Non-invasive quantification of haemodynamic imbalance in acute heart failure Authors: Hiroshi Morishita Published: 2026 ABSTRACT AbstractBackground and AimsAcute heart failure (AHF) is commonly regarded as a worsening stage of chronic heart failure (CHF), yet the haemodynamic transition to decompensation remains incompletely defined. This study evaluated whether a non-invasive physiology-guided approach based on venous return (VR)–cardiac output (CO) balance can provide a framework to characterize transitional circulatory states preceding overt AHF.MethodsA total of 193 patients with chronic heart failure, hypertensive heart disease, or AHF were analysed. Mean circulatory filling pressure and right atrial pressure were estimated non-invasively using inferior vena cava diameter indexed to body surface area and extracellular-to-total body water ratio, respectively. VR–CO imbalance was quantified using a composite mismatch index.ResultsThe distribution of VR–CO mismatch demonstrated a bimodal pattern on Gaussian mixture modelling, with two components (means ± SD: 2.25 ± .96 and 5.25 ± 2.61). All overt AHF cases were localized within the higher-mismatch component. Intermediate states exhibited increasing mismatch while maintaining relatively preserved forward flow, consistent with buffered circulatory states. A subset of patients exhibited marked haemodynamic mismatch without clinical congestion, representing a highly buffered pre-decompensated state.ConclusionsAHF may represent a failure of circulatory compensation in which venous return loading exceeds the capacity of forward flow. A non-invasive VR–CO framework reveals a continuum from compensated to buffered to overtly decompensated states and identifies a previously unrecognized highly buffered state preceding clinical deterioration. This physiology-guided approach may enable earlier detection of circulatory instability and provide a quantitative framework for risk stratification and therapeutic monitoring. Introduction Chronic and acute heart failure (AHF) are traditionally described using clinical classifications that do not fully reflect the underlying circulatory mechanics. In contrast, Guyton’s systems physiology conceptualizes the circulation as an interaction among venous return (VR), right atrial pressure (RAP), and mean circulatory filling pressure (MCFP).1–5 Venous return is governed by the pressure gradient between MCFP and RAP and is strongly influenced by stressed volume and venous capacitance.6,7 These principles emphasize that cardiac output (CO) is fundamentally determined by upstream stressed volume, venous tone, and their balance with RAP. However, the relationships among these determinants have not been translated into a clinically accessible framework. Central venous pressure—commonly interpreted as a marker of preload—does not reliably reflect effective circulating volume or VR dynamics, as emphasized by Magder.8 Instead, changes in venous capacitance, stressed volume, and the RAP–MCFP gradient more directly influence VR, and thus CO. Translating these physiological principles into clinical practice has remained challenging because direct measurement of these determinants is not routinely feasible in most patients. To address this gap, we developed a conceptual haemodynamic field map that positions VR and RAP surrogates (VRs/RAPs) within a unified co-ordinate framework centred on the balanced haemodynamic point (BHP), a conceptual reference representing the equilibrium at which VR and CO are balanced. This representation illustrates how alterations in extracellular volume, venous capacitance, and sympathetic tone may shift the circulation away from equilibrium. By linking clinically obtainable measurements to upstream circulatory forces, this framework provides a physiologic scaffold for examining transitional states that may precede overt decompensation. Acute heart failure is commonly regarded as the decompensated stage of chronic heart failure (CHF), characterized by pulmonary congestion, elevated filling pressures, and impaired perfusion.9,10 However, overt decompensation likely represents only the final clinical expression of a progressive circulatory imbalance. Prior to the onset of overt symptoms, compensatory mechanisms—including