Introduction

Chronic heart failure (CHF) is a multifactorial clinical syndrome characterized by high incidence and mortality rates. It is recognized globally as an epidemic affecting approximately 64.3 million people.1 Diabetes mellitus (DM), a common and frequently occurring disease, is more prevalent among patients with heart failure (HF). The Framingham study indicated a 2.4- to 5-fold increase in HF incidence among patients with DM, while other studies have shown that the prevalence of DM among HF patients is 2–2.5 times higher than in the general population.2,3 The prognosis of HF concomitant with DM is worse than that of either HF or DM alone.4 Therefore, effective risk stratification of patients with HF and DM is essential for optimal management and enhancement of prognosis in this population.

Insulin resistance (IR) has a significant effect on patients with either HF or DM and is closely associated with the development and progression of both diseases.5,6 IR can affect cardiac function through various mechanisms, such as metabolic disruption and nerve damage.7–9 Therefore, IR is an important factor that should be considered in the prognosis of patients with HF and DM. Additionally, HF and DM are both wasting disorders, which commonly lead to complications such as malnutrition. Poor nutritional status may cause muscle wasting, resulting in reduced mobility and worsening clinical outcomes,10 while further aggravating the prognosis of related wasting conditions.11,12 Importantly, IR and malnutrition can also interact with each other,13–15 so it is essential to consider the combined effect of IR and nutritional status on the prognosis of patients with HF and DM.

The triglyceride–glucose index (TyG index) is a simple and reliable surrogate marker for IR, exhibiting a strong correlation with IR as confirmed by hyperinsulinemic–euglycemic clamping experiments.16 Previous studies have confirmed that the TyG index is closely associated with various cardiovascular diseases (CVDs).17–19 The prognostic nutritional index (PNI) is an easily obtainable and noninvasive biomarker that reflects nutritional and immune status. Higher PNI values are commonly associated with better disease prognoses, and PNI has been widely used in various research areas, including prognostic studies in cancer, chronic kidney disease, perioperative outcomes in surgery, and CVDs.20–25

Both IR and nutritional status have significant impacts on the prognosis of chronic cardiovascular diseases. However, existing studies predominantly focus on individual risk factors, such as the TyG index or PNI, with limited research exploring their combined effect. The combined use of the TyG index and PNI has the potential to provide more comprehensive prognostic information, which could lead to more accurate risk stratification.

To the best of our knowledge, this study is the first to investigate the combined effect of the TyG index and PNI on the prognosis of patients with CHF and type 2 DM. Based on these two indicators, we developed a novel and composite index, namely the insulin resistance–nutritional index (IRNI), which may comprehensively reflect both metabolic and nutritional status. In addition to investigating the prognostic prediction potential of this index, we compared its incremental predictive value with those of its constituent factors.

Methods Study Design and Population

This was a retrospective, multicenter cohort study of patients with CHF and type 2 DM who were hospitalized at The First Affiliated Hospital of Henan University of Science and Technology from July 1, 2016, to June 30, 2021, and at The Third Affiliated Hospital from July 1, 2016, to June 30, 2020. CHF was defined in accordance with the 2021 ESC Guidelines for the Diagnosis and Treatment of Acute and Chronic Heart Failure.26 A clinical diagnosis of diabetes was confirmed based on one or more of the following criteria: a prior diagnosis of diabetes, fasting blood glucose (FBG) ≥ 7.0 mmol/L, random blood glucose ≥ 11.1 mmol/L, or the use of hypoglycemic agents. From the initial cohort of 4,313 patients with CHF and DM, 998 were excluded based on the following exclusion criteria: (1) age < 18 years; (2) pregnancy or a diagnosis of type 1 DM or other specific types of diabetes; (3) severe hepatic or renal dysfunction, defined as cirrhosis with ascites or chronic renal failure with dialysis treatment; (4) advanced cancer or connective tissue diseases; (5) missing data on lymphocyte count, albumin, FBG, or triglycerides (TG); and (6) in-hospital mortality or loss to follow-up. After applying these exclusion criteria, particularly criterion (2), the remaining cohort was considered to have CHF and type 2 DM, with 3,315 patients included in the final analysis. The patients were categorized into the following four groups based on their TyG index and PNI levels: (a) Group 1: non-high TyG index and high PNI (n = 660); (b) Group 2: non-high TyG index and non-high PNI (n = 1,549); (c) Group 3: high TyG index and high PNI (n = 445); and (d) Group 4: high TyG index and non-high PNI (n = 661) (Figure 1). The cutoff points for high levels of the TyG index and PNI were based on the 66 percentile values.

