensory structures that play a central role in ventilatory control and significantly contribute to dyspnea perception. They respond to changes in the chemical composition of blood and cerebrospinal fluid, primarily fluctuations in PaO₂, PaCO₂, and pH. Based on their anatomical location, chemoreceptors are classified as peripheral and central, each contributing differently to the regulation of ventilation and the subjective experience of dyspnea [32]. In patients with CHF, tonic overactivation of the SNS accompanied by heightened chemoreceptor sensitivity, is frequently observed. Importantly, this maladaptive neurohumoral activation contributes to disease progression and is strongly associated with adverse prognosis and increased mortality risk [33–35]. The principal peripheral chemoreceptors in humans are the carotid bodies, located at the bifurcation of the common carotid arteries, and the aortic bodies, situated along the aortic arch. These structures consist of clusters of chemosensitive cells that respond predominantly to hypoxemia and, to a lesser extent, to hypercapnia and acidosis [36, 37]. Proper carotid body function is essential for maintaining blood gas homeostasis and adequate tissue perfusion. Upon physiological activation, afferent signals are transmitted to the central nervous system, stimulating respiratory rhythm–generating centers in the brainstem and increasing ventilation. Concurrently, activation of the SNS supports arterial blood pressure and preserves perfusion of vital organs. In CHF, sustained tonic activation of chemoreceptors and exaggerated chemoreflex responses to hypoxia become maladaptive. This hyperresponsiveness promotes pathological breathing patterns, including excessive ventilatory drive and periodic breathing, and contributes to the sensation of “air hunger”, even in the absence of overt pulmonary congestion [38, 39]. As a result, relatively small changes in blood gas parameters may evoke a disproportionate ventilatory response, leading to severe dyspnea despite normal arterial oxygen saturation and carbon dioxide levels. Furthermore, the interaction between chemoreceptor-driven ventilatory stimulation and altered afferent input from pulmonary and chest wall mechanoreceptors amplifies dyspnea perception. This mismatch between respiratory demand and the mechanical capacity of the respiratory system represents a determinant of symptom severity in patients with HF [40]. A schematic summary of the combined effects of chemo- and mechanoreceptors on the generation of dyspnea is presented in Fig. 3.Fig. 3Schematic illustration of how chemoreceptor and mechanoreceptor interactions contribute to the perception of dyspnea Schematic illustration of how chemoreceptor and mechanoreceptor interactions contribute to the perception of dyspnea Comorbidities influencing dyspnea The perception and severity of dyspnea in HF are frequently influenced by the presence of comorbidities, which may independently exacerbate respiratory symptoms, interact with heart failure–related pathophysiology, and complicate both diagnostic evaluation and therapeutic decision-making [6, 41–43]. In many patients, dyspnea reflects not only acute cardiac decompensation but also the cumulative burden of chronic systemic and pulmonary disorders that alter ventilatory mechanics, gas exchange, and respiratory control. Chronic pulmonary diseases represent some of the most important modifiers of dyspnea in AHF. Chronic obstructive pulmonary disease (COPD), asthma, and interstitial lung diseases reduce lung compliance, increase airway resistance, and impair ventilation–perfusion matching [43]. These abnormalities increase the work of breathing and lower the threshold at which pulmonary congestion translates into severe subjective dyspnea [44–46]. Importantly, optimized guideline-directed therapy and close follow-up improve respiratory recovery after the AHF episode [46]. Obesity is another prevalent comorbidity and target for therapy (in heart failure with preserved ejection fraction (HFpEF) patients) that significantly affects dyspnea perception in AHF [42, 47–52]. Excess adipose tissue reduces chest wall compliance, increases intra-abdominal pressure, and limits diaphragmatic excursion, thereby increasing the mechanical load imposed on the respiratory system [53, 54]. In contrast, cachexia and sarcopenia—common in advanced HF—impair respiratory and peripheral muscle strength, reduce ventilatory reserve, and contribute to the early onset of dyspnea during minimal exertion or even at rest [55–59]. Renal dysfunction and anemia, both frequently observed in patients with AHF, further exacerbate dyspnea through complementary mechanisms [60–67]. The kidneys are one of the most important organs for compensating for acid-base imbalances. When they fail, and metabolic acidosis develops, the respiratory system initiates a compensatory