Heart failure (HF) is a multifaceted clinical syndrome defined by structural or functional cardiac dysfunction stemming from diverse etiologies, a condition that may progress to serious adverse cardiovascular events [1]. Despite a stabilization in incidence and mortality rates in developed nations, heart failure persists as a major global public health challenge with a high overall prevalence [2,3]. The results of the latest epidemiological survey in China indicate that the total number of prevalent cases of heart failure increased from 7.03 million in 1990 to 18.51 million in 2019, showing an increasing trend year by year, with a prevalence rate of 1.3% among residents aged 35 and older [4,5]. Characterized by high prevalence, mortality, and readmission rates, heart failure imposes a significant burden on public health, and the early detection and timely intervention of heart failure are crucial for improving patient outcomes and altering the disease trajectory [6].

The pathogenesis of heart failure is characterized by intricate crosstalk among multiple signaling pathways. Cardiac fibrosis, a key pathological feature of heart failure resulting from excessive collagen deposition, plays a crucial role in disease progression [7]. Accumulating evidence has further underscored that a spectrum of acute and chronic inflammatory responses serves as a central mediator in the pathogenesis of cardiac fibrosis [8]. A range of inflammatory mediators, including C-reactive protein and interleukins, have been consistently identified to exhibit a marked upregulation in patients with HF, and these factors are closely correlated with the long-term prognostic outcomes of HF patients [9,10].

HMGB1 was originally characterized as a highly conserved nuclear protein that is abundantly localized within the nucleosomes of eukaryotic organisms. Functionally, high mobility group box 1 (HMGB1) exerts multiple nuclear roles: it contributes to the stabilization of nucleosome assembly, facilitates the initiation and progression of gene transcription, and modulates the transcriptional activity of steroid hormone receptors [11]. Emerging evidence has demonstrated that HMGB1 also functions as a prototypical damage-associated molecular pattern (DAMP) molecule: upon cellular exposure to exogenous stress stimuli (e.g., oxidative stress, ischemia, or mechanical injury), it is translocated from the nucleus to cytoplasm, and subsequently secreted into the extracellular microenvironment [12]. Once extracellular, HMGB1 exerts its biological effects by engaging with pattern recognition receptors such as Toll-like receptors and the receptor for advanced glycation end products, and this signaling axis has been closely linked to the pathogenesis of a broad spectrum of inflammatory and metabolic disorders [13].

Although elevated HMGB1 levels have been documented in both human heart failure patients and murine models [14,15], comprehensive cross-sectional data detailing its levels and dynamic changes in the patient population remain scarce. This study aims to quantify serum HMGB1 levels in a well-characterized heart failure cohort to evaluate its potential as a clinically relevant diagnostic and therapeutic novel biomarker.

Our study aimed to investigate the association between serum HMGB1 and HF. Our key findings revealed that serum HMGB1 levels were markedly elevated in patients with HF compared to the control group (1.53 ng/mL vs. 0.93 ng/mL, p < 0.001), and exhibited a marked dose-dependent correlation with HF risk. Additionally, HMGB1 demonstrated notable predictive value for HF, with an AUC of 0.736, and multivariable logistic regression analyses further confirmed HMGB1 as a novel biomarker for HF. The 30% missing rate of high-sensitivity C-reactive protein (hs-CRP) data in actual collection precluded valid statistical analysis, so we used SIRI as an alternative inflammatory biomarker. Notably, our findings confirmed a positive HMGB1 and SIRI correlation, and their combination improved diagnostic performance. This is consistent with prior studies establishing the prognostic value of composite inflammatory markers in patients with heart failure with reduced ejection fraction, particularly those undergoing device therapy [17], thereby adding to the growing body of evidence for the inflammatory pathogenesis underlying heart failure.