Figure 1 Flow diagram of patients selection. CTD connective tissue diseases, FBG fasting blood glucose, TG triglycerides, CV death cardiovascular death, TyG index triglyceride–glucose index, PNI prognostic nutritional index.

Ethics Statement

This study adhered to the principles outlined in the Declaration of Helsinki and received approval from the ethics committee of the First Affiliated Hospital of Henan University of Science and Technology (2023–03-K0026). Given its retrospective nature, the institutional review board exempted this study from the requirement for informed consent and guaranteed that all patient-related information was anonymized.

Data Collection and Definitions

Data on age, sex, comorbidities, laboratory values (eg, albumin, lymphocyte count, and FBG), and medications were extracted from an electronic medical records system. Body mass index (BMI) was calculated as body weight (kg) divided by the square of body height (m2). Blood pressure was primarily measured using the cuff inflation technique with either electronic sphygmomanometers or mercury sphygmomanometers.27 The mean arterial pressure (MAP) was calculated using the following formula: (systolic blood pressure + 2 × diastolic blood pressure)/3. The TyG index was calculated as ln (fasting TG [mg/dL] × FBG [mg/dL]/2).16 The PNI was calculated as follows: albumin (g/L) + 5 × total lymphocyte counts (109/L).20 Additionally, we constructed a novel indicator, IRNI, defined as TyG index/Ln (PNI).

Follow-Up and Outcomes

The primary end points assessed in this study included all-cause mortality and cardiovascular death. Prognostic information was gathered via telephone interviews or through a review of pertinent electronic medical records over a median follow-up duration of 4.1 years, with an interquartile range (IQR) of 2.6–5.5 years.

Statistical Analysis

The characteristics of the participants were compared across the four groups based on the TyG index and PNI levels. Continuous variables were expressed as mean ± standard deviation (SD), or median with IQR, depending on their normality of distribution. For continuous data, one-way ANOVA was used if data were normally distributed, while the Kruskal–Wallis test was used if data did not conform to normal distribution. Categorical variables were displayed as frequencies and percentages, and differences among groups were assessed using the chi-square test or Fisher’s exact test, as appropriate.

The Kaplan‒Meier analysis was used to estimate the cumulative incidence of the primary end points, and the Log rank test was applied to assess differences between the groups. Univariate and multivariate Cox proportional-hazards models were used to investigate the relationship between IRNI and the incidence of primary outcomes. IRNI was examined both as a categorical variable, using the lowest tertile as the reference, and as a continuous variable, per SD increase. The proportional-hazards assumption was evaluated using Schoenfeld residuals, which indicated no violations. To address missing covariates, multiple imputations with chained equations were used. Propensity score matching (PSM) was used to adjust for covariates, ensuring comparability among the groups when analyzing the baseline characteristics. A restricted cubic spline (RCS) regression model with three knots was applied to explore the relationship between IRNI and the hazard ratio (HR). We conducted exploratory analyses across various subgroups, including different comorbidities and diabetes severity levels (assessed by the adapted Diabetes Complications Severity Index, aDCSI),28 using the likelihood ratio test to evaluate interactions among the subgroups. Finally, the predictive additive value of IRNI for risk stratification was evaluated using the C-statistic, net reclassification index (NRI), integrated discrimination improvement (IDI), and decision curve analysis (DCA) and was compared with those of the TyG index and PNI alone.

Statistical analyses were conducted using R software (version 4.4.0, R Foundation for Statistical Computing, Vienna, Austria). A two-tailed P value lower than 0.05 was considered statistically significant.

Results Characteristics of the Participants

A total of 3,315 participants were included in the analysis. Their average age was 66.8 years, and 60.6% were male. General characteristics of the study population across different groups are presented in Table 1. Among the individuals with both risk factors (Group 4), the proportion of men was significantly reduced, whereas the proportion of insulin use was substantially elevated in comparison to the other groups (both P < 0.05). The patients in the high-nutritional-score groups (Groups 1 and 3) were younger and exhibited higher BMI, albumin, alanine aminotransferase, platelets, lymphocyte counts, estimated glomerular filtration rate, and sodium levels than those in the groups with lower nutritional scores (Groups 2 and 4). Conversely, they exhibited lower levels of creatinine and NT-proBNP, along with a reduced incidence of chronic kidney disease (CKD) (all P < 0.05).