response through hyperventilation to restore acid-base homeostasis. Importantly, this hyperventilation (Kussmaul’s acidotic breathing) does not necessarily cause dyspnea. Impaired renal function promotes sodium and water retention, leading to increased intravascular volume and pulmonary congestion, while simultaneously limiting the effectiveness of diuretic therapy [14, 15]. Anemia reduces the oxygen-carrying capacity and tissue oxygen delivery, leading to compensatory increases in cardiac output and ventilatory drive. This heightened respiratory demand may intensify the sensation of air hunger, even in the absence of significant hypoxemia or hypercapnia [68, 69]. Smoldering, low-grade inflammation plays a significant role in the pathophysiology of HF. Chronic inflammation, characterized by elevated levels of proinflammatory cytokines such as IL-1, TNF-α, IL-6, and hs-CRP, contributes to endothelial dysfunction, increased vascular permeability, and myocardial remodeling. These processes impair left ventricular compliance, leading to elevated filling pressures, reduced cardiac output, and increased pulmonary vascular pressures, contributing to dyspnea even in the absence of pulmonary congestion. Furthermore, systemic inflammation can directly impact skeletal and respiratory muscles, promoting fatigue and consequently lowering the threshold for dyspnea [70, 71]. Quality of life is a critical consideration for patients with HF. It is largely influenced by troublesome symptoms, including peripheral edema and dyspnea, particularly at rest and in the form of orthopnea. These symptoms significantly reduce quality of life, resulting in more frequent hospitalizations [72]. Reduced quality of life is also associated with a higher incidence of neurosis and depressive disorders in this patient population. These often underdiagnosed disorders may exacerbate dyspnea by altering central processing of respiratory sensations and heightening attention to respiratory discomfort [73–75]. A comprehensive understanding of the interactions between AHF and comorbid conditions is essential for the individualized management of dyspnea. Therapeutic strategies should not be limited to correcting hemodynamic and neurohormonal disturbances inherent to HF but should also address modifiable comorbidities that contribute to ventilatory impairment and symptom burden. Such an integrated approach is crucial for optimizing dyspnea relief, improving functional status, and ultimately enhancing clinical outcomes in patients with AHF. Integration of mechanisms across the HF phenotype and activity spectrum The mechanisms described above do not act in isolation, and their relative contribution varies with both HF phenotype and whether dyspnea is assessed at rest or during exertion. Table 3 summarizes these phenotype- and state-dependent differences, integrating hemodynamic, neural, and comorbidity-related pathways discussed throughout this review.Table 3Phenotype and timing dependent mechanisms of dyspnea in heart failureDomainHFrEFHFpEFPredominant timing Filling pressures/congestionFrequently elevated at rest owing to persistent congestionOften normal at rest but increase during exertionHFrEF: Exertional → Resting (progression), HFpEF: Exertional Cardiac outputReduced because of impaired systolic functionPreserved at rest, but impaired reserve during exertionHFrEF: Exertional → Resting (progression), HFpEF: Exertional Neurohormonal activationSympathetic activation and increased chemoreflex sensitivity are well documentedSympathetic activation is present, whereas chemoreflex sensitivity remains less well characterizedBoth Key comorbiditiesAnemia, cachexia, sarcopenia (particularly in advanced HF)Obesity, chronic lung diseaseMainly exertion Low-grade systemic inflammationNot a primary emphasis in current evidenceProposed as a contributing mechanism; precise role remains incompletely understoodBoth Table 3Phenotype and timing dependent mechanisms of dyspnea in heart failureDomainHFrEFHFpEFPredominant timing Filling pressures/congestionFrequently elevated at rest owing to persistent congestionOften normal at rest but increase during exertionHFrEF: Exertional → Resting (progression), HFpEF: Exertional Cardiac outputReduced because of impaired systolic functionPreserved at rest, but impaired reserve during exertionHFrEF: Exertional → Resting (progression), HFpEF: Exertional Neurohormonal activationSympathetic activation and increased chemoreflex sensitivity are well documentedSympathetic activation is present, whereas chemoreflex sensitivity remains less well characterizedBoth Key comorbiditiesAnemia, cachexia, sarcopenia (particularly in advanced HF)Obesity, chronic lung diseaseMainly exertion Low-grade systemic inflammationNot a primary emphasis in current evidenceProposed as a contributing mechanism; precise role remains incompletely understoodBoth Phenotype