venous capacitance adjustments, interstitial compliance, and lymphatic drainage—may temporarily maintain circulatory stability despite rising VR.11,12 Exhaustion of these buffering mechanisms may therefore represent a key transition preceding clinical AHF. Despite advances in haemodynamic monitoring and guideline-directed therapy,13–15 the sequence of physiological changes from compensated to overtly decompensated states remains incompletely characterized. Established clinical classifications—such as those by Killip, Forrester, or Nohria–Stevenson10,16,17—describe congestion phenotypes but do not directly represent the circulatory control dynamics governing VR–CO balance. Consequently, the intermediate phase in which buffering capacity is strained but overt congestion has not yet developed remains poorly defined. To explore this transition, we applied a non-invasive physiology-guided framework integrating surrogate indices derived from echocardiography and bioimpedance analysis.18–25 Inferior vena cava diameter indexed to body surface area (IVC/BSA) was used as a surrogate reflecting venous filling and stressed volume related to MCFP,24–26 whereas the extracellular water-to-total body water ratio (ECW/TBW) was used as a surrogate of systemic volume loading associated with RAP.20–22,27 Previous studies have shown that bioimpedance-derived extracellular water correlates with systemic congestion and volume overload in heart failure populations.28,29 Similarly, inferior vena cava diameter has been widely used as a non-invasive estimate of central venous pressure and RAP.30,31 Cardiac output was estimated using a composite physiological index incorporating tricuspid regurgitant velocity (TRV), left atrial volume index (LAVI), and ECW/TBW, capturing aspects of forward-flow capacity and upstream pressure physiology.18,19,32 Using these indices, patients were mapped within a two-axis VR–CO framework to examine the distribution of circulatory states across a continuum of increasing imbalance between VR and forward flow. Within this framework, AHF may be interpreted as the clinical expression of a progressive imbalance between VR and forward flow that exceeds the capacity of vascular and lymphatic buffering mechanisms. The aim of the present study was therefore to evaluate whether non-invasive physiological surrogates may help identify transitional circulatory states associated with increasing VR–CO imbalance and to characterize the haemodynamic continuum leading to overt AHF. To illustrate the conceptual framework underlying this study, we developed a simplified water-tank model based on Guytonian circulatory physiology (Figure 1A–D). In this representation, extracellular volume (RAPs) and venous filling pressure [MCFP surrogate (MCFPs)] determine VR, while pulmonary interstitial buffering and CO determine the downstream flow response. The model highlights how progressive VR excess can transition from compensated equilibrium to buffered states and eventually to decompensated AHF when lymphatic buffering capacity is exceeded. Conceptual representation of circulatory states in a Guytonian water-tank model. (A) Compensated equilibrium venous return ≈ cardiac output. (B) Buffered state with increased venous return. (C) Highly buffered state with marked venous return–cardiac output imbalance. (D) De-buffered state corresponding to overt acute heart failure Methods Study design and setting This retrospective observational study included consecutive patients evaluated for acute or CHF decompensation at a single cardiovascular clinic between November 2024 and June 2025. All clinical assessments—including transthoracic echocardiography (TTE; Xario α300, Canon Medical Systems, Tochigi, Japan) and bioimpedance analysis (BIA; InBody 720, InBody Co., Seoul, Korea)—were performed under standardized resting conditions within a predefined interval of 1 h to minimize temporal variability. TTE was performed by experienced operators using commercially available ultrasound systems, in accordance with current guideline recommendations. Examinations were conducted in the supine position or, when clinically necessary, in a semi-recumbent position to accommodate patient symptoms. The inferior vena cava was imaged from the subcostal