Table 1 Baseline Characteristics of the Study Population Across Four Groups

The patients in the high-IR-score groups (Groups 3 and 4) exhibited higher MAP, FBG, total cholesterol, TG, and low-density lipoprotein cholesterol levels, along with lower high-density lipoprotein cholesterol values, than those in the groups with lower IR scores (Groups 1 and 2). Moreover, these groups had a higher incidence of myocardial infarction and an increased usage of antiplatelet agents, beta-blockers, and statins (all P < 0.05). Additionally, the proportions of elevated NYHA classification levels and diuretics usage were significantly lower in the group without any risk factors (Group 1) compared with other groups (both P < 0.05).

Combined Effect of the TyG Index and PNI Levels on Mortality

During the follow-up period, the incidence rate of all-cause mortality was 93.3 per 1000 person-years, and the cardiovascular death rate was 58.02 per 1000 person-years. The cumulative incidence rate of both all-cause and cardiovascular death increased with higher tertiles of the TyG index or lower tertiles of PNI (Figure S1, Log rank test, all P < 0.001). Table 2 presents the three Cox regression models used to assess the associations between the combined effect of the TyG index and PNI and the outcomes. In the multivariate model, compared with Group 1, the remaining three groups demonstrated significantly increased risks of all-cause mortality, particularly pronounced in Group 4 (HR, 1.52 [95% CI, 1.27–1.81] for Group 2; 1.55 [95% CI, 1.24–1.95] for Group 3; and 1.91 [95% CI, 1.57–2.32] for Group 4). When using cardiovascular death as the outcome, the results remained consistent (all P < 0.001) (Table 2).

Table 2 Associations Between the Combined Effect of TyG Index and PNI Levels with the Risk of Primary Outcomes in Patients with CHF and Type 2 DM

Correlations Between IRNI and Adverse End Points

The incidence rates of primary events from the lowest to the highest IRNI tertiles were 63.99, 90.9, and 130.14 per 1000 person-years for all-cause death and 35.43, 55.87, and 86.76 per 1000 person-years for cardiovascular death. The Kaplan–Meier analysis demonstrated significant differences in the risk of primary end points across the three IRNI tertiles (Figure 2, Log rank test, both P < 0.001). The RCS regression model also revealed that higher levels of IRNI were associated with an increased risk of all-cause death (model 2: HR per SD increase, 1.31; 95% CI, 1.23–1.40) and cardiovascular death (model 2: HR per SD increase, 1.43; 95% CI, 1.32–1.55) (both P for overall association < 0.001) (Figure S2).

Figure 2 Kaplan-Meier estimation of (a) all-cause death and (b) CV death by tertiles of IRNI. CV death cardiovascular death, IRNI insulin resistance–nutritional index.

The multivariate Cox regression analysis results demonstrated a significant association between IRNI and all-cause mortality (T1 vs T2: HR, 1.42 [95% CI, 1.22–1.66]; T3: HR, 1.93 [95% CI, 1.66–2.26]; all P < 0.001). When cardiovascular death was considered as the end point, the results remained consistent. Namely, the highest IRNI tertile was significantly associated with an increased incidence of CV mortality (model 2: HR, 2.50; 95% CI, 2.05–3.06; P < 0.001, comparing T1 vs T3) (Table 3). To evaluate the stability of the predictive efficacy of IRNI, PSM was used to adjust for primary confounding variables across the three groups (Table S1). The results indicated that even after PSM, IRNI continued to exhibit a strong association with adverse outcomes, with HRs of 1.89 (95% CI, 1.58–2.25, P < 0.001) for all-cause death and 2.58 (95% CI, 2.05–3.25, P < 0.001) for cardiovascular death, comparing T1 vs T3 (Table S2).

Table 3 Univariable and Multivariable Cox Analyses of IRNI Predicting Primary Outcomes in Patients with CHF and Type 2 DM

Implications of IRNI on Adverse Outcomes Across Different Subgroups

We further conducted exploratory analyses of the prognostic efficacy of IRNI across various subgroups. The Kaplan–Meier analysis demonstrated that the risk of primary end points (including all-cause death and cardiovascular death) varied significantly among the three IRNI tertiles across different subgroups, including the subgroups defined by age, sex, hypertension, myocardial infarction, CKD, and obesity (Figures S3 and S4, Log rank test, all P < 0.001). The results of the multivariate Cox proportional-hazards models, examining the relationships between IRNI and all-cause mortality across various subgroups, are presented in Figure 3. The findings consistently revealed that IRNI was significantly associated with all-cause mortality across diverse subgroups (all P for trend < 0.05). The associations between IRNI and cardiovascular mortality also displayed consistent results across the subgroups (Table S3, all P for trend < 0.05).

Figure 3 Forest plot of all-cause death according to tertiles of IRNI across different subgroups adjusted for model 2.

