Your privacy, your choice

We use essential cookies to make sure the site can function. We also use optional cookies for advertising, personalisation of content, usage analysis, and social media.

By accepting optional cookies, you consent to the processing of your personal data - including transfers to third parties. Some third parties are outside of the European Economic Area, with varying standards of data protection.

See our privacy policy for more information on the use of your personal data.

for further information and to change your choices.

Skip to main content

Association between the haemoglobin glycation index and 30-day and 365-day mortality in patients with heart failure admitted to the intensive care unit

Abstract

Background

The hemoglobin glycation index (HGI) represents the difference between the observed and predicted values of haemoglobin A1c (HbA1c). However, the association between HGI and prognosis of heart failure (HF) is not completely clarified yet and requires more investigation. This study aimed to explore the connection between HGI and mortality in HF patients.

Methods

The data for the study were derived from the MIMIC-IV database from 2008 to 2019, a publicly available clinical database in intensive care. A linear regression equation between HbA1c and fasting blood glucose (FBG) was established to calculate predicted HbA1c. The endpoints were 30-day and 365-day all-cause mortality. Kaplan–Meier analysis was utilized to compare survival rates across groups differentiated by their HGI levels. The Cox regression models and restricted cubic spline (RCS) analysis were utilized to analyze the association between HGI and mortality.

Results

The study collected a total of 2846 patients with HF (40.1% male), of whom 305 patients (10.7%) died within 30 days and 954 patients (33.5%) died within 365 days. Kaplan–Meier curves revealed patients with higher HGI had significantly higher mortality risks (log-rank P < 0.001). A high HGI was significantly associated with 30-day mortality (adjusted HR [aHR]: 2.36, 95% CI: 1.74–3.20, P < 0.001) and 365-day mortality (aHR: 1.40, 95% CI: 1.16–1.68, P < 0.001) after adjustment for potential confounders. Likewise, each unit increase in the HGI correlated with a 1.42-fold higher risk of 30-day mortality (aHR: 1.42, 95% CI: 1.28–1.57, P < 0.001) and 1.19-fold higher risk of 365-day mortality (aHR: 1.19, 95% CI: 1.11–1.68, P < 0.001). RCS analysis suggested an L-shaped nonlinear association between HGI and clinical endpoints (P for nonlinearity < 0.001), with an inflection point value of − 1.295. Subgroup analysis and sensitivity analysis revealed that the correlation between HGI and 30-day and 365-day all-cause mortality remained consistent.

Conclusions

In ICU-admitted HF patients, HGI was independently associated with increased risks of 30-day and 365-day mortality and the identification of high HGI (> 0.709) provided a valuable tool for clinicians to detect high-risk populations. Integrating HGI into routine clinical practice might strengthen the prognosis-based decision making improve HF patient outcomes.

Background

Heart failure (HF), the advanced stage of various cardiac diseases, impacts estimated an 64 million individuals worldwide according to 2023 global statistics [1]. Current studies have shown that the 30-day and 1-year mortality for HF are 10.4% and 23.6%, respectively [2, 3]. Despite advancements treatment of HF, this high mortality rate remains a substantial burden on global health care systems [4]. Therefore, early risk stratification and precise identification of high-risk patients with HF are critical for optimizing treatment strategies and improving patient outcomes.

As a strong and independent risk factor for HF, approximately 40% of HF patients were reported to have diabetes mellitus (DM) [5,6,7,8,9]. The Framingham Heart Study found that individuals with diabetes had a higher risk of HF than those without DM, even after taking into account other risk factors [10]. Moreover, patients diagnosed with concurrent HF and DM experienced worse prognosis than those without DM [8]. Hemoglobin A1c (HbA1c), an index reflecting average blood glucose level during the last three months, is the most widely used marker of long-term glucose control [11]. Poor glycemic control is strongly linked to an elevated risk of HF, as research found that each 1% increase in HbA1c can cause 8% to 36% increasing risk of incident HF hospitalization among participants without DM or HF [8, 12]. However, HbA1c levels are always above or below and do not align with fasting blood glucose (FBG) levels in some populations [13]. This inconsistency between HbA1c levels and blood sugar levels may limit the accuracy of HbA1c measurements in guiding treatment regimens. With a better understanding of HbA1c, researchers discovered that a variety of factors like the differences in glucose transmembrane gradients, erythrocyte lifespan, glucose metabolism, passive hemoglobin glycation rates, enzyme abnormalities, and genetic factors can cause HbA1c changes [14,15,16,17]. Therefore, HbA1c level as a biomarker for prognostic assessment is inaccurate for all populations.

To improve the accuracy, Hempe et al. introduced the hemoglobin glycation index (HGI), a metric that quantifies the difference between observed HbA1c and blood sugar levels [13]. This method reflects the disparity between actual observed HbA1c and predicted HbA1c. Variations in the glycation of hemoglobin among people with identical blood sugar levels can be evaluated by the calculation of HGI [18]. Some research has implied that higher HGI levels were linked to diabetes complications, metabolic disorders and cardiovascular disease (CVD) [19, 20]. Wang M et al. discovered a higher HGI suggests an elevated risk of developing NAFLD in people with type 2 diabetes [20]. HGI, as a metabolic indicator, may have an impact on cardiovascular event risk through multiple mechanisms, such as inflammatory responses, oxidative stress, or the accumulation of glycosylation end products, and this effect may not vary linearly. For example, Wang Y et al. found both low and high HGIs were associated with an increased risk of major adverse cardiovascular events (MACE) in patients with type 2 diabetes with a non-linear relationship [19]. Qing et al., meanwhile, found that HGI levels had a non-linear relationship with cardiovascular and all-cause mortality in individuals with hypertension [21].

Not only that, the relationship between HGI and adverse events has shown varying and inconsistent results across different research populations. Some studies suggested that higher HGI is associated with a higher risk of cardiovascular events [22,23,24]. However, this hypothesis was challenged by research conducted by Sigrid et al., which found that patients with higher HGI exhibited a decreased risk of death after undergoing intensive treatment [25]. While glycemic variability and HbA1c have been associated with adverse outcomes in various chronic diseases, their predictive value in the context of HF remains underexplored. To date, there is only one study analyzing the relationship between HGI and heart failure [26]. Additionally, most studies have focused on stable outpatients, leaving a critical knowledge gap regarding the utility of HGI in ICU-admitted HF patients who face higher short-term mortality risks and more complex metabolic derangements. Therefore, the present study aimed to establish whether HGI is independently associated with short- and long-term mortality in HF patients in an ICU setting, aiming to offer new insights for prognostic assessment and enhancing the clinical management of heart failure.