long-axis view. The maximum IVC diameter was measured at end-expiration, ∼2–3 cm proximal to its junction with the right atrium, and averaged over at least three consecutive cardiac cycles.24–26 Care was taken to optimize image quality and avoid oblique sections. Left ventricular ejection fraction (LVEF) was assessed using the Simpson’s biplane method. Left atrial volume was indexed to body surface area (LAVI), and TRV was measured using continuous-wave Doppler, in accordance with standard echocardiographic protocols.18,19,33 All measurements were performed using standardized acquisition protocols to minimize inter-observer variability. Representative values were selected based on image quality. BIA measurements were obtained in the standing resting position, as required by device specifications.20–23,27 All measurements were conducted after a predefined period of resting stabilization. Because body posture influences venous capacitance and fluid distribution, differences in measurement position (supine or semi-recumbent for TTE vs standing for BIA) may affect haemodynamic parameters, including IVC diameter. Postural changes alter the distribution between stressed and unstressed blood volume through shifts in venous capacitance, thereby modifying MCFP and VR. Supine positioning increases central venous volume, whereas standing reduces VR due to gravitational pooling. In patients with AHF, semi-recumbent positioning may further influence VR and fluid redistribution. Although measurements were performed under standardized conditions within a short time interval, residual postural effects cannot be fully excluded. Definition of acute heart failure AHF was defined as the rapid onset or worsening of dyspnoea, orthopnoea, pulmonary congestion, or peripheral oedema developing within hours to days and requiring urgent escalation of therapy, according to ESC and AHA/ACC/HFSA criteria.14,15 All AHF cases met Framingham criteria9 and demonstrated ≥ 1 of the following: Elevated natriuretic peptides32,34 Elevated natriuretic peptides32,34 Echocardiographic evidence of elevated filling pressures17,18 Echocardiographic evidence of elevated filling pressures17,18 Radiographic pulmonary congestion. Radiographic pulmonary congestion. Patients with isolated non-cardiac volume overload (renal failure, cirrhosis, nephrotic syndrome) were excluded. Non-invasive haemodynamic assessment A simplified Guytonian framework was applied to evaluate the relationship between VR, RAP, and circulatory output (CO) using clinically obtainable surrogate indices. The following surrogate variables were defined: RAPs: ECW/TBW, obtained using multifrequency BIA20–23,28,29,35 RAPs: ECW/TBW, obtained using multifrequency BIA20–23,28,29,35 MCFPs: IVC diameter indexed to body surface area (IVC/BSA), measured from the subcostal view by TTE.24–26,36 MCFPs: IVC diameter indexed to body surface area (IVC/BSA), measured from the subcostal view by TTE.24–26,36 To approximate VR and forward flow dynamics non-invasively, the following composite indices were calculated: VRs: a composite index integrating MCFP-related and RAP-related components VRs: a composite index integrating MCFP-related and RAP-related components CO surrogate (COs): a composite index incorporating TRV, LAVI, ECW/TBW, and IVC/BSA to approximate effective forward flow capacity. CO surrogate (COs): a composite index incorporating TRV, LAVI, ECW/TBW, and IVC/BSA to approximate effective forward flow capacity. The indices were defined as follows: Detailed derivation and physiological rationale for these parameters are provided in the Supplementary Methods. Because these indices share underlying physiological components, VRs and COs should be interpreted as integrative physiological surrogates, rather than independent direct haemodynamic measurements. The difference between VRs and COs (VRs − COs) was used as an exploratory index reflecting the degree of imbalance between VR drive and effective forward flow capacity. Definition of zone-based haemodynamic states Patients were categorized into four physiologic zones representing increasing VR–CO imbalance: Zone C (compensated): (VRs − COs) < 3.0 Zone C (compensated): (VRs − COs) < 3.0 Zone B (buffered state): ≥ 3.0 to < 4.0 Zone B (buffered