Abbreviations: HR, hazard ratio; CI, confidence interval; IRNI, insulin resistance–nutritional index; CKD, chronic kidney disease.

We further explored whether diabetes severity, measured by aDCSI, affects the predictive ability of IRNI for survival outcomes. A higher score on aDCSI indicates a greater number of diabetic complications and more severe disease.28 The participants were categorized into three groups based on their aDCSI score levels, namely ≤ 3, 4–5, and ≥ 6. We found that IRNI consistently maintained a strong association with adverse outcomes across all levels of diabetes severity (Table S4, all P for trend < 0.05), but there was an increasingly pronounced correlation between IRNI and all-cause mortality in the groups with higher aDCSI scores. The HRs increased across the groups from low to high scoring 1.54 (95% CI, 1.20–1.99), 2.01 (95% CI, 1.56–2.58), and 2.45 (95% CI, 1.78–3.39) in T1 vs T3 comparisons, with a significant interaction (P = 0.028). The association between IRNI and cardiovascular death demonstrated a consistent trend across varying aDCSI levels, with HRs of 1.74 (95% CI, 1.27–2.40), 2.63 (95% CI, 1.91–3.64), and 3.25 (95% CI, 2.09–5.05) with T1 vs T3 (Figure 4, P for interaction = 0.010).

Figure 4 Forest plot of primary outcomes according to tertiles of IRNI stratified by different aDCSI scores adjusted for model 2.

Abbreviations: HR hazard ratio; CI, confidence interval; IRNI, insulin resistance–nutritional index; CV death, cardiovascular death; aDCSI, adapted Diabetes Complications Severity Index.

Evaluation of the Incremental Predictive Value of IRNI on Risk Stratification

The incremental predictive value of IRNI for mortality is shown in Table 4. The addition of IRNI to the baseline risk model improved risk prediction, with the C-statistic increasing from 0.656 to 0.679 (P < 0.01). Analysis of NRI and IDI revealed statistically significant improvements in predictive accuracy; namely, NRI was 0.287 (95% CI, 0.211–0.362; P < 0.01), and IDI was 0.016 (95% CI, 0.012–0.021; P < 0.01). Next, we evaluated the predictive capabilities of IRNI in comparison with those of the TyG index and PNI alone. We found that IRNI consistently exhibited superior predictive performance, in terms of both numerical increase and statistical significance (Table 4, Figure 5, both P for difference < 0.05). DCA further demonstrated that adding IRNI to the baseline model markedly enhanced the net benefit, exceeding that provided by TyG or PNI alone (Figure S5).

Table 4 Evaluation of the Incremental Effect of Adding IRNI, TyG Index and PNI to the Baseline Risk Model for Predicting 3-Year Mortality

Figure 5 ROC curves evaluating the incremental effect of IRNI, TyG index and PNI beyond the baseline risk model. ROC curve receiver operator characteristic curve, IRNI insulin resistance–nutritional index, TyG index triglyceride–glucose index, PNI prognostic nutritional index. The baseline risk model included age, gender, NYHA classification, LVEF, NT-proBNP, creatinine, hypertension, atrial fibrillation, previous MI, COPD, ACEI/ARB/ARNI, β-blocker and SGLT2 inhibitors.

Discussion

This study is the first to explore the combined effect of the TyG index and PNI on the prognosis of patients with CHF and type 2 DM. The key findings of our study are as follows: (1) The combined effect of the TyG index and PNI on prognosis was significantly greater than the effect of either factor alone, with the highest risk of adverse outcomes observed in the population exhibiting higher TyG levels in conjunction with lower PNI levels. (2) Based on the TyG index and PNI, we developed a novel composite indicator, IRNI, and validated its effective and robust prognostic value. (3) IRNI provided a significantly enhanced prognostic incremental value and was superior to its basic constituent indicators, including the TyG index and PNI. In summary, our study demonstrates that IRNI can serve as a novel and valuable prognostic factor for patients with CHF and type 2 DM.

IR and nutritional status are significant factors affecting the prognosis of patients with chronic diseases such as CHF and DM. Chronic wasting diseases such as CHF or DM are closely associated with IR and nutritional imbalances.6,29–32 These issues are particularly pronounced in populations suffering from comorbidities of both conditions. The TyG index, based on FBG and TG, is a simple, easily obtainable, and reliable surrogate marker for IR. Previous studies have demonstrated that the TyG index has high sensitivity (96.5%) and specificity (85.0%) for identifying IR compared with the hyperinsulinemic–euglycemic clamp technique, which is the gold standard for measuring IR.16 Currently, the TyG index has been widely used in various research fields, such as cancer, cerebrovascular diseases, CKD, and CVDs. A large-scale study involving 155,167 cancer patients demonstrated a graded positive association between the TyG index and CVD hospitalization; namely, for each unit increase in the TyG index, the hospitalization rates for CVD and acute myocardial infarction increased by 16% and 45%, respectively.33 Chen et al used the Medical Information Mart for Intensive Care IV (Version 2.2) repository to include 1,537 patients with HF and CKD. Kaplan–Meier survival analysis showed that the group with a higher TyG index had significantly lower survival rates than the group with a lower TyG index (Log rank test: P < 0.001). Additionally, there was a nonlinear association between the TyG index and all-cause mortality.34