Methods

Data source

This study was conducted using data from the MIMIC-IV database, which is publicly accessible and contains de-identified critical care data from patients admitted to the ICU of the Beth Israel Deaconess Medical Center (BIDMC) from 2008 to 2019. One author (Shuoyan An) obtained access to the database after meeting all required certifications (Record ID 39674606). MIMIC-IV has been approved by the Institutional Review Boards of BIDMC and MIT [27]. The data extracted from the MIMIC-IV database contained demographic information, medical records, lab results, medical therapies, survival data and so forth. Due to the de-identification of the data and informed consent has been waived, written informed consent was not required.

Study population

Patients who were diagnosed with HF based on the ninth and tenth revision of the International Classification of Diseases (ICD), were incorporated into this study. The specific ICD codes for HF are shown in Additional file 1: Table S1. The inclusion criteria required a diagnosis of "heart failure" as listed in the long title of the diagnostic dictionary. Exclusion criteria included: (1) age < 18 years, (2) individuals without ICU records, (3) patients with short ICU stay (≤ 24 h), (4) patients diagnosed as post-operative HF, (5) patients missing HbA1c levels or FBG values. For the patients admitted multiple times due to HF, only their initial admission was extracted. Ultimately, 2846 patients were included in the final analysis.

Data extraction and Clinical outcomes

We collected data from the MIMIC-IV database through the pgAdmin PostgreSQL tools. The variables extracted for analysis included: (1) Baseline characteristics: age, gender, ethnicity, BMI; (2) Vital statistics: heart rate, systolic/diastolic blood pressure (SBP/DBP), SpO2; (3) Comorbidities: hypertension, coronary artery disease (CAD), DM, atrial fibrillation, peripheral vascular disease, cerebrovascular disease, ventricular arrhythmia, cardiac arrest, chronic pulmonary disease, chronic kidney disease, liver disease, sepsis; (4) Laboratory parameters: WBC, RBC, platelet, hemoglobin (HGB), alanine aminotransferase (ALT), aspartate transaminase (AST), bilirubin, creatinine, blood urea nitrogen (BUN), FBG, HbA1C, sequential organ failure assessment score (SOFA); (5) Therapeutic interventions: mechanical ventilation, PCI, coronary artery bypass grafting (CABG), vasoactive agents, aspirin, clopidogrel, statin, β-blocker, warfarin; (6) Patient outcomes: dates of hospital and ICU admission and discharge, death date and time. Variables with greater than 20% missing values were deleted and the random forest imputation method was used to impute.

The clinical outcomes of this study were defined as 30-day and 365-day all-cause death from the moment of admission to the ICU. All-cause death included both cardiovascular and non-cardiovascular deaths.

HGI calculation and grouping

A linear regression equation between HbA1c and FBG was developed and the predicted HbA1c was obtained by the equation: predicted HbA1c = 0.442 * FBG (mmol/L) + 3.12 (Fig. 1) The HGI was then calculated by subtracting the predicted HbA1c from the actual measured HbA1c. Patients were divided into three groups based on HGI tertiles: tertile 1 (T1), HGI < −0.151; tertile 2 (T2), −0.151 ≤ HGI < 0.483; and tertile 3 (T3), HGI ≥ 0.483.

Fig. 1
figure 1

The linear regression equation between FBG and HbA1c: predict HbA1c = 0.442 * FBG (mmol/L) + 3.12 (r2 = 0.37 and P < 0.001)

Statistical analysis

Continuous variables were expressed as median (25th, 75th percentile) and compared using the Kruskal–Wallis test. Categorical variables were expressed as frequencies (n) and percentages (%) and compared using chi-square test. The Kaplan–Meier (KM) methods were employed to calculate the cumulative risk of mortality, and the log-rank test was utilized to differentiate across the three tertiles of the HGI. Cox regression models were performed to evaluate the connection between the HGI tertiles and 30-day and 365-day all-cause mortality. The HGI was included in the COX models as a continuous and categorical variable, with results presented as hazard ratios (HR) and 95% confidence intervals (CI). Restricted cubic splines (RCSs) were employed to investigate the linear or nonlinear correlation between the HGI and 30-day and 365-day all-cause mortality. While traditional linear models may underestimate or ignore the specific effects of HGI in high-risk populations, nonlinear analysis can more fully reveal this complex relationship and provide a basis for precision medicine interventions. Subgroup analysis was performed to investigate the heterogeneity among subgroups. Furthermore, we conducted a sensitivity analysis by propensity score matching (PSM) to reduce potential bias and balance baseline characteristics between groups [28]. First, we performed receiver operating characteristic curve (ROC) analysis to determine the optimal cut-off value of the HGI for predicting 30-day and 365-day mortality. The cut-off value was then used to divide the patients into two groups. Most of the covariates listed in Table 1 were included in the PSM, with the exclusion of FPG and HbA1c. All statistics were conducted using R (version 4.2.1) and two-tailed P < 0.05 was regarded as statistically significant.

Table 1 Baseline characteristics of the patients according to HGI tertiles

Results

Baseline characteristics

In this study, we initially identified 10,780 HF patients from the MIMIC-IV database and 9243 (85.7%) patients were excluded based on the exclusion criteria, leaving a final cohort of 2846 (14.3%) HF patients (Fig. 2). The median age of the patients was 72 years, with 40.1% being female, 79.3% with hypertension, and 47.6% with DM (Table 1). The patients were grouped according to HGI tertiles: T1: HGI < + 0.151 (n = 949); T2: − 0.151 ≤ HGI < 0.483 (n = 948); and T3: HGI ≥ 0.483 (n = 949). For each group, the baseline information for the patients is presented in Table 1. The statistically significant differences were observed between the three groups in terms of age, ethnicity, BMI, HR, hypertension, CAD, DM, AF, peripheral vascular disease, cerebrovascular disease, WBC, RBC, ALT, BUN, FBG, HbA1c, CABG, aspirin, clopidogrel, statin and so forth. The mortalities of 30-day and 365-day were 10.7% (n = 305) and 33.5% (n = 954), respectively (Table 1). The low-HGI group (HGI < − 0.151) was relatively young but had a higher BMI. The percentages of patients with hypertension, CAD, and DM were larger in the low HGI group.