and timing dependent mechanisms of dyspnea in heart failure Overall, two broad phenotypic patterns of dyspnea can be distinguished in HF. In heart failure with reduced ejection fraction (HFrEF), dyspnea is often initially exertional, but with disease progression it may also occur at rest. It results from persistent volume overload and reduced cardiac output, together with SNS activation, increased chemoreflex sensitivity, and common comorbidities such as anemia, cachexia, and sarcopenia, which are well described in advanced HF. In HFpEF, dyspnea is more typically exertional and emerges under conditions of hemodynamic stress when filling pressures rise. With progression, resting dyspnea may also develop as filling pressures become chronically elevated. In this phenotype, symptoms appear to be more closely linked to comorbid conditions, particularly obesity and chronic lung disease. Low-grade systemic inflammation has also been proposed as a contributing mechanism, although its role in symptom generation remains incompletely understood [70, 71]. This distinction mainly reflects the underlying chronic hemodynamic trajectory rather than the clinical picture during acute decompensation, when resting dyspnea predominates regardless of ejection fraction phenotype. From a clinical standpoint, this remains important, as assessment of dyspnea at rest alone may underestimate both symptom burden and underlying pathophysiological complexity in a substantial proportion of patients with HFpEF outside acute decompensation. This supports the use of exercise-based or provocative hemodynamic assessment in selected patients, in line with current diagnostic algorithms. Assessment scales for dyspnea and their clinical relevance Due to the subjective nature of dyspnea and the absence of a single objective parameter that accurately reflects its severity, standardized rating scales play a crucial role in both routine clinical practice and the design and interpretation of clinical trials in AHF. A variety of instruments with differing levels of sensitivity have been developed to enable structured assessment of dyspnea severity, its temporal changes, and response to treatment. These tools allow for quantitative evaluation of dyspnea perception by integrating complex pathophysiological mechanisms into a single, measurable patient-reported outcome [31, 76]. Table 4 summarizes the most commonly used scales, highlighting their principal advantages and limitations, which directly influence their applicability in everyday clinical practice. In routine care, simple, easy-to-use instruments that impose minimal cognitive burden on the patient are most commonly used. One such tool is the Numerical Rating Scale (NRS), typically ranging from 0 to 10, where 0 denotes no dyspnea and 10 represents the worst imaginable dyspnea. Its intuitive structure enables repeated assessments during hospitalization, facilitating monitoring of symptom dynamics. The NRS demonstrates good reproducibility and correlates well with other dyspnea assessment tools. Owing to its simplicity, it can be applied across a broad spectrum of patients, including those with mild cognitive impairment, which explains its widespread use in clinical practice [77–79]. Another commonly used tool is the modified Borg scale, particularly in settings involving physical exertion, rehabilitation, and the assessment of functional capacity. This ordinal scale ranges from 0 to 10 and is anchored by descriptive terms reflecting increasing dyspnea intensity, from “no dyspnea” to “maximal dyspnea.” It is widely employed during exercise testing, providing insight into the subjective ventilatory response to objective workload. Although less sensitive to subtle changes in dyspnea at rest, the Borg scale is valuable for evaluating exercise tolerance during later stages of hospitalization, mobilization, and rehabilitation. The combined use of these scales enables a more comprehensive assessment of dyspnea and supports individualized therapeutic decision-making in patients with AHF [80, 81].Table 4Comparison of dyspnea assessment scales used in heart failureScaleTypeTypical UseUtility in Daily PracticeUtility in Clinical TrialsStrengthsLimitations Numerical Rating Scale (NRS)Numeric, ordinal (0–10)Bedside assessment, hospitalizationHighModerateSimple and intuitive;minimal cognitive burden;repeatable measurements;good reproducibilityLimited sensitivity to small changes;subjective interpretation of numeric values Modified Borg ScaleOrdinal, descriptive (0–10)Exercise testing, rehabilitation, mobilizationModerateLow–moderateUseful during exertion; reflects ventilatory effort; well-validated in exercise physiologyLess sensitive at rest; limited applicability in bedridden patients; ordinal nature limits statistical analysis Likert Scale (5-point)Ordinal, categoricalClinical trials (short-term outcomes)LowHighEasy