The PNI, which quantifies serum albumin and lymphocyte counts through simple calculations to assess systemic nutritional status and immunocompetence, has garnered widespread attention.35 This index was first introduced by Onodera in 1984 and was used as a comprehensive nutritional indicator for assessing the surgical risk in patients with gastrointestinal tumors.36 It has attracted increasing attention. Initially applied extensively in cancer research, PNI has progressively been used in various non-oncological fields, including trauma, diabetes, and cardiomyopathy, demonstrating promising prospects for broader applications. In a previous study, researchers included 3,351 patients aged 45 years and older who underwent hip fracture surgery.37 The patients were divided into low, medium, and high groups based on their PNI levels. The results showed that the patients in the high-PNI group had significantly fewer postoperative complications (odds ratio, 0.61; 95% CI, 0.40–0.93) and lower mortality rates (HR, 0.61; 95% CI, 0.42–0.88) compared to those in the low-PNI group. Ning et al included a total of 5,916 adult patients with DM, with an average follow-up period of 8 years. After grouping based on the quartile levels of PNI, it was found that with increasing PNI levels (ie, Q2, Q3, and Q4), all-cause mortality decreased by 24%, 38%, and 28%, respectively, and cardiovascular mortality decreased by 30%, 27%, and 26%, respectively.38 Another study analyzed the effect of PNI in patients with hypertrophic cardiomyopathy, revealing that a higher PNI was significantly associated with improved survival outcomes. After adjusting for potential confounders, PNI was independently associated with both all-cause mortality and cardiovascular mortality, with HRs per SD of 0.46 (95% CI, 0.34–0.62) and 0.44 (95% CI, 0.30–0.63), respectively.39

Although numerous studies across diverse fields have shown a correlation between the TyG index and poor prognosis and an inverse correlation between PNI and adverse outcomes, research on the joint application of these two predictive markers for disease risk stratification is still limited. A previous study combined the TyG index and PNI, among other indicators, to develop a nomogram predicting contrast-induced nephropathy after percutaneous coronary intervention in patients with type 2 DM and acute coronary syndrome.40 The results demonstrated high discriminative capability, with the area under the curve reaching 0.785 (95% CI, 0.729–0.841) in the training cohorts and 0.802 (95% CI, 0.699–0.905) in the validation cohorts. That study highlighted the potential of the combined use of the TyG index and PNI for stratifying disease risk. However, this combined approach has not yet been applied to other diseases.

Our study showed that in patients with CHF and type 2 DM, the combined effect of the TyG index and PNI on prognosis was significantly superior to that of either single indicator alone, suggesting that their combined use might produce a synergistic enhancement of the effect. Most importantly, to facilitate clinical use and encourage broader adoption, we developed a new composite indicator, IRNI, derived from these two indicators. This new indicator demonstrated a strong correlation with adverse outcomes and exhibited significant predictive value. We propose that the underlying mechanisms can be explained through the following aspects: First, IRNI positively correlates with IR. Numerous studies have confirmed that IR is not only associated with a high incidence of HF but also closely related to poor prognosis.5,41 IR can influence cardiac function in multiple ways, including disruption of myocardial energy metabolism, damage to cardiac sympathetic nerves, endothelial dysfunction, oxidative stress, and inflammatory responses.7–9 Second, IRNI positively correlates with poor nutritional status. The incidence of malnutrition in HF patients is notably high at approximately 46%, and compared with patients without malnutrition, those with nutritional deficiencies have an increased risk of all-cause mortality (HR, 2.15; 95% CI, 1.89–2.45).29 Additionally, poor nutritional status may often be accompanied by significant muscle wasting, which mainly manifests as reduced mobility and weakened muscle strength, severely affecting clinical outcomes.10 Third, IRNI may reflect both IR and nutritional status. According to previous studies, IR can trigger malnutrition by affecting protein-energy metabolism.13 Moreover, malnutrition may indirectly affect insulin sensitivity through various mechanisms,15 including altering the function and number of immune cells, such as T cells. This can ultimately lead to or exacerbate IR. Thus, there is a mutual influence, even a vicious cycle, between IR and malnutrition. Therefore, considering these factors together can provide more comprehensive prognostic information.