Fig. 2
figure 2

Selection of study population from MIMIC-IV database

Survival analysis

The survival difference of patients was compared using KM survival analysis according to tertiles of HGI (Fig. 3). The T3 group has a significantly higher 30-day mortality than the other groups (log-rank P < 0.001) (Fig. 3A). Similarly, the 365-day mortality rate was markedly elevated in T3 group relative to T1 and T2. These results indicated that high HGI was correlated to poorer 30-day and 365-day survival outcomes in patients with HF (Fig. 3B).

Fig. 3
figure 3

Kaplan–Meier survival analysis for 30-day (A) and 365-day (B) all-cause mortality. A table below the Kaplan–Meier curves shows the number of patients at risk at different time points (in days)

The association between HGI and 30-day and 365-day all-cause mortality

Additional file 1: Table S2 listed the variables being included in the multivariable analysis, which were clinically relevant or statistically significant in the univariate model. To evaluate the independent association of the HGI with mortality risk, we employed three Cox regression models (Table 2). When HGI was analyzed as a categorical variable, patients in tertile 3 had a significantly higher rate of 30-day all-cause death (HR: 3.15, 95% CI: 2.34–4.24, P < 0.001) and 365-day all-cause death (HR: 1.74, 95% CI: 1.46–2.09, P < 0.001) than did the T1 group in the unadjusted Cox regression model (Model 1), while the all-cause death was similar in the T1 and T2 groups. After accounting for age, gender and BMI, we observed a consistent relationship between HGI and mortality outcomes (Model 2). In fully adjusted analysis (Model 3), higher HGI remained significantly associated with increased 30-day mortality (T3 vs. T1: aHR: 3.00, 95% CI: 2.23–4.04, P < 0. 001), and 365-day mortality (T3 vs. T1: aHR: 1.66, 95% CI:1.38–1.99, P < 0.001). When modeled as a continuous variable, per 1 standard deviation HGI increment was linked to a 42% increase in 30-day mortality (aHR: 1.42, 95% CI: 1.28–1.57, P < 0.001) and a 19% increase in 365-day mortality (aHR: 1.19, 95% CI: 1.11–1.28, P < 0.001) (Table 2).

Table 2 Predictive relationship of HGI with 30-day and 365-day mortality in cox regression models

We used RCS to assess the non-linear association between HGI and clinical endpoints. Figure 4 shows that, after adjusting for all covariates, a non-linear and L-shaped relationship between HGI and all-cause mortality at 30 days and 365 days, with inflection point values of −1.295. On the left of the inflection point, an increase in HGI was not linked to all-cause mortality and the 95% CI of HGI included 1. After the inflection point, a significant and sudden escalation in HR was recorded in the risk of mortality with increasing HGI (Fig. 4). Considering that HGI reflects interindividual glycemic variability, more attention should be paid to the extent of glycemic changes in patients during their stay in the ICU.

Fig. 4
figure 4

RCS analysis of the association of HGI with 30-day (A) and 365-day (B) all-cause mortality. The horizontal dotted black line represents the HR = 1. The orange curve shows the value of HR. The orange shaded area represents the 95% CI. The adjustment strategy is the same as the Model 3. The plot demonstrates an "L-shaped" association, where the risk stabilizes below the inflection point (HGI = − 1.295) and increases sharply above this threshold. RCS, restricted cubic spline

Stratified analysis and sensitivity analysis

Subgroup analysis was performed based on demographic features and risk factors, including age (< 65 vs. ≥ 65 years), gender, BMI (< 25 vs. ≥ 25 kg/m2), hypertension, DM, and CAD. A significant interaction between DM and HGI was observed in 30-day mortality (P for interaction = 0.047). No significant interactions were detected between HGI and any other variables (Fig. 5, 6).

Fig. 5
figure 5

Forest plots of subgroup analysis of HGI and 30-day all-cause mortality. Adjusting for the same covariates as in Model 3 except for the stratification variables. HT: hypertension; DM: diabetes mellitus; CAD, coronary artery disease

Fig. 6
figure 6

Forest plots of subgroup analysis of HGI and 365-day all-cause mortality. Adjusting for the same covariates as in Model 3 except for the stratification variables

To assess the robustness of our primary findings, we performed a sensitivity analysis. Before that, the ROC analysis was used to determine the optimal cut-off value of the HGI as a grouping condition (Additional file 2: Figure S1). This analysis identified an optimal cut-off value of 0.709 (30-day mortality: sensitivity 47.2% and specificity 80.1%; 365-day mortality: sensitivity 34.7% and specificity 81.0%). The baseline and clinical characteristics of both the low HGI (< 0.709) and high HGI (≥ 0.709) groups are presented in Table S3. After the adequate matching, a total of 671 pairs of patients in the two groups were successfully matched, resulting in cohorts with highly similar baseline characteristics (Additional file 1: Table S3). In the PSM cohort, the high TyG index remained significantly associated with an increased risk of 30-day and 365-day mortality (Additional file 1: Table S4).

Discussion

The present study investigated the relationship between HGI and the 30-day and 365-day risk of mortality in HF patients and found that patients with elevated HGI had higher mortality rates. Even after adjusting age, BMI, hypertension and other important potential confounders, HGI remained a significant independent risk factor for 30-day and 365-day all-cause mortality in HF. These correlations manifested a non-linear L-shaped pattern, with the risk of mortality significantly increasing when HGI was > − 1.295. These results highlight the significant clinical importance of HGI for assessing adverse outcomes, risk stratification, or disease management of patients with HF.

Diabetes is one of the major risk factors for HF and a recent study found that people with DM had a twofold increase in the incidence of HF [29]. The HbA1c testing is the gold standard for assessing long-term glycemic fluctuation in diabetes. However, the HbA1c level clinically detected represents only approximately 60–80% of the average blood sugar levels, and the remaining 20–40% of glycated hemoglobin variation may not be explained by HbA1c, which may contribute to adverse events [19, 30,31,32]. HGI was proposed to quantify this variation in HbA1c. This variation may arise from several factors beyond blood glucose concentration, such as redox status in red blood cells, and the activity of glycosylation process, erythrocyte genetic predisposition, hemoglobin oxygenation status [13, 33, 34]. These factors may influence the process of glucose metabolism and contribute to interindividual variability in the relationship between HbA1c and FBG [14].