to use;captures subjective change over time; suitable for treatment response evaluationLimited granularity; ordinal scale may miss subtle changes;recall bias Likert Scale (7-point)Ordinal, categoricalClinical trialsLowHighGreater sensitivity than 5-point scale;widely accepted; intuitive relative to baselineStill ordinal;intermediate categories may be interpreted subjectively;not ideal for frequent bedside assessments Visual Analog Scale (VAS)Continuous (0–100 mm)Clinical trials, early treatment responseLowVery highHighly sensitive to small changes; continuous measurement;strong statistical powerHigher cognitive demands;less practical for unstable patients;limited bedside repeatability Multidimensional Dyspnea Profile (MDP)Multidimensional (numeric, descriptor-based)Cardiopulmonary and respiratory research, emerging use in HFLowHighSeparates sensory and affective components; flexible temporal framingHigher complexity; limited routine use Dyspnea-12 (D-12)Multidimensional (summed ordinal score)Chronic respiratory disease (COPD, interstitial lung disease (ILD), asthma); limited use in HFLowHighBrief multidimensional assessment; physical and affective domainsLess detailed than MDP;more limited granularity Table 4Comparison of dyspnea assessment scales used in heart failureScaleTypeTypical UseUtility in Daily PracticeUtility in Clinical TrialsStrengthsLimitations Numerical Rating Scale (NRS)Numeric, ordinal (0–10)Bedside assessment, hospitalizationHighModerateSimple and intuitive;minimal cognitive burden;repeatable measurements;good reproducibilityLimited sensitivity to small changes;subjective interpretation of numeric values Modified Borg ScaleOrdinal, descriptive (0–10)Exercise testing, rehabilitation, mobilizationModerateLow–moderateUseful during exertion; reflects ventilatory effort; well-validated in exercise physiologyLess sensitive at rest; limited applicability in bedridden patients; ordinal nature limits statistical analysis Likert Scale (5-point)Ordinal, categoricalClinical trials (short-term outcomes)LowHighEasy to use;captures subjective change over time; suitable for treatment response evaluationLimited granularity; ordinal scale may miss subtle changes;recall bias Likert Scale (7-point)Ordinal, categoricalClinical trialsLowHighGreater sensitivity than 5-point scale;widely accepted; intuitive relative to baselineStill ordinal;intermediate categories may be interpreted subjectively;not ideal for frequent bedside assessments Visual Analog Scale (VAS)Continuous (0–100 mm)Clinical trials, early treatment responseLowVery highHighly sensitive to small changes; continuous measurement;strong statistical powerHigher cognitive demands;less practical for unstable patients;limited bedside repeatability Multidimensional Dyspnea Profile (MDP)Multidimensional (numeric, descriptor-based)Cardiopulmonary and respiratory research, emerging use in HFLowHighSeparates sensory and affective components; flexible temporal framingHigher complexity; limited routine use Dyspnea-12 (D-12)Multidimensional (summed ordinal score)Chronic respiratory disease (COPD, interstitial lung disease (ILD), asthma); limited use in HFLowHighBrief multidimensional assessment; physical and affective domainsLess detailed than MDP;more limited granularity Comparison of dyspnea assessment scales used in heart failure Simple and intuitive; minimal cognitive burden; repeatable measurements; good reproducibility Limited sensitivity to small changes; subjective interpretation of numeric values Useful during exertion; reflects ventilatory effort; well-validated in exercise physiology Easy to use; captures subjective change over time; suitable for treatment response evaluation Limited granularity; ordinal scale may miss subtle changes; recall bias Greater sensitivity than 5-point scale; widely accepted; intuitive relative to baseline Still ordinal; intermediate categories may be interpreted subjectively; not ideal for frequent bedside assessments Highly sensitive to small changes; continuous measurement; strong statistical power Higher cognitive demands; less practical for unstable patients; limited bedside repeatability Less detailed than MDP; more limited granularity In large randomized clinical trials involving patients with AHF, more sensitive instruments are preferred to detect even subtle changes in symptom severity. The Likert scale is among the most commonly used tools for this purpose, allowing qualitative and quantitative assessment of changes in dyspnea severity. It is typically presented as a five- or seven-point scale, enabling patients to rate their dyspnea as “markedly worsened,” “moderately worsened,” “unchanged,” “moderately improved,” or “markedly improved” compared with baseline. The Likert scale is particularly useful for evaluating short-term treatment response, as it captures patient-perceived symptom dynamics [82, 83]. Another widely used