state): ≥ 3.0 to < 4.0 Zone A1 (highly buffered state): ≥ 4.0 without overt congestion Zone A1 (highly buffered state): ≥ 4.0 without overt congestion Zone A2 (de-buffered HF/AHF): ≥ 4.0 with overt congestion. Zone A2 (de-buffered HF/AHF): ≥ 4.0 with overt congestion. Negative values (< 0) were infrequent and were included within Zone C. Thresholds were derived from the empirical distribution of VR–CO mismatch using Gaussian mixture modelling and are considered exploratory and cohort-specific. Clinical variables Clinical characteristics, vital signs, laboratory values, medications, and underlying cardiac diagnoses were extracted from the electronic medical record. Statistical analysis Continuous variables are presented as mean ± standard deviation (SD) or median [interquartile range (IQR)], as appropriate based on data distribution. Normality was assessed using the Shapiro–Wilk test. Between-group comparisons were performed using Welch’s t-test for normally distributed variables and the Mann–Whitney U test for non-normally distributed variables. Categorical variables were compared using the χ2 test or Fisher’s exact test, as appropriate. Haemodynamic indices were visualized using scatter plots to illustrate the distribution of circulatory states within the proposed framework. To explore the distribution of VR–CO imbalance, the difference between VRs and COs (VRs − COs) was analysed descriptively. Thresholds used for classification were derived from empirical distributional patterns and are therefore considered exploratory. All analyses were performed using Python 3.11 (SciPy, NumPy, and Matplotlib). A two-sided P < .05 was considered statistically significant. Ethical approval The protocol was approved by the Institutional Review Board of Morishita Heart Clinic (approval no. MHC-2025-02). Given the retrospective observational design, the requirement for written informed consent was waived. Results Haemodynamic distribution across the venous return–cardiac output gradient The distribution of (VRs − COs) exhibited a long rightward tail (Figure 2). Gaussian mixture modelling identified two components with an inflection near 3.0, and values ≥4.0 were predominantly observed in patients with overt AHF. Histogram of (venous return surrogate − cardiac output surrogate) with overt acute heart failure cases highlighted. The distribution showed a right-skewed pattern with two components identified by Gaussian mixture modelling. Values ≥ 4.0 were predominantly observed in patients with overt acute heart failure Patient distribution and clinical characteristics Among 193 evaluable patients, 138 (72%) were classified in Zone C, 25 (13%) in Zone B, 15 (8%) in Zone A1, and 15 (8%) in Zone A2. Patients in Zone A2 were older than those in Zone C (86.2 ± 5.0 vs 76.7 ± 9.8 yr, P < .001). BNP-equivalent levels were markedly higher in A2 (894.5 ± 976.5 vs 158.4 ± 291.6 pg/mL, P = .009). Estimated glomerular filtration rate showed a downward trend (46.9 ± 20.9 vs 53.5 ± 16.5 mL/min/1.73 m2), although the difference did not reach statistical significance. Left ventricular ejection fraction remained similar across zones (∼60%), indicating that the observed circulatory imbalance occurred largely independent of systolic pump function. Haemodynamic gradients A progressive rise in upstream haemodynamic surrogates was observed across zones (Table 1), including RAPs (.394–.410, P < .001), MCFPs (8.00–14.32, P < .001), and VRs (2.12–7.89, P < 10−15). COs values increased modestly across zones (from .25 to 1.30, P < 10−15) but remained insufficient to offset the larger rise in VRs. (VRs − COs) consequently increased markedly from 1.85 ± .73 (Zone C) to 6.62 ± 2.65 (Zone A2), demonstrating a substantial increase in mismatch between VR and effective forward flow. Table 1Baseline characteristics of patients stratified by haemodynamic zones based on the VR–CO gradient Table 1 clinical and haemodynamic characteristics across VR–CO imbalance zonesVR–CO zones were defined according to the (VRs − COs) gradient derived from non-invasive haemodynamic surrogatesZone C: (VRs − COs)< 3.0CompensatedZone B: (VRs − COs)≥ 3.0 to < 4.0Buffered stateZone A1: (VRs − COs)≥ 4.0 (without congestion)Highly buffered stateZone A2: (VRs − COs)≥ 4.0 (with