In our subgroup analysis, IRNI consistently maintained robust predictive efficacy across different subgroups. However, its ability to identify high-risk individuals was particularly pronounced in populations with higher severity of diabetes. The enhanced predictive capability may be attributed to the fact that individuals with more severe diabetes complications may often exhibit higher degrees of IR and poorer nutritional status.42 This finding reminds us that in clinical practice, the management of such high-risk populations should place greater emphasis on simultaneously addressing both IR and malnutrition. Finally, IRNI significantly enhances risk stratification to a greater extent than either of its components alone. Not only does this finding support the application of this novel index in our study population, but it also suggests its potential utility in other chronic diseases, pending further validation.

Strengths and Limitations

Our study has several advantages. First, the use of cohort data from two centers enhanced representativeness. Second, it was the first to explore the combined effect of the TyG index and PNI on the prognosis of patients with CHF and type 2 DM. Third, we innovatively constructed a new index, IRNI, and validated its effectiveness and robustness. Fourth, the analysis not only included a wide range of baseline characteristics to adjust for confounding factors but also used PSM analysis to enhance comparability between groups.

It is also important to acknowledge several limitations to the study. First, given the retrospective nature of the study, comprehensive baseline data for patients outside of the hospital setting could not be obtained. Second, the absence of insulin measurement data and other anthropometric indicators such as arm and hip circumference precluded the ability to perform a multidimensional comparison. Third, the follow-up results may be subject to varying degrees of recall or reporting bias. Fourth, this retrospective observational study may contain unmeasured confounding factors, and therefore, interpretations of the results should be approached with caution while also recognizing that causal relationships cannot be definitively established. Fifth, due to the limitations of the retrospective study design, the specific brands of the measurement instruments used were not available. Finally, although our study developed the new indicator IRNI, its effectiveness and universality, especially regarding the potential impact of population or ethnic differences, still need to be confirmed through other cohort studies and prospective research.

Conclusions

In patients with CHF and type 2 DM, the combination of the TyG index and PNI enhances the ability to predict adverse risks. Furthermore, our newly developed index, IRNI, demonstrates robust and effective predictive potential, surpassing the performance of its individual components. Not only does IRNI enhance the precision in assessing and managing disease risks, but it also warrants further investigation for its potential application in various other chronic or wasting disease populations.

Data Sharing Statement

The datasets utilized and analyzed in this study are available from the corresponding author upon reasonable request.

Ethics

This study adhered to the principles outlined in the Declaration of Helsinki and received approval from the ethics committee of the First Affiliated Hospital of Henan University of Science and Technology. Due to its retrospective nature, the institutional review board exempted this study from the requirement for informed consent and guaranteed that all patient-related information was anonymized.

Acknowledgments

There is no acknowledgement.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Funding

This work was supported by the National Engineering Research Center of Medical Big Data of Chinese PLA General Hospital, the Medical Engineering Laboratory of Chinese PLA General Hospital, and the Independent Research Project of Medical Engineering Laboratory of Chinese PLA General Hospital (2022SYSZZKY07).

Disclosure

The authors declare no conflicts of interest in this work.

References

1. James SL, Abate D, Abate KH, et al. Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: a systematic analysis for the global burden of disease study 2017. Lancet. 2018;392(10159):1789–1858.

2. Kannel WB, Hjortland M, Castelli WP. Role of diabetes in congestive heart failure: the Framingham study. Am J Cardiol. 1974;34(1):29–34. doi:10.1016/0002-9149(74)90089-7

3. Park JJ. Epidemiology, pathophysiology, diagnosis and treatment of heart failure in diabetes. Diabet Metabol J. 2021;45(2):146–157. doi:10.4093/dmj.2020.0282

4. Siao WZ, Chen YH, Tsai CF, Lee CM, Jong GP. Diabetes mellitus and heart failure. J Pers Med. 2022;12(10):1698. doi:10.3390/jpm12101698

5. Vardeny O, Gupta DK, Claggett B, et al. Insulin resistance and incident heart failure the ARIC study (atherosclerosis risk in communities). JACC Heart Fail. 2013;1(6):531–536. doi:10.1016/j.jchf.2013.07.006

6. Abdul-Ghani M, Maffei P, DeFronzo RA. Managing insulin resistance: the forgotten pathophysiological component of type 2 diabetes. Lancet Diabetes Endocrinol. 2024;12(9):674–680. doi:10.1016/S2213-8587(24)00127-X

