BIA-estimated hypervolemia correlates with signs of congestion in AHF, including raised levels of natriuretic peptides, and peripheral or lung oedema [12]. While BIA may not add significant value when congestion is clinically evident, it may enhance its detection in subtler forms. In this cohort, subclinical congestion was more prevalent (57%) than previously reported (17%) [13]. This low-grade congestion may not be detected on physical examination, even by experienced physicians [4], highlighting the importance of multiparametric congestion assessment using biomarkers and non-invasive tools [5, 13,14,15]. BIA appears more accurate in detecting peripheral oedema than natriuretic peptides [16]. In this study, although NT-proBNP correlated with ECW/TBW, it did not differ between congestive groups, suggesting that mild-to-moderate elevation of natriuretic peptides may not effectively distinguish these states [17], a situation not rarely seen in clinical practice. Although half of the cohort was overweight/obese, which limits volemic assessment by physical examination, this did not correlate with ECW/TBW. BIA has shown good performance in detecting congestion and guiding diuretic therapy in obese HF patients [18], thereby broadening its potential in the highly heterogeneous HFpEF population. Diagnosing HFpEF in patients with chronic unexplained dyspnoea remains challenging, as it often relies on probability estimates derived from a combination of multiple variables. Even an elevated NT-proBNP is insufficient for HFpEF diagnosis, particularly in obese patients [19]. We speculate that BIA could aid in the reclassification of patients with an intermediate HFpEF probability by identifying subtle congestion, as demonstrated in a recent computed tomography algorithm for detecting pulmonary congestion [20]. This approach could potentially reduce the need for the often inaccessible invasive hemodynamic stress testing and facilitate HFpEF diagnosis.

BIA-guided therapy enables a safer transition of AHF patients to outpatient care by detecting residual congestion [15, 21]. BIA-estimated congestion predicts WHF events in outpatient settings [13, 22,23,24]. Notably, LVEF was not associated with the degree of congestion or outcomes [17]. A recent prospective study in HFrEF outpatients showed that BIA-estimated hypervolemia was independently associated with WHF events, outperforming other surrogate markers of congestion, including NT-proBNP, and indexes of LV filling pressure and central venous pressure (CVP). In contrast, in our study, the predictive accuracy of NT-proBNP and ECW/TBWz‑score was comparable. Although Rodriguez-Lopez’s study did not specifically address subclinical congestion, a volume excess of 1.2 L predicted WHF events [13]. Similarly, in our study ECW/TBWz‑score > 2 corresponds to an average absolute volume overload of 1 L. In addition, all patients identified with congestion had ECW/TBW > 0.39, a cutoff previously linked to adverse HF outcomes [23]. Challenging this evidence, Curbelo et al. found that NTproBNP, lung ultrasound and CVP, but not ECW/TBW, were associated with WHF events in CHF outpatients [25], though their study reported 60% ECW, which is a paradoxical water distribution, indicating a potential methodological issue.

HF presents with two primary congestion patterns—intravascular and interstitial—which often coexist [1]. In HFrEF, interstitial congestion predominates and results from saturation of different fluid buffering mechanisms, such as venous capacitance, lymphatic drainage, and interstitial hydrostatic pressure [26]. Conversely, rapid fluid redistribution into the intravascular space, with increased pulmonary and cardiac filling pressures, usually presents with acute pulmonary oedema [1]. Identifying the dominant congestion phenotype has therapeutic implications, particularly for diuretic and vasodilator therapy. While BIA correlates with body weight and lung ultrasound changes, Curtain et al. showed that it does not correlate with single measurements of pulmonary capillary wedge pressure (PCWP) [27]. This discrepancy can be attributed to the use of BIA devices that provide only arbitrary units of impedance, which are not suitable for single-point measurements. However, this limitation can be overcome by tracking relative impedance changes over time or by quantifying fluid volumes in specific compartments using advanced BIA devices, as demonstrated in our study.

HF progression is associated with a loss of muscle mass and a ‘counterintuitive’ reduction in both ECW and TBW can be observed [28]. As muscle mass is a major contributor to TBW, ECW/TBW indexes should be the preferred metric to assess fluid redistribution and congestion in HF. Although muscle function was not assessed in this study—a key requirement for diagnosing sarcopenia [10], 14% of the patients had low ASMI, consistent with previous reports in CHF cohorts [29]. Furthermore, low ASMI correlated with the degree of congestion and other surrogate markers of HF severity. However, it is important to note that current BIA devices may underestimate the true prevalence of sarcopenia in HF, particularly in congestive states where BIA-derived muscle mass tends to be overestimated [30]. BIA allows the identification of patients at risk of sarcopenia who may benefit from interventions, such as dietary and exercise programs [28]. Another valuable BIA parameter is phase angle (PhA), a marker of cellular health, with PhA < 4.2° linked to poor nutritional status and adverse events in HF [14]. In our study, PhA was inversely correlated with congestion severity. However, larger studies are needed to validate PhA as as a reliable indicator of cellular health in HF patients, particularly in cohorts with comparable fluid status.