congestion)De-buffered/AHF P-valuePopulation Number of patients138251515Demographics Sex, n (%) Male sex77 (56%)9 (16%)1 (7%)6 (40%).548 Female sex61 (44%)16 (52%)14 (93%)9 (60%) Age, yr76.7 ± 9.880.5 ± 10.087.1 ± 5.286.2 ± 5.05.2 × 10−5 BW, kg60.2 ± 14.653.9 ± 13.351.4 ± 12.053.8 ± 16.7.025Clinical variables SBP, mmHg128.6 ± 11.1129.2 ± 7.4132.7 ± 14.8124.1 ± 18.9.354 PR, b.p.m.70.1 ± 14.172.1 ± 10.371.6 ± 11.569.6 ± 25.0.916 BNP-eq, pg/mL158.4 ± 291.691.3 ± 89.1209.2 ± 296.1894.5 ± 976.56.2 × 10−11 eGFR, mL/min/1.73 m253.5 ± 16.559.1 ± 14.246.0 ± 6.846.9 ± 20.9.323 Atrial fibrillation, n (%)16 (12%)0 (0%)5 (33%)10 (67%)1.0 × 10−9Haemodynamic variables RAPs.394 ± .010.397 ± .013.403 ± .008.410 ± .0111.8 × 10−7 MCFPs7.998 ± .8239.589 ± .38011.647 ± 1.29014.316 ± 3.397< 10−15 VRs2.116 ± .7933.658 ± .2405.587 ± 1.1757.892 ± 3.217< 10−15 COs.254 ± .145.298 ± .163.530 ± .1991.277 ± .937< 10−15 (VRs − COs)1.852 ± .7323.360 ± .2315.057 ± 1.1156.615 ± 2.651< 10−15Echocardiographic parameters LVEF, %62.7 ± 9.762.7 ± 8.963.7 ± 8.756.9 ± 13.1.166 LVMI, g/m2121.6 ± 39.9109.8 ± 35.8110.4 ± 38.0141.7 ± 41.4.020 RWT.51 ± .13.48 ± .07.54 ± .11.59 ± .20.138 E/e′10.26 ± 2.9510.81 ± 2.6012.14 ± 4.1515.37 ± 5.318.7 × 10−4 LAVI, ml/m241.72 ± 17.9439.81 ± 21.7852.38 ± 16.6590.00 ± 54.073.1 × 10−12TRV, m/s2.26 ± .382.30 ± .392.73 ± .422.92 ± .346.1 × 10−10Underlying cardiac diseases, n (%) CAD23 (11%)4 (16%)0 (0%)2 (13%).843 HHD73 (61%)13 (52%)9 (60%)8 (52%) VHD27 (20%)5 (20%)3 (20%)2 (13%)DCM6 (3%)1 (4%)1 (7%)1 (7%) HCM9 (6%)2 (8%)2 (13%)2 (13%)HF/HHD phenotypes, n (%) HFrEF8 (6%)0 (0%)0 (0%)2 (13%).015 HFmrEF10 (7%)3 (12%)2 (13%)0 (0%) HFpEF53 (38%)11 (44%)10 (67%)12 (80%) HHD67 (49%)11 (44%)3 (20%)1 (7%)Medication use, n (%) SGLT2 inhibitor54 (39%)6 (24%)3 (20%)5 (33%).903 Loop diuretics16 (12%)5 (20%)4 (27%)10 (67%)3.9 × 10−5 Thiazide29 (21%)0 (0%)1 (7%)5 (33%).205 MRA29 (21%)10 (40%)9 (60%)6 (40%).118 CCB52 (38%)0 (0%)4 (27%)1 (7%).667 β-Blocker87 (63%)10 (40%)4 (27%)5 (33%).720 ARNI57 (41%)6 (24%)4 (27%)12 (80%).0047 ACEi/ARB84 (60%)12 (48%)3 (20%)3 (20%).010Patients were classified into four zones according to the VR–CO gradient (VRs − COs): Zone C (n = 138), Zone B (n = 25), Zone A1 (n = 15), and Zone A2 (n = 15).Variables include demographics, vital signs, natriuretic peptide levels, haemodynamic surrogate indices, underlying cardiac disease, echocardiographic parameters, atrial fibrillation, medication use, and heart failure phenotype. Values are presented as mean ± SD or n (%).BNP-eq indicates B-type natriuretic peptide or N-terminal pro–B-type natriuretic peptide (NT-proBNP) divided by 3.5.CAD, coronary artery disease; HHD, hypertensive heart disease; VHD, valvular heart disease; DCM, dilated cardiomyopathy; HCM, hypertrophic cardiomyopathy; ECW/TBW, extracellular-to-total body water ratio; IVC/BSA, inferior vena cava diameter indexed to body surface area; VRs, venous return surrogate; COs, cardiac output surrogate; LVEF, left ventricular ejection fraction; LVMI, left ventricular mass index; RWT, relative wall thickness; E/e′, ratio of early transmitral flow velocity to early diastolic mitral annular velocity; LAVI, left atrial volume index; TRV, tricuspid regurgitation velocity; AF, atrial fibrillation; SGLT2i, sodium–glucose cotransporter-2 inhibitor; MRA, mineralocorticoid receptor antagonist; CCB, calcium channel blocker; BB, beta-blocker; ARNI, angiotensin receptor–neprilysin inhibitor; ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; HFpEF, heart failure with preserved ejection fraction; HFmrEF, heart failure with mildly reduced ejection fraction; HFrEF, heart failure with reduced ejection fraction. Baseline characteristics of patients stratified by haemodynamic zones based on the VR–CO gradient Patients were classified into four zones according to the VR–CO gradient (VRs − COs): Zone C (n = 138), Zone B (n = 25), Zone A1 (n = 15), and Zone A2 (n = 15). Variables include demographics, vital signs, natriuretic peptide levels, haemodynamic surrogate indices, underlying cardiac disease, echocardiographic parameters, atrial fibrillation, medication use, and heart failure phenotype. Values are presented as mean ± SD or n (%). BNP-eq indicates B-type natriuretic peptide or N-terminal pro–B-type natriuretic peptide (NT-proBNP) divided by 3.5. CAD, coronary artery disease; HHD, hypertensive heart disease; VHD, valvular heart disease; DCM, dilated cardiomyopathy; HCM, hypertrophic cardiomyopathy; ECW/TBW, extracellular-to-total body water ratio; IVC/BSA, inferior vena cava diameter indexed to body surface area; VRs, venous return surrogate; COs, cardiac output surrogate; LVEF, left ventricular ejection fraction; LVMI, left ventricular mass index; RWT, relative wall thickness; E/e′, ratio of early transmitral flow velocity to early diastolic mitral annular velocity; LAVI, left atrial volume index; TRV, tricuspid regurgitation velocity; AF, atrial fibrillation; SGLT2i, sodium–glucose cotransporter-2 inhibitor; MRA, mineralocorticoid receptor antagonist; CCB, calcium channel blocker; BB, beta-blocker; ARNI, angiotensin receptor–neprilysin inhibitor; ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; HFpEF, heart failure with preserved ejection fraction; HFmrEF, heart failure with mildly reduced ejection fraction; HFrEF, heart failure with reduced ejection fraction. Echocardiographic correlates Markers of left-sided filling pressure and pulmonary vascular load increased progressively across zones, including E/e′ (10.3–15.4, P < .001), LAVI (41.7–90.0 mL/m2, P < 10−12), and TRV (2.26–2.92 m/s, P < 10−9). These changes paralleled increases in VRs, suggesting a consistent increase in left atrial and pulmonary pressures. Medication profiles Medication usage reflected increasing haemodynamic severity. Loop-diuretic prescriptions increased markedly from 12% (Zone C) to 67% (Zone A2) (P < .001). ARNI use increased from 41% to 80% (P = .0047), whereas SGLT2 inhibitors and β-blockers showed no significant differences between zones. Zone B exhibited intermediate therapy patterns, consistent with early congestion. Two-axis distribution of haemodynamic states To visualize the interaction between VR and forward flow, patients were mapped on a two-axis co-ordinate system defined by the VR surrogate (VRs) and the CO surrogate (COs) (Figure 3). The diagonal reference lines represent thresholds of (VRs − COs) = 3.0 and 4.0, corresponding to the onset of buffering and severe circulatory mismatch, respectively. Mapping patients on the RAPs–VRs plane (Figure 4) demonstrated a continuous physiologic spectrum. Zones C and B clustered around the empirically derived equilibrium point (ECW/TBW ≈ .393; VRs ≈ 2.36). Scatterplot of venous return surrogate vs cardiac output surrogate. Dashed diagonal lines indicate reference levels of venous return–cardiac output difference. The balanced haemodynamic point is shown as a reference equilibrium Scatterplot of venous return surrogate vs right atrial pressure surrogate (extracellular-to-total body water ratio). The balanced haemodynamic point is indicated as a reference equilibrium. Data points are color-coded according to haemodynamic classification Patients in Zones A1 and A2 shifted upward and rightward on the map, consistent with higher RAPs and disproportionately elevated VRs, consistent with excess stressed volume and impaired compensatory buffering. Dynamic haemodynamic transitions Five patients with paired measurements during acute decompensation and post-treatment recovery demonstrated clear zone migration, with VRs decreasing from 5.81 ± .64 to 3.90 ± 1.04 (P < .001) and (VRs − COs) decreasing from 7.01 ± .70 to 5.17 ± 1.02 (P < .001). Reductions in RAPs and VRs paralleled clinical recovery, illustrating the dynamic nature of the proposed haemodynamic framework. Discussion Mechanistic interpretation The observed haemodynamic transitions can be interpreted within a Guytonian framework of VR–cardiac function interactions.1–4 Acute circulatory imbalance may arise from either an increase in VR relative to CO or an insufficient CO response to VR (Figure 5A–D). Conceptual representation of Guytonian venous return and cardiac function curves illustrating acute and compensated states of circulatory imbalance. (A) Acute venous return excess. (B) Compensated venous return excess. (C) Acute low cardiac output. (D) Compensated low cardiac output To further conceptualize these dynamics, we employed a two-axis field map centred on a BHP, integrating surrogates of VR and RAP. Displacement from the BHP