instrument in clinical trials is the Visual Analog Scale (VAS). This continuous scale is usually represented by a 100-mm line, anchored at one end by “no dyspnea” and at the other by “the worst imaginable dyspnea.” Patients indicate symptom severity by marking a point on the line, which is then quantified in millimeters. Owing to its continuous nature, the VAS is highly sensitive to small changes in dyspnea intensity and is especially useful for assessing early therapeutic effects, such as during the first hours of treatment for AHF [76, 84]. These methods are often complemented by more specific tools, including the Position-Based Rating Scale (PRS) and the Provoked Dyspnea Assessment (PDA), which account for dyspnea severity in different body positions or under defined clinical conditions [76]. In addition, the New York Heart Association (NYHA) functional classification, although not a direct measure of dyspnea intensity, remains widely used to assess the functional impact of dyspnea on exercise tolerance and daily activities in both observational and interventional studies [85, 86]. Although NRS, the Borg scale, Likert scale, and VAS allow quantitative assessment of dyspnea severity as a unidimensional measure, its multidimensional nature—encompassing sensory, affective, and emotional components, similar to pain—is increasingly recognized [31, 87]. In response, more complex assessment tools have been developed, such as the Multidimensional Dyspnea Profile (MDP) and the Dyspnea-12 (D-12), which allow separation of dyspnea intensity, sensory quality, and associated emotional response [88, 89]. A key limitation of most scales is the lack of a clearly defined temporal context for assessment. While the Borg scale was developed and validated specifically for exertional settings, NRS, Likert, and VAS are most commonly used to capture the patient’s current, momentary perception, often in the resting, bedside setting [77]. In contrast, MDP and D-12 allow an explicit definition of the recall period (e.g., “right now”, “during usual daily activities”, or “over the past week”), which increases their flexibility in distinguishing resting from exertional dyspnea within the same instrument [88, 89]. It should be noted that MDP has been validated across different clinical populations, including respiratory diseases and other conditions associated with dyspnea, whereas D-12 has been used primarily in chronic respiratory diseases (particularly COPD, interstitial lung disease, and asthma) [90]. Accordingly, available evidence for the use of both instruments in acute heart failure remains limited and less well established. Overall, the use of standardized dyspnea assessment scales enables partial objectification of this inherently subjective symptom, improves comparability across clinical studies, and facilitates interpretation of therapeutic efficacy. These scales do not distinguish dyspnea according to its etiology but rather assess symptom severity and its changes over time. Careful selection of the assessment tool is essential, as individual scales differ in sensitivity, reproducibility, and clinical relevance, depending on the study objective and clinical context. The initial assessment of severe dyspnea at hospital admission, regardless of the scale used, helps identify patients requiring prompt medical intervention. Available clinical trial data indicate that improvement in dyspnea is frequently observed irrespective of the therapeutic strategy applied, but this improvement does not translate into better long-term prognosis. Importantly, the absence of dyspnea relief identifies patients at increased risk of adverse clinical outcomes [91, 92]. Dyspnea and its significance in large clinical trials of acute heart failure Despite the central role of dyspnea in AHF, relatively few large randomized trials have used it as a formal endpoint, including VMAC, PROTECT, ASCEND-HF, RELAX-AHF, EMPA-RESPONSE-AHF, and DAPA-RESPONSE-AHF. Other major trials in AHF, such as TRUE-AHF, RELAX-AHF-2, and EMPULSE, did not assess dyspnea directly but still help to place early symptomatic or clinical changes in the context of longer-term outcomes. These studies differed with respect to the methodology used to assess dyspnea, the timing of symptom evaluation, and the prognostic relevance of the reported outcomes. From a methodological perspective, using dyspnea as an endpoint is associated with significant limitations. Dyspnea is a highly subjective symptom with complex pathophysiology, and the lack of standardized assessment tools, heterogeneity of applied scales, and variability in assessment time points further hinder cross-trial comparisons and limit clinical interpretability. Table 5 summarizes major clinical trials in AHF that assessed dyspnea, related patient-centered outcomes, or their relationship to hard clinical endpoints.Table 5Summary of