As a marker of insulin resistance, HGI has been widely used in clinical practice. McCarter et al. found that elevated HGI was related to a 3- and sixfold increase in risk for retinal and renal microangiopathy, respectively [35]. Van et al. have found that a percentage increase in HGI corresponded to a 16% increase in risk of cardiovascular mortality [22, 24]. In addition, there are many studies have demonstrated that higher HGI was independently related to a high prevalence of cardiovascular diseases, including CAD, MACEs, stroke, and peripheral vascular disease, in individuals with prediabetes or early type 2 diabetes [23, 36, 37]. Consistent with previous research, our study found that patients with higher HGI had a significantly higher likelihood of cardiovascular and cerebrovascular diseases, such as cerebrovascular disease, ventricular arrhythmia, peripheral vascular disease and Cardiac arrest.

Several underlying mechanisms might mediate the association between HGI and clinical prognosis. Individuals with a high HGI often have higher HbA1c levels at the same blood sugar level compared to those with a low HGI; thus, the HGI might express the tendency of nonenzymatic glycosylation processes, promoting increased levels of advanced glycation end-products (AGEs), which play a major role in the progression of many chronic diseases, including diabetes, lipid disorders, coronary artery disease, and heart failure, etc. [36]. The AGEs trigger a signaling cascade that suppresses mitochondrial respiration, activates NF-kB-mediated inflammatory pathways (including TNFα converting enzyme), and promotes endothelial dysfunction while generating reactive oxygen species [38, 39]. AGEs can also promote the modification of low-density lipoproteins, contributing to oxidative stress and activating Toll-like receptor 4-mediated proinflammatory pathways [40]. Specifically, AGEs interact with the receptor for advanced glycation end-products (RAGE) to activate signaling pathways, such as NADPH oxidase, leading to excessive reactive oxygen species (ROS) production and oxidative damage to vascular and cardiac tissues; activate NF-κB signaling, which triggers the release of pro-inflammatory cytokines (e.g., IL-6, TNF-α), exacerbating systemic inflammation; impair endothelial function by reducing nitric oxide (NO) bioavailability, promoting vascular stiffness, and disrupting microcirculatory integrity [41]. All of these elevate cardiovascular risk. Another potential mechanism is the “metabolic memory”, that is that high blood sugar levels and HbA1c variability are “remembered”. Individuals with greater HbA1c variability may accumulate risk in the periods when HbA1c values are above the normal range, mediated by the same mechanism underlying the metabolic abnormalities, including oxidative stress [42, 43]. Stratton et al. found that HbA1c variability exerts a significant impact on outcomes more than average HbA1c as the risk of microvascular complications increases exponentially when HbA1c rises [43]. In addition, both insulin sensitivity and pancreatic β-cell function might be impaired in subgroups with elevated glucose [44]. Compared with persistent hyperglycemia, blood glucose fluctuations significantly exacerbate β-cell dysfunction and apoptosis, leading to reduced insulin secretion, further aggravating metabolic abnormalities and inducing glucose toxicity [45]. Also, HGI may be involved in aging acceleration by significantly decreasing the length of the telomere, which is associated with inflammation, oxidative stress and aging process and possibly mediated through TNFα signaling [46].

Pan et al. noted that both the lowest and highest tertiles of HGI values were correlated to a significantly increased risk of poor prognosis compared with moderate HGI in diabetic patients with ischemic stroke [47]. However, the ADVANCE study has suggested that a high HGI may associated with a lower risk of death in diabetic patients including type 1 diabetes and type 2 diabetes [25]. The ACCORD trial, a randomized trial that enrolled 10,251 participants with type 2 diabetes, found that only patients with the highest HGI tertile had a higher risk of mortality [48]. So the impact of HGI on CVD and mortality is still debated. This may partly be because these studies on HGI differed in study populations, causing inconsistencies in findings. In addition, glucose treatment strategies were different. For example, hypoglycemic therapy in ACCORD was intensified when the HbA1c level was ≥ 42 mmol/mol or self-monitored fasting plasma glucose (FPG) or 2-h glucose level was above a certain threshold [49]. The ADVANCE trial took into account both HbA1c and blood sugar levels [50]. When the HbA1c level was > 47 mmol/mol but FPG was relatively low, mealtime interventions were optimized. Other than that, the additional treatments differed between studies (i.e. all participants used sulfonylurea at the start in ADVANCE, while thiazolidinedione was frequently used in ACCORD). Thus, the different study designs could have an impact on the results. So far, there was only one study conducted by Weimeng et al. looking at the relationship between HGI and HF over the past decade and considered to be related to a reduction in cardiac death and all-cause deaths [26]. However, this previous study included special study populations with more than 80% of the patients with good blood sugar control (HbA1c less than 7.0%). Therefore, more prospective studies are required to verify the conclusions of Weimeng et al.

Our study focused on a population study taken from over 50,000 ICU admissions at BIDMC from 2008 to 2019, and patients at risk for relatively short-term (30 days and 365 days) mortality, unlike the study by Weimeng et al., which reports long-term outcomes (5 years). Our study found that individuals in the highest HGI tertile had a significantly increased risk of mortality, even after adjusting for confounders. These findings were consistent with previous studies suggesting a poor outcome in type 2 diabetes patients with the highest HbA1c [23, 37]. Further multivariate regression analysis showed that high HGI level is an independent risk factor.

Interestingly, we found a higher HR for T3 vs. T1 at the 30-day follow-up (HR = 2.36 (1.74–3.20)) compared to the 365-day follow-up (HR = 1.40 (1.16–1.68)). The higher HR observed at 30 days could be due to patients with heart failure being prone to malignant events in the acute phase, such as arrhythmia, cardiogenic shock, or acute kidney injury. The relatively lower HR at 365 days may indicate a dilution of the HGI-mortality association over time due to several factors: (1) competing risks: As time progresses, other comorbidities or competing risks (e.g., infections, cancer, or non-cardiovascular deaths) may play a larger role in determining long-term mortality, potentially attenuating the impact of HGI on outcomes. (2) adaptive changes: Patients with higher HGI might receive more intensive medical management or lifestyle interventions following acute events, which could mitigate their risk over the longer term. (3) survivor bias: Patients who survive the acute phase (30 days) may represent a more resilient subset, leading to a weaker association between HGI and mortality at 365 days. These findings highlight the importance of differentiating between short-term and long-term risks in heart failure management. HF patients with high HGI need closer monitoring and intervention during the acute phase to reduce the short-term risk of death. During long-term follow-up, intervention strategies should be adjusted by considering other risk factors and competition risk.