reflects progressive imbalance between VR and forward flow, as illustrated in Figure 6. In the present cohort, many patients maintained relatively preserved forward flow despite increasing VR (Zones B and A1). This pattern is consistent with buffered equilibria, in which compensatory mechanisms temporarily preserve circulatory stability. Conceptual representation of venous return–right atrial pressure interactions showing the effects of cardiac function and venous return determinants. The balanced haemodynamic point is indicated as a reference equilibrium. Shifts in cardiac function and venous return curves illustrate potential changes in circulatory state Mechanistic basis of decompensation In this study, overt AHF was characterized by a disproportionate increase in the venous-return surrogate relative to CO indices, accompanied by elevations in echocardiographic markers of filling pressure (E/e′, LAVI, TRV). Although decompensation may be triggered by diverse clinical factors, the resulting haemodynamic pattern appears consistent: excessive VR relative to effective forward flow, associated with upstream pressure elevation and pulmonary congestion, as observed in large registries such as OPTIMIZE-HF.34 A zone-based continuum of circulatory imbalance The distribution of (VRs − COs) demonstrated a bimodal structure, indicating the presence of relatively stable compensated and high-mismatch circulatory states. The statistical inflection near 3.0 corresponded to the onset of measurable imbalance, whereas higher values were predominantly associated with overt AHF. Within this framework, patients in the intermediate range exhibited features consistent with a buffered circulatory state, in which compensatory mechanisms—such as venous capacitance and lymphatic drainage—remain active despite increasing VR load. In contrast, patients in the high-mismatch range demonstrated characteristics of overt decompensation. These findings support the concept of a physiologic continuum of circulatory imbalance rather than a binary transition between compensated and decompensated states. Latent and highly buffered states A notable finding was the identification of patients in Zone A1, who demonstrated marked haemodynamic mismatch but remained clinically stable. These patients exhibited a haemodynamic substrate similar to overt AHF but appeared to retain sufficient vascular or lymphatic buffering to temporarily prevent congestion. This observation is consistent with previous studies showing that subclinical congestion often precedes overt AHF episodes.20–23,28,29 The identification of such buffered states may therefore provide an opportunity for earlier therapeutic intervention. Integration with existing clinical classifications Traditional AHF classifications—including Killip, Forrester, and Nohria–Stevenson10,16,17—describe clinical congestion phenotypes but do not directly quantify the underlying circulatory balance between VR and forward flow. The framework proposed here may complement these classifications by providing a quantitative representation of circulatory imbalance using non-invasive measurements. In this study, most hypertensive heart disease and HFpEF patients clustered within Zone C, whereas overt AHF cases shifted towards regions characterized by higher venous-return surrogates and elevated filling pressure. Importantly, left ventricular ejection fraction remained largely preserved across zones, indicating that circulatory imbalance may develop independently of systolic pump function. Because the present cohort was characterized by a relatively high mean LVEF, the findings primarily reflect haemodynamic patterns typical of HFpEF and hypertensive heart disease. Whether the proposed VR–CO framework applies similarly to HFrEF populations remains to be determined and warrants further investigation in cohorts with more advanced systolic dysfunction. Methodological considerations The indices used in this study are intended to approximate relationships between venous filling pressure, extracellular volume, and forward flow using non-invasive measurements. These indices should be interpreted as integrative physiological surrogates rather than direct haemodynamic measurements. Because certain variables are mathematically