7. Guo CA, Guo S. Insulin receptor substrate signaling controls cardiac energy metabolism and heart failure. J Endocrinol. 2017;233(3):R131–r143. doi:10.1530/JOE-16-0679

8. Paolillo S, Rengo G, Pellegrino T, et al. Insulin resistance is associated with impaired cardiac sympathetic innervation in patients with heart failure. Eur Heart J Cardiovasc Imaging. 2015;16(10):1148–1153. doi:10.1093/ehjci/jev061

9. Ormazabal V, Nair S, Elfeky O, Aguayo C, Salomon C, Zuniga FA. Association between insulin resistance and the development of cardiovascular disease. Cardiovasc Diabetol. 2018;17(1):122. doi:10.1186/s12933-018-0762-4

10. Fulster S, Tacke M, Sandek A, et al. Muscle wasting in patients with chronic heart failure: results from the studies investigating co-morbidities aggravating heart failure (SICA-HF). Eur Heart J. 2013;34(7):512–519. doi:10.1093/eurheartj/ehs381

11. Tirandi A, Montecucco F, Liberale L. Malnutrition predicts mortality in heart failure patients. Intern Emerg Med. 2023;18(4):979–980. doi:10.1007/s11739-023-03305-9

12. Saitoh M, Dos Santos MR, Ebner N, et al. Nutritional status and its effects on muscle wasting in patients with chronic heart failure: insights from studies investigating co-morbidities aggravating heart failure. Wien Klin Wochenschr. 2016;128(Suppl 7):497–504. doi:10.1007/s00508-016-1112-8

13. Siew ED, Ikizler TA. Insulin resistance and protein energy metabolism in patients with advanced chronic kidney disease. Semin Dial. 2010;23(4):378–382. doi:10.1111/j.1525-139X.2010.00763.x

14. Rao RH, Menon RK. Chronic malnutrition impairs insulin sensitivity through both receptor and postreceptor defects in rats with mild streptozocin diabetes. Metabolism. 1993;42(6):772–779. doi:10.1016/0026-0495(93)90248-M

15. Gerriets VA, MacIver NJ. Role of T cells in malnutrition and obesity. Front Immunol. 2014;5:379. doi:10.3389/fimmu.2014.00379

16. Guerrero-Romero F, Simental-Mendia LE, Gonzalez-Ortiz M, et al. The product of triglycerides and glucose, a simple measure of insulin sensitivity. Comparison with the euglycemic-hyperinsulinemic clamp. J Clin Endocrinol Metab. 2010;95(7):3347–3351. doi:10.1210/jc.2010-0288

17. Barzegar N, Tohidi M, Hasheminia M, Azizi F, Hadaegh F. The impact of triglyceride-glucose index on incident cardiovascular events during 16 years of follow-up: Tehran lipid and glucose study. Cardiovasc Diabetol. 2020;19(1):155. doi:10.1186/s12933-020-01121-5

18. Tao LC, Xu JN, Wang TT, Hua F, Li JJ. Triglyceride-glucose index as a marker in cardiovascular diseases: landscape and limitations. Cardiovasc Diabetol. 2022;21(1):68. doi:10.1186/s12933-022-01511-x

19. Kurniawan LB. Triglyceride-glucose index as a biomarker of insulin resistance, diabetes mellitus, metabolic syndrome, and cardiovascular disease: a review. Eur J Int Fed Clin Chem Lab Med. 2024;35(1):44–51.

20. Du Z, Sun H, Zhao R, et al. Combined with prognostic nutritional index and IgM for predicting the clinical outcomes of gastric cancer patients who received surgery. Front Oncol. 2023;13:1113428. doi:10.3389/fonc.2023.1113428

21. Zhao L, Zhao X, Tian P, et al. Prognostic utility of the prognostic nutritional index combined with serum sodium level in patients with heart failure. Nutr Metab Cardiovasc Dis. 2022;32(8):1894–1902. doi:10.1016/j.numecd.2022.04.004

22. Peng P, Chen L, Shen Q, Xu Z, Ding X. Prognostic nutritional index (PNI) and controlling nutritional status (CONUT) score for predicting outcomes of breast cancer: a systematic review and meta-analysis. Pak J Med Sci. 2023;39(5). doi:10.12669/pjms.39.5.7781

23. Yu J-H, Chen Y, Yin M-G. Association between the prognostic nutritional index (PNI) and all-cause mortality in patients with chronic kidney disease. Renal Failure. 2023;45(2). doi:10.1080/0886022X.2023.2264393