We found an L-shaped relationship between HGI level and risk of mortality in the RCS-based analysis. The possible mechanism is that when the HGI is in the lower range (HGI < − 1.295), the risk may stabilize because HbA1c near or below the predicted value may reflect better blood sugar control or a relatively stable metabolic state. In addition, patients with low HGI typically have a lower metabolic burden, which may have reduced the incidence of acute events. When HGI was above − 1.295, HbA1c was significantly higher than predicted, suggesting potential insulin resistance or increased blood sugar fluctuations. This metabolic disorder may increase inflammation, oxidative stress, and endothelial dysfunction, which significantly increases the risk of death in heart failure patients. HGI provides additional information beyond traditional glucose monitoring and holds the potential for identifying HF patients who may benefit from more aggressive management and therapeutic interventions, especially those with good blood sugar control but severe complications. Though the RCS curves presented a non-linear relationship while the results of group analysis showed a linear relationship, the linear and nonlinear models supplemented each other. The categorical analysis using tertiles is easy to interpret and provides intuitive clinical information, but obscure localized variations in risk, particularly near the inflection point, and may result in HRs that appear more linear when comparing tertiles, which may reflect either sample size limitations. RCS analysis, on the other hand, provides more accurate risk trends and is especially suitable for finding threshold effects or non-linear risk change points. Future studies could employ granular categorizations or larger sample sizes to better capture localized risk variations and validate the nonlinear relationship suggested by our RCS analysis.

Regarding the clinical value of HGI, some studies suggested that HGI was only a surrogate for HbA1c. However, a recent study suggested that HGI was independently related to cardiovascular events in patients with type 2 diabetes [51]. As an indicator reflecting genetic variation, HGI may help clinicians to better understand why some individuals with poor glycemic control escape complications while others with apparently good control develop severe complications and develop personalized treatment strategies [52]. One study found that high HGI was accompanied by a higher prevalence of obesity, higher TG levels and lower HDL-C levels [53]. From this perspective, statin therapy may help to reduce HGI levels. Furthermore, many studies have reported the active and regulatory role of ACEI/ARBs or SGLT2 inhibitors for glucose and metabolic abnormalities [54]. HGI and HbAc1 have something in common, suggesting that these drugs may also affect HGI levels. Additionally, we propose to integrate HGI calculation formulas into the clinical information system and automatically generate HGI values and their risk categories after patient HbA1c and FBG data are entered. An automatic reminder function can also be designed in the electronic medical record (EMR) to send an alert to the clinician when the HGI exceeds a predefined threshold. Training is also provided to clinicians and medical staff to help them understand the significance of HGI and its practical application in patient management. Similarly, we look forward to more clinical research on HGI in different populations to provide more evidence of causality between HGI and adverse outcomes in the future. In addition, HGI may have potential complementary effects with other biomarkers or metrics. For example, NT-proBNP is a marker that reflects myocardial tension, while HGI provides additional information on metabolic imbalances and chronic inflammation, mechanisms that work together to advance heart failure. Other emerging markers, such as galectin-3 [55], which is mainly associated with cardiac fibrosis and inflammation, may also enhance the predictive power of HGI in combination with HGI.

This study still has certain limitations. First, we retrieved corresponding clinical data about US seriously ill hospitalized patients with HF from the MIMIC-IV database and HGI was computed based on this study population, therefore may not be representative of the general US population and may have inherent selection bias. In addition, numerous epidemiological studies have shown higher HbA1c levels in Hispanic, Asian and African descents compared with Caucasians [21]. Our study population was primarily from the US population with a large percentage of whites, which may limit the generalizability of our findings to other ethnic groups or other populations. Therefore, our findings need to be further verified by more studies. Second, the lipid levels may be risk factors for poor prognosis; however, we did not analyze them in our analysis because of too many missing values (> 30%), which could compromise the robustness of statistical analysis. Third, although this study had a relatively large study sample and accounted for confounding factors in the data analysis, some bias may still exist due to the single-center and retrospective design issue. Unmeasured or inadequately measured factors may also influence the observed associations, such as lifestyle habits, dietary composition patterns and physical activities. Additionally, there was a potential for selection bias, as the analysis was limited to participants with both HbA1c and FBG measurements available for calculating HGI. Given the HGI in this study was computed based on the specific study population, its generalizability to other populations is limited. For each population, new regression models should be derived. We hope the regression models can be constructed based on several huge datasets to compute the HGI for each type of population in the future and future research can prioritize strategies to minimize missing data, such as prospective data collection or imputation methods, to better assess the interplay between glycemic variability, lipid metabolism, and clinical outcomes.

Conclusions

In conclusion, HGI may serve as a good indicator of blood glucose management and adverse prognosis in patients with HF. This may contribute to enhancing the management of blood glucose in HF patients, especially in patients with diabetes, and provide a decision basis for early intervention, such as optimizing blood glucose control or adjusting medication regimens. Further studies are needed to validate these findings in other populations and countries and explore HGI's interaction with other therapies.

Data availability

The data sets supporting the results of the present study are available in the MIMIC-IV database (website: https://mimic.physionet.org/).

Abbreviations

HGI:

Hemoglobin glycation index

HbA1c:

Haemoglobin A1c

HF:

Heart failure

DM:

Diabetes mellitus

FBG:

Fasting blood glucose

ICD:

International classification of diseases

BMI:

Body mass index

SBP/DBP:

Systolic/diastolic blood pressure

SpO2:

Oxygen saturation

CAD:

Coronary artery disease

WBC:

White blood cell

RBC:

Red blood cell

HGB:

Hemoglobin

ALT:

Alanine aminotransferase

AST:

Aspartate transaminase

BUN:

Blood urea nitrogen

SOFA:

Sequential organ failure assessment score

PCI:

Percutaneous coronary intervention

CABG:

Coronary artery bypass grafting

HR:

Hazard ratios

CI:

Confidence intervals

KM:

Kaplan–Meier

RCS:

Restricted cubic spline

References

  1. Dou J, Guo C, Wang Y, Peng Z, Wu R, Li Q, et al. Association between triglyceride glucose-body mass and one-year all-cause mortality of patients with heart failure: a retrospective study utilizing the MIMIC-IV database. Cardiovasc Diabetol. 2023;22(1):309. https://doi.org/10.1186/s12933-023-02047-4.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  2. Savarese G, Becher PM, Lund LH, Seferovic P, Rosano GMC, Coats AJS. Global burden of heart failure: a comprehensive and updated review of epidemiology. Cardiovasc Res. 2023;118(17):3272–87. https://doi.org/10.1093/cvr/cvac013.