related, VRs and COs reflect co-ordinated aspects of circulatory function rather than independent parameters. Accordingly, the (VRs − COs) index should be understood as an approximate co-ordinate describing deviation from circulatory equilibrium within a Guytonian framework, rather than a simple difference between two independent measurements. This approach is consistent with the concept that individual pressure or flow variables are best interpreted within the broader context of VR–CO interactions, as emphasized in prior physiological studies.8 Clinical implications The ability to identify intermediate buffered states (Zones B and A1) may have several potential clinical implications: Identification of haemodynamic stress before overt congestion develops Identification of haemodynamic stress before overt congestion develops Visualization of treatment responses through zone migration Visualization of treatment responses through zone migration Physiologic stratification of patients with preserved ejection fraction Physiologic stratification of patients with preserved ejection fraction Potential integration into outpatient monitoring strategies. Potential integration into outpatient monitoring strategies. These observations suggest that serial assessment of circulatory balance may help to identify patients approaching decompensation. Limitations Several limitations of this study should be acknowledged. First, this was a single-centre, cross-sectional study, which limits generalizability and precludes assessment of temporal or causal relationships. Second, the proposed surrogate indices were not validated against invasive haemodynamic measurements and should therefore be interpreted as physiology-guided approximations rather than direct measures of VR or CO. Third, the framework does not incorporate resistance to VR, an important determinant in Guytonian physiology that was not directly assessed in this study. Fourth, the non-invasive measurements used in this study—including IVC/BSA and ECW/TBW—are subject to both technical and physiological variability. Postural differences between echocardiographic assessment (supine or semi-recumbent) and bioimpedance analysis (standing) may influence haemodynamic parameters, particularly IVC diameter and fluid distribution. In patients with AHF, semi-recumbent positioning for symptom relief may further modify VR and volume redistribution. Although all measurements were obtained under standardized resting conditions within a short predefined interval, residual postural effects cannot be fully excluded. Semi-recumbent positioning may influence the apparent magnitude of estimated VR, although the direction and magnitude of this effect remain uncertain. Fifth, the number of overt AHF cases was relatively small, which may limit the robustness of subgroup analyses. Finally, the thresholds used to define circulatory states were derived from empirical distributional patterns within the study population and should be considered exploratory. Accordingly, these findings should be interpreted as hypothesis-generating, and the prognostic value and clinical applicability of the proposed VR–CO framework require validation in larger prospective studies incorporating clinical outcomes. These limitations should be considered when interpreting the generalizability of the present findings. Future directions Future studies should examine whether zone-based circulatory mapping predicts clinical outcomes such as HF hospitalization or response to therapy. Conclusions Acute heart failure represents a state of circulatory compensation failure in which venous return loading exceeds the capacity of forward flow. Mapping this imbalance within a two-axis VR–CO framework reveals a continuum from compensated to buffered to overtly decompensated states. This physiology-guided approach provides a physiologically grounded extension of existing clinical classifications by enabling earlier identification of circulatory instability and offering a quantitative framework for tracking disease progression and therapeutic response. Further studies are warranted to determine whether this framework can improve risk stratification and guide therapeutic decision-making in heart failure. Supplementary Material