Heart failure, as the end stage of multiple cardiovascular diseases, is characterized by complex structural, functional, and molecular pathologies that collectively impair cardiac output [18]. Among the key drivers of HF progression, inflammation and myocardial fibrosis have been increasingly recognized as central pathological processes [19], which are mediated through intricate interactions between macrophages and fibroblasts via paracrine signaling [20]. In this context, high mobility group protein 1, a DAMP, has garnered increasing interest for its potential role in HF pathogenesis. We also acknowledge that a single inflammatory indicator cannot fully explain the complexity of HF. Currently, numerous studies integrate multiple indices to predict heart failure, thereby enhancing the persuasiveness of their findings. For instance, a previous study validated albumin’s prognostic role in HF patients treated with cardiac resynchronization therapy defibrillators [21], as albumin integrates nutritional and inflammatory status, it effectively improves risk stratification. Consequently, various composite indices based on albumin have also been shown to predict long-term mortality in HF patients [22,23]. Similarly, the combination of HMGB1 and SIRI in our study strengthens the inflammatory mediating role of HMGB1, further reinforcing the credibility of such multi-index integration strategies.

Multiple studies have shown that macrophages secrete substantial HMGB1 into the cytoplasm or extracellular space upon external stimulation, mediating various diseases via activation of downstream signaling pathways [24,25]. Notably, elevated HMGB1 levels significantly influence prognosis and long-term recovery in Intensive Care Unit (ICU) patients [26], highlighting the clinical value of early HMGB1 modulation. In cardiovascular diseases, HMGB1 has been increasingly implicated: a study of 258 patients found significantly elevated serum HMGB1 within 24 h of onset for unstable angina and non-ST-segment elevation, with follow-up levels independently predicting cardiovascular mortality [27]; another study identified HMGB1 as a biomarker for acute myocardial infarction regardless of HF status [28]; and a prospective study demonstrated elevated HMGB1 in HF patients, with levels independently predicting 12-month mortality [29]. While previous studies have identified HMGB1 as a serum biomarker, most were conducted in cohorts of patients with MI and featured relatively small sample sizes. The novelty of our study resides in its specific focus on patients presenting with chest pain: we systematically assessed the diagnostic value of HMGB1 for HF in this distinct population and established its optimal diagnostic cutoff value via RCS and ROC curve analyses. We found that after sequential adjustment for demographic, clinical, and laboratory covariates, HMGB1 remained independently associated with HF (fully adjusted OR = 2.273, 95% CI: 1.410–3.663, p < 0.001), confirming robustness independent of major confounders. Notably, unlike some traditional cardiovascular biomarkers such as NT-proBNP that are more closely linked to hemodynamic changes, HMGB1 reflects core inflammatory and fibrotic pathological processes of HF, complementing existing markers for more comprehensive risk stratification [30]. But the role of HMGB1 as an inflammatory DAMP in the pathogenesis of HF remains to be further explored: it is currently unclear whether HMGB1 is a causal factor, a secondary outcome of heart failure, or both [31]. This unresolved controversy leads to the inherent limitations of a single serum HMGB1 test, restricting the clinical application of our research findings. While our study’s sample size sufficiently validated this association, the study design limited our ability to establish temporal causality and long-term prognostic value; we plan to address this gap through prospective follow-up studies to further clarify HMGB1’s utility in HF management.

Our correlation analysis demonstrated that elevated HMGB1 levels were significantly associated with adverse cardiac remodeling, as reflected by increased LAD and LVEDD, coupled with a decrease in LVEF. These structural and functional impairments are well-established hallmarks of heart failure severity [32]. Concurrently, HMGB1 showed a positive correlation with the SIRI, reinforcing its pro-inflammatory role [33]. Although multiple multivariate scoring systems and biochemical biomarkers are currently available for predicting heart failure progression [34], our study highlights HMGB1’s unique advantage, achieving a predictive AUC of 0.736 for heart failure onset. Furthermore, a combined model integrating HMGB1 with LVEF, LVEDD, LAD, and SIRI demonstrated superior predictive accuracy (AUC = 0.807). In our study, patients at baseline exhibited mild left ventricular diastolic dysfunction, as shown by the mitral inflow E/A ratio and E/e′ ratio, which is a typical feature of heart failure. Notably, the E/e′ ratio rather than the E/A ratio showed significant differences between groups, highlighting the higher sensitivity of the former to diastolic load. After adjusting for these diastolic function indices, the association between HMGB1 and heart failure remained robust, indicating that its pathogenic effects are independent of the baseline diastolic dysfunction. However, we acknowledge the study limitation is the inability to directly compare HMGB1 with the gold-standard NT-proBNP in ROC and multivariate analyses [35], due to extensive missing NT-proBNP data within 24 hours of admission and exclusion of delayed data to avoid early treatment bias. The New York Heart Association classification, a key clinical parameter that categorizes heart failure severity based on symptoms and physical activity, was not available due to the retrospective, real-world design of this study. Given its critical role in heart failure assessment [36], future research will systematically incorporate these indicators.