    Article  CAS  PubMed  Google Scholar 

  3. Benjamin EJ, Blaha MJ, Chiuve SE, Cushman M, Das SR, Deo R, et al. Heart disease and stroke statistics-2017 update: a report from the american heart association. Circulation. 2017;135(10):e146–603. https://doi.org/10.1161/cir.0000000000000485.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Heidenreich PA, Bozkurt B, Aguilar D, Allen LA, Byun JJ, Colvin MM, et al. 2022 AHA/ACC/HFSA Guideline for the Management of Heart Failure: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation. 2022;145(18):e895–1032. https://doi.org/10.1161/cir.0000000000001063.

    Article  PubMed  Google Scholar 

  5. Lehrke M, Marx N. Diabetes mellitus and heart failure. Am J Med. 2017;130(6s):S40-s50. https://doi.org/10.1016/j.amjmed.2017.04.010[publishedOnlineFirst:2017/05/21].

    Article  CAS  PubMed  Google Scholar 

  6. Jankauskas SS, Kansakar U, Varzideh F, Wilson S, Mone P, Lombardi A, et al. Heart failure in diabetes. Metabolism. 2021;125: 154910. https://doi.org/10.1016/j.metabol.2021.154910.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Pandey A, Khan MS, Patel KV, Bhatt DL, Verma S. Predicting and preventing heart failure in type 2 diabetes. Lancet Diabetes Endocrinol. 2023;11(8):607–24. https://doi.org/10.1016/s2213-8587(23)00128-6.

    Article  CAS  PubMed  Google Scholar 

  8. Dunlay SM, Givertz MM, Aguilar D, Bhatt DL, Verma S. Type 2 diabetes mellitus and heart failure: a scientific statement from the American heart association and the heart failure society of America: this statement does not represent an update of the 2017 ACC/AHA/HFSA heart failure guideline update. Circulation. 2019;140(7):e294–324. https://doi.org/10.1161/cir.0000000000000691.

    Article  CAS  PubMed  Google Scholar 

  9. Echouffo-Tcheugui JB, Xu H, DeVore AD, Schulte PJ, Butler J, Yancy CW, et al. Temporal trends and factors associated with diabetes mellitus among patients hospitalized with heart failure: findings from get with the guidelines-heart failure registry. Am Heart J. 2016;182:9–20. https://doi.org/10.1016/j.ahj.2016.07.025.

    Article  PubMed  Google Scholar 

  10. Kannel WB, McGee DL. Diabetes and cardiovascular disease. The Framingham study. JAMA. 1979;241(19):2035–8. https://doi.org/10.1001/jama.241.19.2035.

    Article  CAS  PubMed  Google Scholar 

  11. Genuth S, Alberti KG, Bennett P, Buse J, Defronzo R, Kahn R, et al. Follow-up report on the diagnosis of diabetes mellitus. Diabetes Care. 2003;26(11):3160–7. https://doi.org/10.2337/diacare.26.11.3160.

    Article  PubMed  Google Scholar 

  12. van Melle JP, Bot M, de Jonge P, de Boer RA, van Veldhuisen DJ, Whooley MA. Diabetes, glycemic control, and new-onset heart failure in patients with stable coronary artery disease: data from the heart and soul study. Diabetes Care. 2010;33(9):2084–9. https://doi.org/10.2337/dc10-0286.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Hempe JM, Gomez R, McCarter RJ Jr, Chalew SA. High and low hemoglobin glycation phenotypes in type 1 diabetes: a challenge for interpretation of glycemic control. J Diabetes Compl. 2002;16(5):313–20. https://doi.org/10.1016/s1056-8727(01)00227-6.

    Article  Google Scholar 

  14. Wei X, Chen X, Zhang Z, Wei J, Hu B, Long N, et al. Risk analysis of the association between different hemoglobin glycation index and poor prognosis in critical patients with coronary heart disease-a study based on the MIMIC-IV database. Cardiovasc Diabetol. 2024;23(1):113. https://doi.org/10.1186/s12933-024-02206-1.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Malka R, Nathan DM, Higgins JM. Mechanistic modeling of hemoglobin glycation and red blood cell kinetics enables personalized diabetes monitoring. Sci Transl Med. 2016;8(359):359ra130. https://doi.org/10.1126/scitranslmed.aaf9304.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Khera PK, Joiner CH, Carruthers A, Lindsell CJ, Smith EP, Franco RS, et al. Evidence for interindividual heterogeneity in the glucose gradient across the human red blood cell membrane and its relationship to hemoglobin glycation. Diabetes. 2008;57(9):2445–52. https://doi.org/10.2337/db07-1820.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Brown SM, Smith DM, Alt N, Thorpe SR, Baynes JW. Tissue-specific variation in glycation of proteins in diabetes: evidence for a functional role of amadoriase enzymes. Ann N Y Acad Sci. 2005;1043:817–23. https://doi.org/10.1196/annals.1333.094.

    Article  CAS  PubMed  Google Scholar 

  18. Hempe JM, Yang S, Liu S, Hsia DS. Standardizing the haemoglobin glycation index. Endocrinol Diabetes Metab. 2021;4(4): e00299. https://doi.org/10.1002/edm2.299.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Wang Y, Liu H, Hu X, Wang A, Wang A, Kang S, et al. Association between hemoglobin glycation index and 5-year major adverse cardiovascular events: the REACTION cohort study. Chin Med J (Engl). 2023;136(20):2468–75. https://doi.org/10.1097/cm9.0000000000002717.

    Article  CAS  PubMed  Google Scholar 

  20. Wang M, Li S, Zhang X, Li X, Cui J. Association between hemoglobin glycation index and non-alcoholic fatty liver disease in the patients with type 2 diabetes mellitus. J Diabetes Investig. 2023;14(11):1303–11. https://doi.org/10.1111/jdi.14066.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Shangguan Q, Yang J, Li B, Chen H, Yang L. Association of the hemoglobin glycation index with cardiovascular and all-cause mortality in individuals with hypertension: findings from NHANES 1999–2018. Front Endocrinol (Lausanne). 2024;15:1401317. https://doi.org/10.3389/fendo.2024.1401317.

    Article  PubMed  Google Scholar 

  22. van Steen SC, Schrieks IC, Hoekstra JB, Lincoff AM, Tardif JC, Mellbin LG, et al. The haemoglobin glycation index as predictor of diabetes-related complications in the AleCardio trial. Eur J Prev Cardiol. 2017;24(8):858–66. https://doi.org/10.1177/2047487317692664.

