Heart Failure (HF) is a prevalent and serious cardiovascular disease characterized by high incidence and mortality rates globally (1). Previous research indicates that approximately 40% of patients with HF also have diabetes, and HF is a significant contributor to morbidity and mortality in individuals with Type 2 Diabetes Mellitus (T2DM) (2). Early identification of individuals at high risk for HF is essential for timely intervention and the development of effective preventive strategies.

Blood glucose control has been shown to provide long-term cardiovascular benefits focusing on fasting blood glucose(FPG)and glycated hemoglobin (HbA1c) control (3). FPG alone has a low predictive value for cardiovascular outcomes. HbA1c is currently the most reliable measure for evaluating long-term blood glucose control (4, 5). HbA1c, utilized in the diagnosis and management of DM, reflects an individual’s average blood glucose over a three-month period and is currently the most commonly used surrogate marker of the effectiveness of glucose-lowering interventions (6, 7). HbA1c is not only utilized for the diagnosis and treatment of DM but also plays a significant role in the individualized management of patients with HF (8). However, the relationship between HbA1c and mortality in patients with both DM and HF has yielded conflicting results. Some cohort studies indicate that elevated HbA1c levels increase the risk of death in HF patients (9), while others suggest that higher HbA1c levels may reduce the risk of death (10, 11) Additionally, some studies have reported a U-shaped relationship between the two (12) Given that various factors, such as ethnicity, age, genetic variation, red blood cell (RBC) lifespan, iron deficiency, and anemia, can cause discrepancies between HbA1c and blood glucose levels, Hempe et al. first proposed the use of the Hemoglobin Glycation Index (HGI) to quantify individual variations in HbA1c levels (13). The HGI, which defined as is the difference between observed and predicted HbA1c, is derived from either mean or FPG and serves as a predictive tool for severe hypoglycemia and major adverse cardiovascular events (14).

Previous studies have indicated that the HGI is closely associated with adverse cardiovascular events (15, 16). However, both high and low levels of HGI have been linked to adverse outcomes in patients with CHF to varying degrees. Therefore, in the present study, the MIMIC-IV was database was utilized to construct linear regression equations to calculate the HGI and analyze the correlation between the HGI and adverse outcomes in patients with critical CHF.

A multitude of studies have demonstrated that the HGI is associated with a negative prognosis in cardiovascular diseases (CVDs), particularly in cases of coronary heart disease (CHD) and arrhythmia (19–21). While previous research has established a link between elevated HGI levels and an increased risk of CVD (22–24), evidence regarding the association between HGI and mortality in patients with congestive heart failure (CHF) remains inadequate. Therefore, this study developed a linear regression model utilizing MIMIC-IV data to evaluate the HGI in individuals with severe CHF and to investigate the relationship between HGI and adverse outcomes in these patients.

The data we gathered indicated that the mortality rate among patients was 15.9% within the initial 30 days of hospital admission, rising to 19.3% by 60 days, 22.1% at 90 days, and escalating to 30.7% by 365 days post-admission. In the fully adjusted Cox proportional hazards model, HGI demonstrated a significant correlation with mortality rates at both the 30-day and 60-day intervals for patients suffering from severe CHF. An elevated HGI functions as a protective factor, enhancing the survival outlook for individuals with CHF. These findings could be instrumental in developing management strategies for CHF patients.

In our evaluation rooted in the RCS curve analysis, a ‘U’-shaped connection has been identified between the HGI and the risk ratio for outcome events, with inflection points nearing zero, aligning with earlier research (13, 25). Within a specific cohort, HbA1c levels exceeding 10% demonstrate a progressively non-linear association with HF development, while levels falling below 6% have been linked to a heightened risk of HF (1). Currie et al. previously undertook a retrospective analysis of T2DM patients via the UK General Practice Research Database (GPRD) (26, 27) revealing a U-shaped correlation between average HbA1c levels and overall mortality as well as cardiac incidents. A recent investigation has indicated that patients exhibiting low HbA1c alongside high glycemic variability show poorer left ventricular diastolic function than those with elevated HbA1c but low glycemic variability (28). Inadequate or subpar blood glucose management can lead to microvascular changes and cardiac fibrosis, increasing the HF risk (29). These observations imply that both elevated and diminished HbA1c values can forecast HF progression. Consequently, it is essential in clinical practice to develop a linear regression model to compute the HGI for a substantial patient population.

Previous studies have identified several risk factors for the development of HF in T2DM, including increasing age, body mass index (BMI), SBP, baseline creatinine levels, and a history of coronary artery disease (CAD) (30, 31). Our univariate analysis indicates that age, race, MI, cerebrovascular disease, DM, paraplegia, renal disease, CCI, SBP, respiratory rate, WBC, RBC, and creatinine are risk factors for heart rate (HR). Even after multivariate adjusting the above risk factors, Medium to high level HGI was still independently related to mortality in CHF. Additionally, our cox regression analysis suggest that SBP is one of the risk factors for HF patients. Specifically, maintaining lower SBP levels is significantly linked to a reduced risk of acute MI and stroke in this population (32). Moreover, SBP has been correlated with a reduced risk of cardiovascular events and mortality (33). The Systolic Blood Pressure Intervention Trial (SPRINT) provided further evidence, illustrating that targeting SBP levels below the standard guidelines for non-diabetic individuals can lead to a decreased risk of cardiovascular outcomes and death (34). Lebeche et al. indicates that elevated HGI levels correspond with a greater prevalence of HF in hypertensive patients (35). Consequently, HGI may represent an independent risk factor for individuals experiencing severe CHF.

Our analysis of subgroups indicates that patients with HF who also have diabetes may experience a more pronounced protective effect from higher levels of the HGI compared to HF patients without diabetes. In individuals with HF, the presence of DM has been shown to independently correlate with increased overall mortality, cardiovascular deaths, and a higher frequency of hospital readmissions (36, 37). Most evidence suggests that hyperglycemia, regardless of the treatment administered, can alter the quantitative relationship between HbA1c and blood glucose levels, resulting in a significant increase in HGI levels (38–40). Consequently, diabetic patients are more likely to exhibit elevated HGI levels and may benefit from improved glycemic control interventions. This observation suggests that the HGI could hold greater prognostic significance for patients suffering from advanced heart failure who do not have diabetes.

The study presents several limitations that warrant acknowledgment. Firstly, BMI and other factors have not been measured or may have been inadequately assessed, which could still affect the associations we observed. Secondly, we did not implement a time-dependent model to analyze the use of antihyperglycemic medications. As a result, we are unable to ascertain whether the relationship between these medications and the progression of heart failure (HF) is influenced by factors. Lastly, while the study included data on all-cause mortality, we did not collect information specifically regarding cardiovascular mortality or hospitalizations due to heart failure. This gap highlights the need for further investigation to uncover more nuanced insights into the causes of mortality within this patient demographic.

In conclusion, by utilizing patient data retrieved from the MIMIC-IV database, the authors were able to demonstrate a nonlinear relationship HGI and all-cause mortality in patients with CHF. A higher HGI has a more protective effect than a low HGI for patients with CHF and is closely related to both short-term mortality rates. We believe that HGI is a good prognostic indicator for poor outcomes in CHF patients and could serve as a potential marker for stratifying the risk of short-term and long-term mortality in this patient population. These findings may aid in the management of CHF patients and their blood sugar levels and directing limited healthcare resources towards those who require more intensive management.