A key finding of our study was the significant positive correlation between HMGB1 levels and advancing age. Patients in the highest HMGB1 tertile had a median age of 70 years, substantially older than the cohort average of 65 years. Current research indicates that one hallmark of aging is the gradual accumulation of chronic, low-level systemic inflammation [37]. This association is biologically plausible given that aging is characterized by chronic low-grade inflammation driven by senescent cells releasing pro-inflammatory mediators via the senescence-associated secretory phenotype (SASP) [37,38]. The age-related rise in HMGB1 suggests that aging may amplify its detrimental effects on cardiac function. Importantly, the consistent association between HMGB1 and HF across all age subgroups underscores its robust predictive value independent of aging-related inflammatory changes. Moreover, the influence of renal function on HMGB1 levels represents a significant consideration. Studies indicate that impaired kidney function can reduce the clearance capacity of HMGB1, leading to its accumulation in plasma [39]. Furthermore, in patients with chronic kidney disease, serum HMGB1 levels are markedly elevated and closely associated with disease progression and prognosis [24]. However, in the present study, the potential impact of this factor was mitigated by the exclusion of patients with severe kidney disease during participant enrollment.

As noted above, our study has several limitations that should be acknowledged, along with corresponding future improvements. First, despite the relatively large sample size supporting the assessment of serum HMGB1’s association with HF, the single-center design may limit the generalizability of the results. Second, the control group consisted of patients with chest symptoms but without concurrent HF. If these patients also had other cardiovascular diseases that could elevate HMGB1 levels, this might introduce bias; subsequent research should include age and sex matched healthy controls to minimize confounding. Third, HMGB1 levels were only measured within 24 hours of hospitalization, precluding the tracking of dynamic changes throughout disease progression. The longitudinal monitoring of HMGB1 during treatment and follow-up would help clarify its role in disease trajectories. Additionally, a potential limitation of this study is that cardiac function parameters such as NT-proBNP and LVEF may have been influenced by concomitant medications in our patient cohort, which could introduce bias into the results [40]. To address these questions, our team plans to initiate a prospective follow-up study. This study will involve the systematic collection of comprehensive clinical data, including detailed HF etiology, disease duration, and other potential confounders. With these data, we will perform more rigorous multivariable statistical adjustments. Furthermore, a study focusing on predicting myocardial injury in elderly patients demonstrated that machine learning could enhance prediction accuracy [41]. In future research, we may refer to the machine learning approach employed in this study to further optimize our model by capturing nonlinear interactions. By integrating the interactions between HMGB1 and other inflammatory indicators, we will further verify the association between HMGB1 and HF, explore its potential causal role, and analyze key clinical outcomes, thereby endowing this study with greater clinical practical value.

Our study investigated the significance of HMGB1 in heart failure, providing predictive value that complements current clinical parameters. As a readily measurable circulating biomarker linked to fundamental disease mechanisms, HMGB1 holds significant promise for improving HF risk assessment and may serve as a potential biomarker for the development of innovative therapeutic strategies.

The datasets used and analyzed during the current study can be obtained from the corresponding author upon reasonable request.

XTJ, XF designed the study, collected the data, and wrote the manuscript; and WL conducted the ELISA experiments; SLG contributed to data analysis and helped in analyzing the data; XPP, XW made substantial contributions to conception, design, and coordination of the study. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

This study adhered to the Helsinki Declaration and was approved by the Ethical Review Board of The First Affiliated Hospital of Nanchang University (Ethical Number: IIT2024849-1). Written informed consent was obtained from all participants or their legal guardians.

Not applicable.

This research is supported by the National Natural Science Foundation of China [Grant Nos. 82400462].

The authors declare no conflicts of interest.

During the preparation of this manuscript, ChatGPT-3.5 was used exclusively for spelling and grammatical checks. The authors have subsequently reviewed and revised all content independently and take full responsibility for the final version of the manuscript and all its scientific assertions.

Supplementary material associated with this article can be found, in the online version, at https://doi.org/10.31083/RCM49717.