    Article  PubMed  Google Scholar 

  23. Klein KR, Franek E, Marso S, Pieber TR, Pratley RE, Gowda A, et al. Hemoglobin glycation index, calculated from a single fasting glucose value, as a prediction tool for severe hypoglycemia and major adverse cardiovascular events in DEVOTE. BMJ Open Diabetes Res Care. 2021. https://doi.org/10.1136/bmjdrc-2021-002339.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Østergaard HB, Mandrup-Poulsen T, Berkelmans GFN, der Graaf YV, Visseren FLJ, Westerink J, et al. Limited benefit of haemoglobin glycation index as risk factor for cardiovascular disease in type 2 diabetes patients. Diabetes Metab. 2019;45(3):254–60. https://doi.org/10.1016/j.diabet.2018.04.006.

    Article  CAS  PubMed  Google Scholar 

  25. van Steen SC, Woodward M, Chalmers J, Li Q, Marre M, Cooper ME, et al. Haemoglobin glycation index and risk for diabetes-related complications in the action in diabetes and vascular disease: preterax and diamicron modified release controlled evaluation (ADVANCE) trial. Diabetologia. 2018;61(4):780–9. https://doi.org/10.1007/s00125-017-4539-1.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Cheng W, Huang R, Pu Y, Li T, Bao X, Chen J, et al. Association between the haemoglobin glycation index (HGI) and clinical outcomes in patients with acute decompensated heart failure. Ann Med. 2024;56(1):2330615. https://doi.org/10.1080/07853890.2024.2330615.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Johnson AEW, Bulgarelli L, Shen L, Gayles A, Shammout A, Horng S, et al. MIMIC-IV, a freely accessible electronic health record dataset. Sci Data. 2023;10(1):1. https://doi.org/10.1038/s41597-022-01899-x.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Xie E, Ye Z, Wu Y, Zhao X, Li Y, Shen N, et al. The triglyceride-glucose index predicts 1-year major adverse cardiovascular events in end-stage renal disease patients with coronary artery disease. Cardiovasc Diabetol. 2023;22(1):292. https://doi.org/10.1186/s12933-023-02028-7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. McAllister DA, Read SH, Kerssens J, Livingstone S, McGurnaghan S, Jhund P, et al. Incidence of hospitalization for heart failure and case-fatality among 3.25 million people with and without diabetes mellitus. Circulation. 2018;138(24):2774–86. https://doi.org/10.1161/circulationaha.118.034986.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Nathan DM, Kuenen J, Borg R, Zheng H, Schoenfeld D, Heine RJ, et al. Translating the A1C assay into estimated average glucose values. Diabetes Care. 2008;31(8):1473–8. https://doi.org/10.2337/dc08-0545.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Delpierre G, Veiga-da-Cunha M, Vertommen D, Buysschaert M, Schaftingen EV. Variability in erythrocyte fructosamine 3-kinase activity in humans correlates with polymorphisms in the FN3K gene and impacts on haemoglobin glycation at specific sites. Diabetes Metab. 2006;32(1):31–9. https://doi.org/10.1016/s1262-3636(07)70244-6.

    Article  CAS  PubMed  Google Scholar 

  32. Gonzalez-Covarrubias V, Sánchez-Ibarra H, Lozano-Gonzalez K, Villicaña S, Texis T, Rodríguez-Dorantes M, et al. Transporters, TBC1D4, and ARID5B variants to explain glycated hemoglobin variability in patients with type 2 diabetes. Pharmacology. 2021;106(11–12):588–96. https://doi.org/10.1159/000517462.

    Article  CAS  PubMed  Google Scholar 

  33. Soros AA, Chalew SA, McCarter RJ, Shepard R, Hempe JM. Hemoglobin glycation index: a robust measure of hemoglobin A1c bias in pediatric type 1 diabetes patients. Pediatr Diabetes. 2010;11(7):455–61. https://doi.org/10.1111/j.1399-5448.2009.00630.x.

    Article  CAS  PubMed  Google Scholar 

  34. Fiorentino TV, Marini MA, Succurro E, Succurro E, Andreozzi F, Sciacqua A, Hribal ML, et al. Association between hemoglobin glycation index and hepatic steatosis in non-diabetic individuals. Diabetes Res Clin Pract. 2017;134:53–61. https://doi.org/10.1016/j.diabres.2017.09.017.

    Article  CAS  PubMed  Google Scholar 

  35. McCarter RJ, Hempe JM, Gomez R, Chalew SA. Biological variation in HbA1c predicts risk of retinopathy and nephropathy in type 1 diabetes. Diabetes Care. 2004;27(6):1259–64. https://doi.org/10.2337/diacare.27.6.1259.

    Article  CAS  PubMed  Google Scholar 

  36. Ahn CH, Min SH, Lee DH, Oh TJ, Kim KM, Moon JH, et al. Hemoglobin glycation index is associated with cardiovascular diseases in people with impaired glucose metabolism. J Clin Endocrinol Metab. 2017;102(8):2905–13. https://doi.org/10.1210/jc.2017-00191.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Xu S, Qin Z, Yuan R, Cui X, Zhang L, Bai J, et al. The hemoglobin glycation index predicts the risk of adverse cardiovascular events in coronary heart disease patients with type 2 diabetes mellitus. Fronti Cardiovasc Med. 2022;9: 992252. https://doi.org/10.3389/fcvm.2022.992252.

    Article  CAS  Google Scholar 

  38. Akhter F, Chen D, Akhter A, Yan SF, Yan SD. Age-dependent accumulation of dicarbonyls and advanced glycation endproducts (AGEs) associates with mitochondrial stress. Free Radical Biol Med. 2021;164:429–38. https://doi.org/10.1016/j.freeradbiomed.2020.12.021.

    Article  CAS  Google Scholar 

  39. Jiang JX, Chen X, Fukada H, Serizawa N, Devaraj S, Török NJ. Advanced glycation endproducts induce fibrogenic activity in nonalcoholic steatohepatitis by modulating TNF-α-converting enzyme activity in mice. Hepatology. 2013;58(4):1339–48. https://doi.org/10.1002/hep.26491.

    Article  CAS  PubMed  Google Scholar 

  40. Hodgkinson CP, Laxton RC, Patel K, Ye S. Advanced glycation end-product of low density lipoprotein activates the toll-like 4 receptor pathway implications for diabetic atherosclerosis. Arterioscler Thromb Vasc Biol. 2008;28(12):2275–81. https://doi.org/10.1161/atvbaha.108.175992.