24. Liu M, Sun M, Zhang T, et al. Prognostic nutritional index (PNI) as a potential predictor and intervention target for perioperative ischemic stroke: a retrospective cohort study. BMC Anesthesiol. 2023;23(1). doi:10.1186/s12871-023-02216-8

25. Zhang X, Zhang J, Liu F, et al. Prognostic nutritional index (PNI) as a predictor in patients with metabolic syndrome and heart failure. Diabetes Metab Syndr Obes. 2023;16:2503–2514. doi:10.2147/DMSO.S420924

26. McDonagh TA, Metra M, Adamo M, et al. 2021 ESC guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J. 2021;42(36):3599–3726. doi:10.1093/eurheartj/ehab368

27. Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on clinical practice guidelines. Circulation. 2018;138(17):e426–e483. doi:10.1161/CIR.0000000000000597

28. Chang H-Y, Weiner JP, Richards TM, Bleich SN, Segal JB. Validating the adapted diabetes complications severity index in claims data. Am J Manag Care. 2012;18(11):721–726.

29. Lv S, Ru S. The prevalence of malnutrition and its effects on the all-cause mortality among patients with heart failure: a systematic review and meta-analysis. PLoS One. 2021;16(10):e0259300. doi:10.1371/journal.pone.0259300

30. Lima L. Insulin resistance underlying type 2 diabetes. Lancet Diabetes Endocrinol. 2019;7(6):424. doi:10.1016/S2213-8587(19)30147-0

31. Aroor AR, Mandavia CH, Sowers JR. Insulin resistance and heart failure: molecular mechanisms. Heart Failure Clin. 2012;8(4):609–617. doi:10.1016/j.hfc.2012.06.005

32. Sanz Paris A, Garcia JM, Gomez-Candela C, et al. Malnutrition prevalence in hospitalized elderly diabetic patients. Nutr Hosp. 2013;28(3):592–599. doi:10.3305/nh.2013.28.3.6472

33. Jung MH, Yi SW, An SJ, et al. Associations between the triglyceride-glucose index and cardiovascular disease in over 150,000 cancer survivors: a population-based cohort study. Cardiovasc Diabetol. 2022;21(1):52. doi:10.1186/s12933-022-01490-z

34. Chen Y, Li S, Yang K, et al. Triglyceride-glucose index and prognosis in individuals afflicted with heart failure and chronic kidney disease. ESC Heart Fail. 2024;11:3120–3132. doi:10.1002/ehf2.14898

35. Pan Y, Ma Y, Dai G. The prognostic value of the prognostic nutritional index in patients with advanced or metastatic gastric cancer treated with immunotherapy. Nutrients. 2023;15(19):4290. doi:10.3390/nu15194290

36. Onodera T, Goseki N, Kosaki G. Prognostic nutritional index in gastrointestinal surgery of malnourished cancer patients. Nihon Geka Gakkai Zasshi. 1984;85(9):1001–1005.

37. Wang Y, Jiang Y, Luo Y, et al. Prognostic nutritional index with postoperative complications and 2-year mortality in hip fracture patients: an observational cohort study. Int J Surg. 2023;109(11):3395–3406.

38. Ning Y, Pan D, Guo J, et al. Association of prognostic nutritional index with the risk of all-cause mortality and cardiovascular mortality in patients with type 2 diabetes: NHANES 1999–2018. BMJ Open Diabetes Res Care. 2023;11(5):e003564. doi:10.1136/bmjdrc-2023-003564

39. Wang Z, Zhao L, He S. Prognostic nutritional index and the risk of mortality in patients with hypertrophic cardiomyopathy. Int J Cardiol. 2021;331:152–157. doi:10.1016/j.ijcard.2021.01.023

40. Zhu Y, He H, Qiu H, Zhang X, Wang L, Li W. Prognostic nutritional index combined with triglyceride-glucose index to contrast a nomogram for predicting contrast-induced kidney injury in type 2 diabetes mellitus patients with acute coronary syndrome after percutaneous coronary intervention. Clin Interv Aging. 2023;18:1663–1673. doi:10.2147/CIA.S429957

41. Doehner W, Rauchhaus M, Ponikowski P, et al. Impaired insulin sensitivity as an independent risk factor for mortality in patients with stable chronic heart failure. J Am Coll Cardiol. 2005;46(6):1019–1026. doi:10.1016/j.jacc.2005.02.093

42. Zhang SS, Tang ZY, Fang P, Qian HJ, Xu L, Ning G. Nutritional status deteriorates as the severity of diabetic foot ulcers increases and independently associates with prognosis. Exp Ther Med. 2013;5(1):215–222. doi:10.3892/etm.2012.780