major clinical trials in acute heart failure evaluating dyspnea and clinical outcomesStudyDesignPatients (n)InterventionDyspnea EndpointEffect on DyspneaClinical Outcomes VMACDouble-blind RCT489nesiritide vs. nitroglycerin vs. placeboLikert scale (subjective) at predefined time pointsRapid subjective improvement in dyspneaSymptomatic relief did not translate into improved prognosis PROTECT (Pilot)Pilot RCT~ 200rolofylline7-point Likert scale (24–48 h)Early reduction in dyspneaN/A PROTECT (Main)Double-blind RCT2033rolofylline vs. placebo7-point Likert scaleNo superiority over placeboNo improvement in prognosis or composite endpoints ASCEND-HFDouble-blind RCT7141nesiritide vs. placebo7-point Likert scale at 6 and 24 hModest improvement in dyspneaNo effect on 30-day mortality or rehospitalization; higher hypotension risk TRUE-AHFDouble-blind RCT2157ularitide vs. placeboNot assessed directlyN/A (dyspnea not a formal endpoint)No reduction in cardiovascular mortality RELAX-AHFDouble-blind RCT1161serelaxin vs. placeboVAS + AUC (first 5 days)Significant improvement in dyspneaNo impact on short-term mortality or rehospitalization RELAX-AHF-2Double-blind RCT6545serelaxin vs. placeboNot assessedN/A (dyspnea not assessed)No difference in cardiovascular mortality or worsening HF EMPA-RESPONSE-AHFPilot RCT80empagliflozin vs. placeboVAS (co-primary)No significant differenceNo difference in diuretic response, NT-proBNP, or length of stay EMPULSEDouble-blind RCT530empagliflozin vs. placeboIndirect assessment via KCCQ-TSSSymptom burden improved within compositeHigher probability of clinical benefit (hierarchical endpoint: death, HF events, health status) DAPA-RESPONSE-AHFDouble-blind RCT87dapagliflozin vs. placeboVAS + AUC (first 4 days)Faster, greater reduction in dyspneaImproved decongestion markers (diuresis, NT-proBNP); no effect on mortality or serious adverse eventsRCT randomized controlled trial, VAS Visual Analog Scale, AUC area under the curve NT-proBNP N-terminal pro–B-type natriuretic peptide, KCCQ-TSS Kansas City Cardiomyopathy Questionnaire Total Symptom Score, HF heart failure Table 5Summary of major clinical trials in acute heart failure evaluating dyspnea and clinical outcomesStudyDesignPatients (n)InterventionDyspnea EndpointEffect on DyspneaClinical Outcomes VMACDouble-blind RCT489nesiritide vs. nitroglycerin vs. placeboLikert scale (subjective) at predefined time pointsRapid subjective improvement in dyspneaSymptomatic relief did not translate into improved prognosis PROTECT (Pilot)Pilot RCT~ 200rolofylline7-point Likert scale (24–48 h)Early reduction in dyspneaN/A PROTECT (Main)Double-blind RCT2033rolofylline vs. placebo7-point Likert scaleNo superiority over placeboNo improvement in prognosis or composite endpoints ASCEND-HFDouble-blind RCT7141nesiritide vs. placebo7-point Likert scale at 6 and 24 hModest improvement in dyspneaNo effect on 30-day mortality or rehospitalization; higher hypotension risk TRUE-AHFDouble-blind RCT2157ularitide vs. placeboNot assessed directlyN/A (dyspnea not a formal endpoint)No reduction in cardiovascular mortality RELAX-AHFDouble-blind RCT1161serelaxin vs. placeboVAS + AUC (first 5 days)Significant improvement in dyspneaNo impact on short-term mortality or rehospitalization RELAX-AHF-2Double-blind RCT6545serelaxin vs. placeboNot assessedN/A (dyspnea not assessed)No difference in cardiovascular mortality or worsening HF EMPA-RESPONSE-AHFPilot RCT80empagliflozin vs. placeboVAS (co-primary)No significant differenceNo difference in diuretic response, NT-proBNP, or length of stay EMPULSEDouble-blind RCT530empagliflozin vs. placeboIndirect assessment via KCCQ-TSSSymptom burden improved within compositeHigher probability of clinical benefit (hierarchical endpoint: death, HF events, health status) DAPA-RESPONSE-AHFDouble-blind RCT87dapagliflozin vs. placeboVAS + AUC (first 4 days)Faster, greater reduction in dyspneaImproved decongestion markers (diuresis, NT-proBNP); no effect on mortality or serious adverse eventsRCT randomized controlled trial, VAS Visual Analog Scale, AUC area under the curve NT-proBNP N-terminal pro–B-type natriuretic peptide, KCCQ-TSS Kansas City Cardiomyopathy Questionnaire Total Symptom Score, HF heart failure Summary of major clinical trials in acute heart failure evaluating dyspnea and clinical outcomes RCT randomized controlled trial, VAS Visual Analog Scale, AUC area under the curve NT-proBNP N-terminal pro–B-type natriuretic peptide, KCCQ-TSS Kansas City Cardiomyopathy Questionnaire Total Symptom Score, HF heart failure One of the first clinical trials to assess dyspnea as an endpoint was the VMAC (Vasodilation in the Management of Acute Congestive Heart Failure) trial. This double-blind study enrolled 489 patients hospitalized at 55 centers in the United States with significant dyspnea at rest due to AHF. The study compared the effects of intravenous nesiritide