    Article  CAS  PubMed  Google Scholar 

  41. Fishman SL, Sonmez H, Basman C, Singh V, Leonid PL. The role of advanced glycation end-products in the development of coronary artery disease in patients with and without diabetes mellitus: a review. Mol Med. 2018;24(1):59. https://doi.org/10.1186/s10020-018-0060-3.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Ihnat MA, Thorpe JE, Ceriello A. Hypothesis: the “metabolic memory”, the new challenge of diabetes. Diabet Med. 2007;24(6):582–6. https://doi.org/10.1111/j.1464-5491.2007.02138.x.

    Article  CAS  PubMed  Google Scholar 

  43. Stratton IM, Adler AI, Neil HA, Matthews DR, Manley SE, Cull CA, et al. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study. BMJ. 2000;321(7258):405–12. https://doi.org/10.1136/bmj.321.7258.405.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Xin S, Zhao X, Ding J, Zhang X. Association between hemoglobin glycation index and diabetic kidney disease in type 2 diabetes mellitus in China: a cross- sectional inpatient study. Front Endocrinol. 2023;14:1108061. https://doi.org/10.3389/fendo.2023.1108061.

    Article  Google Scholar 

  45. Kim MK, Jung HS, Yoon CS, Ko JH, Jun HJ, Kim TK, et al. The effect of glucose fluctuation on apoptosis and function of INS-1 pancreatic beta cells. Korean Diabetes J. 2010;34(1):47–54. https://doi.org/10.4093/kdj.2010.34.1.47.

    Article  PubMed  PubMed Central  Google Scholar 

  46. Lyu L, Yu J, Liu Y, He S, Zhao Y, Qi M, et al. High hemoglobin glycation index is associated with telomere attrition independent of HbA1c, mediated by TNFα. J Clin Endocrinol Metab. 2022;107(2):462–73. https://doi.org/10.1210/clinem/dgab703.

    Article  PubMed  Google Scholar 

  47. Pan Y, Jing J, Wang Y, Liu L, Wang Y, He Y. Association of hemoglobin glycation index with outcomes of acute ischemic stroke in type 2 diabetic patients. Neurol Res. 2018;40(7):573–80. https://doi.org/10.1080/01616412.2018.1453991.

    Article  CAS  PubMed  Google Scholar 

  48. Hempe JM, Liu S, Myers L, McCarter RJ, Buse JB, Fonseca V. The hemoglobin glycation index identifies subpopulations with harms or benefits from intensive treatment in the ACCORD trial. Diabetes Care. 2015;38(6):1067–74. https://doi.org/10.2337/dc14-1844.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  49. Gerstein HC, Riddle MC, Kendall DM, Cohen RM, Robin G, Feinglos MN, et al. Glycemia treatment strategies in the action to control cardiovascular risk in diabetes (ACCORD) trial. Am J Cardiol. 2007;99(12A):34i–43i. https://doi.org/10.1016/j.amjcard.2007.03.004.

    Article  PubMed  Google Scholar 

  50. Patel A, MacMahon S, Chalmers J, Bruce N, Billot L, Woodward M, et al. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. N Engl J Med. 2008;358(2):2560–72. https://doi.org/10.1056/NEJMoa0802987.

    Article  CAS  PubMed  Google Scholar 

  51. Kim MK, Jeong JS, Yun JS, Kwon HS, Baek KH, Song KH, et al. Hemoglobin glycation index predicts cardiovascular disease in people with type 2 diabetes mellitus: a 10-year longitudinal cohort study. J Diabetes Complicat. 2018;32(10):906–10. https://doi.org/10.1016/j.jdiacomp.2018.08.007.

    Article  Google Scholar 

  52. Brownlee M. Advanced protein glycosylation in diabetes and aging. Annu Rev Med. 1995;46:223–34. https://doi.org/10.1146/annurev.med.46.1.223.

    Article  CAS  PubMed  Google Scholar 

  53. Marini MA, Fiorentino TV, Succurro E, Pedace E, Andreozzi F, Sciacqua A, et al. Association between hemoglobin glycation index with insulin resistance and carotid atherosclerosis in non-diabetic individuals. PLoS ONE. 2017;12(4): e0175547. https://doi.org/10.1371/journal.pone.0175547.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Li J, Xin Y, Li J, Zhou L, Qiu H, Shen A, et al. Association of haemoglobin glycation index with outcomes in patients with acute coronary syndrome: results from an observational cohort study in China. Diabetol Metab Syndr. 2022;14(1):162. https://doi.org/10.1186/s13098-022-00926-6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  55. Seropian IM, El-Diasty M, El-Sherbini AH, González GF, Rabinovich GA. Central role of Galectin-3 at the cross-roads of cardiac inflammation and fibrosis: implications for heart failure and transplantation. Cytokine Growth Factor Rev. 2024;80:47–58. https://doi.org/10.1016/j.cytogfr.2024.10.002.

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

We are grateful to all participants involved in this study.

Funding

This project received backing from the National High-Level Hospital Clinical Research Funding (2024-NHLHCRF-YS-01, 2024-NHLHCRF-JBGS-WZ-06, 2023-NHLHCRF-YXHZ-ZRMS-09), Capital's Funds for Health Advancement and Research (No.2022-1-4062), National Natural Science Foundation of China (No.82270352) and Beijing Research Ward Construction Clinical Research Project (2022-YJXBF-04-03).

Author information

Authors and Affiliations

Authors

Contributions

Z.G. and Y.L. conceived the research, analyzed the data, and drafted the manuscript. Y.L. made critical revisions to the manuscript. S.A. extracted data from the MIMIC-IV database. J.Z. played a role in the study's conceptualization and design, as well as in revising the manuscript. The final manuscript was reviewed and approved by all the authors.

Corresponding author

Correspondence to Jingang Zheng.

Ethics declarations

Ethics approval and consent to participate

This study was conducted according to the guidelines of the Declaration of Helsinki. The MIMIC-IV protocol was revised and approved by the Ethics Review Committee of the Massachusetts Institute of Technology and Beth Israel Deaconess Medical Center. Because the data were publicly available, ethics approval statements and informed consent were not required.

Consent for publication

All the authors read and approved the manuscript.

Competing interests

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Guo, Z., Li, Y., An, S. et al. Association between the haemoglobin glycation index and 30-day and 365-day mortality in patients with heart failure admitted to the intensive care unit. Diabetol Metab Syndr 17, 87 (2025). https://doi.org/10.1186/s13098-025-01661-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13098-025-01661-4

Keywords