with placebo and nitroglycerin infusion at predefined time points. A subset of patients underwent invasive hemodynamic assessment using a Swan–Ganz catheter, including measurement of pulmonary capillary wedge pressure (PCWP), while dyspnea severity was assessed using a Likert scale. The trial demonstrated that subjective improvement in dyspnea could be achieved rapidly after initiation of vasodilator therapy; however, differences between treatment groups were small and clinically insignificant. Importantly, the study also showed that symptomatic improvement does not necessarily translate into improved prognosis [93]. These findings led to the continued use of dyspnea as an endpoint in subsequent AHF trials, while simultaneously highlighting its limitations as a stand-alone measure of treatment efficacy. In the following years, the results of the PROTECT pilot study were published, followed by the main PROTECT trial (Placebo-Controlled Randomized Study of the Selective A1 Adenosine Receptor Antagonist Rolofylline in Patients With Acute Heart Failure and Renal Dysfunction), which enrolled more than 2,000 patients hospitalized with AHF and significant dyspnea. In the pilot phase, a reduction in dyspnea was observed during the first 24–48 h of hospitalization, assessed using a 7-point Likert scale. However, in the main PROTECT trial, rolofylline did not demonstrate superiority over placebo with respect to symptomatic improvement or the composite endpoint, including early dyspnea relief. Despite systematic assessment of dyspnea severity, the rate of improvement was comparable between groups, and treatment did not translate into improved clinical outcomes or prognosis [94, 95]. Overall, the PROTECT findings demonstrated that even carefully defined and systematically assessed improvement in dyspnea is not a sufficiently sensitive or specific endpoint to establish the efficacy of a novel therapy in AHF. Another study evaluating the effects of nesiritide in patients with AHF was ASCEND-HF (Acute Study of Clinical Effectiveness of Nesiritide in Decompensated Heart Failure). This large, randomized trial enrolled 7141 hospitalized patients with AHF who were randomly assigned to receive an intravenous infusion of nesiritide or placebo in addition to standard care. The primary endpoints included changes in dyspnea severity at 6 and 24 h, assessed using a 7-point Likert scale, and a composite endpoint of death or rehospitalization for HF within 30 days. Moderate or marked improvement in dyspnea was observed at both time points in the nesiritide group compared with placebo; however, these differences were marginal and not statistically significant. Moreover, nesiritide did not demonstrate superiority over placebo with respect to 30-day mortality or rehospitalization and was associated with a higher incidence of hypotension. Overall, the ASCEND-HF results indicate that although improvement in dyspnea may be subjectively more frequent with active therapy, its clinical relevance is limited and does not translate into benefits in major clinical endpoints [96]. TRUE-AHF was another randomized, placebo-controlled trial evaluating a vasodilator in AHF, in this case ularitide, a synthetic analogue of the natriuretic peptide urodilatin, in patients hospitalized with AHF. The trial had two coprimary endpoints: death from cardiovascular causes and a hierarchical clinical composite assessing the early clinical course. Ularitide produced significant early hemodynamic and biomarker changes consistent with decongestion, including reductions in NT-proBNP levels, though it was associated with a higher incidence of hypotension. Despite these early effects, ularitide did not reduce cardiovascular mortality. These findings support the observation that favorable short-term physiological effects do not necessarily translate into improved long-term clinical outcomes in AHF [97]. In the following years, the results of another large, randomized trial, RELAX-AHF (RELAXin in Acute Heart Failure), were published, including 1,161 patients hospitalized with AHF, symptomatic dyspnea, and congestion. Patients were randomly assigned to receive an intravenous infusion of serelaxin (recombinant human relaxin-2) or placebo within the first 48 h of hospitalization, in addition to standard therapy. The primary endpoint was subjective improvement in dyspnea, assessed using a VAS scale and area under the curve (AUC) analysis over the first five days of hospitalization. The study demonstrated that serelaxin significantly improved dyspnea compared with placebo. However, as in other AHF trials, treatment did not significantly affect short-term mortality or rehospitalization rates. An exploratory analysis had suggested a potential reduction in cardiovascular mortality at 180 days, although this was not a prespecified endpoint [98]. RELAX-AHF showed that appropriately tailored therapy and systematic, quantitative assessment of dyspnea can provide meaningful symptomatic relief. To further investigate this signal, the larger RELAX-AHF-2 trial enrolled 6,545 patients hospitalized for AHF with dyspnea and congestion, who were randomly assigned to serelaxin or placebo in addition to standard therapy. Cardiovascular mortality at 180 days was a co-primary endpoint, alongside worsening HF at day 5. The trial did not show a difference in mortality between the groups, and dyspnea was not included as a formal endpoint [99]. These findings did not confirm the mortality difference observed in the original RELAX-AHF trial. Beyond vasoactive agents such as serelaxin, dyspnea and related outcomes have also been investigated in trials of SGLT2 inhibitors. Earlier evidence for SGLT2 inhibitors in AHF came from the EMPA-RESPONSE-AHF pilot trial, which randomized 80 hospitalized patients to empagliflozin or placebo. The co-primary endpoints included changes in dyspnea assessed by the VAS, diuretic response, and NT-proBNP levels. No significant differences were observed between the treatment groups in any of these outcomes, nor in length of hospital stay. Although no symptomatic or decongestive benefit was demonstrated, the study supported the feasibility and safety of initiating SGLT2 inhibitor therapy during hospitalization for AHF [100]. In contrast, the larger EMPULSE trial randomized 530 patients with AHF to empagliflozin or placebo after clinical stabilization during hospitalization. The primary endpoint was a hierarchical composite of clinical benefit, incorporating death, HF events, and health status assessed by the Kansas City Cardiomyopathy Questionnaire Total Symptom Score (KCCQ-TSS) at 90 days. Empagliflozin was associated with a significantly greater probability of benefit than placebo [101]. Returning to dyspnea-specific assessment, the most recent study to include this symptom as a formal endpoint is DAPA-RESPONSE-AHF. This randomized, double-blind trial enrolled 87 hospitalized patients with AHF and significant dyspnea. Patients were randomly assigned within 24 h of admission to receive dapagliflozin or placebo for 30 days. The primary endpoint was the difference in the AUC of the VAS dyspnea scale during the first four days of hospitalization. The study demonstrated a faster and significantly greater reduction in dyspnea severity in the dapagliflozin group compared with placebo. Additionally, beneficial effects on congestion parameters were observed, including greater diuresis and larger reductions in NT-proBNP levels. However, no significant differences were seen in short-term mortality or the incidence of serious adverse events [102]. DAPA-RESPONSE-AHF provides evidence that dapagliflozin can induce early relief of dyspnea in patients with AHF, confirming the value of this symptom as a patient-centered endpoint, although without clear impact on hard clinical outcomes in the short term. Conclusions Dyspnea is a key and highly burdensome clinical symptom in patients with HF and often represents the primary target of initial therapy, aimed at improving patient comfort and quality of life. Its severity reflects a complex interaction between pulmonary congestion, hemodynamic and neurohormonal disturbances, and the presence of comorbidities, making the perception of dyspnea highly subjective and difficult to correlate with objective clinical parameters. Assessment scales allow partial objectification of this symptom and enable monitoring of treatment response trajectories in both clinical practice and research settings. Evidence indicates that in AHF, early improvement in dyspnea is common after hospital admission, but does not always translate into better long-term outcomes, whereas refractory or residual dyspnea is associated with poorer prognosis [91, 92]. This dissociation, as illustrated by trials such as ASCEND-HF, PROTECT, and RELAX-AHF, likely reflects the fact that early symptom relief is primarily driven by decongestion and hemodynamic stabilization, whereas long-term outcomes are determined by the underlying disease process and neurohormonal activation, which are not necessarily modified by short-term therapy. These consistent findings suggest that dyspnea alone has limited value as a surrogate marker for hard clinical outcomes and should not be used as the sole basis for evaluating treatment efficacy in AHF. Nonetheless, its assessment remains clinically relevant as a patient-centered measure of comfort and treatment response. In routine practice, patient evaluation should be based on periodic monitoring of dyspnea trajectory, complemented by objective measurements and prognostic markers. Effective relief of dyspnea and optimization of patient care require an integrated approach that addresses both the complex pathophysiology of the symptom and the influence of comorbidities.