Skip to main content
Download PDF
Download ePub

Insulin resistance quantified by estimated glucose disposal rate predicts cardiovascular disease incidence: a nationwide prospective cohort study

Cardiovascular Diabetology volume 24, Article number: 161 (2025) Cite this article

Abstract

Background

Insulin resistance (IR) is an important pathologic component in the occurrence and development of cardiovascular disease (CVD). The estimated glucose disposal rate (eGDR) is a measure of glucose handling capacity, that has demonstrated utility as a reliable marker of IR. The study aimed to determine the predictive utility of IR assessed by eGDR for CVD risk.

Methods

This nationwide prospective cohort study utilized data of 6416 participants from the China Health and Retirement Longitudinal Study (CHARLS) who were free of CVD but had complete data on eGDR at baseline. The Boruta algorithm was performed for feature selection. Multivariate Cox proportional hazards regression models and restricted cubic spline (RCS) analysis were conducted to examine the associations between eGDR and CVD, and the results were expressed with hazard ratio (HR) and 95% confidence interval (CI) values. The area under the receiver operating characteristic (ROC) curve (AUC), calibration curve, Hosmer–Lemeshow test, net reclassification improvement (NRI), and decision curve analysis (DCA) were employed to evaluate the clinical efficacy of eGDR in identifying CVD. Subgroup analysis was performed to explore the potential association of with CVD in different populations.

Results

During a median follow-up of 106.5 months, 1339 (20.87%) incident CVD cases, including 1025 (15.96%) heart disease and 439 (6.84%) stroke, were recorded from CHARLS. The RCS curves demonstrated a significant and linear relationship between eGDR and all endpoints (all P for nonlinear > 0.05). After multivariate adjustment, the lower eGDR levels were found to be significantly associated with a greater prevalence of CVD. Compared to the lowest quartile, the highest eGDR quartile was associated with a decreased risk of CVD (HR 0.686, 95% CI 0.545–0.862). When assessed as a continuous variable, individuals with a unit increasement in eGDR was related to a 21.2% (HR 0.788, 95% CI 0.669–0.929) lower risk of CVD, a 18.3% (HR 0.817, 95% CI 0.678–0.985) decreased risk of heart disease, and 39.5% (HR 0.705, 95% CI 0.539–0.923) lower risk of stroke. The eGDR had an excellent predictive performance according to the results of ROC (AUC = 0.712) and χ2 likelihood ratio test (χ2 = 4.876, P = 0.771). NRI and DCA analysis also suggested the improvement from eGDR to identify prevalent CVD and the favorable clinical efficacy of the multivariate model. Subgroup analysis revealed that the trend in incident CVD risk were broadly consistent with the main results across subgroups.

Conclusion

A lower level of eGDR was found to be associated with increased risk of incident CVD, suggesting that eGDR may serve as a promising and preferable predictor for CVD.

Graphical Abstract

Research insights

AbstractSection What is currently known about this topic?

Insulin resistance (IR) constitutes a critical pathophysiological mechanism that significantly contributes to the pathogenesis and progression of cardiovascular disease (CVD).

AbstractSection What is the key research question?

What is the relationship between IR assessed by estimated glucose disposal rate (eGDR) and the incident CVD risk in China?

AbstractSection What is new?

The study demonstrated a significant association between reduced eGDR levels and elevated CVD risk in the Chinese population.

AbstractSection How might this study influence clinical practice?

Our findings underscore the critical need to expand cardiovascular risk stratification paradigms beyond conventional parameters.

Introduction

Cardiovascular disease (CVD) remains the primary cause of global morbidity and mortality [1]. According to the Global Burden of Disease Study 2021, ischemic heart disease and stroke cause approximately 8.8 and 7.1 million premature deaths, as well as 188.3 and 160.4 million disability-adjusted life years, respectively [2]. China and India collectively accounted for over one-third of global CVD mortality in 2020, constituting a disproportionate share of this public health burden [3]. China’s epidemiological transition, characterized by heterogeneous economic development over three decades, has precipitated marked escalations in modifiable CVD risk determinants through dynamic shifts in socioeconomic stratification, lifestyle patterns, and healthcare accessibility [4]. Notably, age-standardized CVD mortality surged 46% between 1990 and 2013, a trajectory attributable to synergistic interactions between demographic aging and socioeconomic determinants of health [5]. Current estimates indicate approximately 330 million prevalent CVD cases nationwide, representing one of the largest chronic disease cohorts globally [6]. Crucially, despite coordinated efforts to mitigate conventional risk factors, persistent residual risks underscore the need for novel predictive biomarkers and targeted interventions [7, 8].

Insulin resistance (IR) is considered as an independent risk factor for microvascular complications of diabetes mellitus and CVD, and that there is a close association between IR and cardiovascular risk [9, 10]. Conventional approaches for detecting IR mainly include the hyperinsulinemic–euglycemic (HIEG) clamp technique and the homeostasis model assessment for IR (HOMA-IR) [11]. Nevertheless, the technique’s invasiveness or time-consuming render it impractical for regular and extensive measurement in clinical settings. Recently, the estimated glucose disposal rate (eGDR), determined by easily accessible clinical parameters including hemoglobin A1c (HbA1c), hypertension, and waist circumference (WC), has emerged as a simpler surrogate marker for IR [12]. Previous studies have found that eGDR was inversely related to risks of stroke, renal disease, and all-cause mortality, particularly in diabetic individuals [13,14,15]. Yet the relationship between eGDR and incident CVD in the broader population still required to be further determined.

Therefore, to address these concerns, we enrolled participants from the China Health and Retirement Longitudinal Study (CHARLS), a nationwide, population based, prospective cohort study [16], to identify the predictive utility of eGDR for CVD incident in China. The findings of this study may aid in the development of novel predictive techniques as well as provide vital new insights into the role of eGDR in predicting cardiovascular events.

Materials and methods

Study design and participants

The data for the study came from the CHARLS cohort study of Chinese residents aged 45 years and older. The CHARLS covers 150 districts and 450 villages/urban communities, including 10,257 households, and collects a wide range of information on residents’ demographic and social characteristics, health status, socioeconomic status, family structure and social activities. In brief, 17,708 participants aged 45 years or above from 28 provinces in China were recruited by a four-stage stratified cluster sampling, adopting a multistage probability sampling technology in the first wave (W1) between 2011 and 2012 with a more than 80% response rate. These participants underwent regular follow-ups every two years through face-to-face interviews conducted by trained interviewers using computer-assisted guidance. Four subsequent follow-up waves were conducted in 2013, 2015, 2018, and 2020. The 19,395 individuals were recruited with available follow-up data from the fifth wave (W5, 2020). The protocol of CHARLS has been described in detail elsewhere (https://charls.pku.edu.cn/). The study population comprised participants with CVD at Wave 5. Patients who meet one of the following criteria will be excluded: [1] those with pre-existing CVD at baseline; [2] individuals with missing baseline eGDR data; [3] participants demonstrating incomplete baseline data on sociodemographic factors, health behaviors, anthropometrics, or critical biomarkers; and [4] subjects with unavailable CVD outcomes during longitudinal follow-up.

Data collection and definitions

The CHARLS investigators collected variables according to pre-specified standards. Demographic data comprised age, gender, marital status, and blood pressure. Both systolic blood pressure (SBP) and diastolic blood pressure (DBP) were calculated as the average of the three-time measurements taken in a sitting position after resting for five minutes. Body mass, height, and WC also were measured while participants wore lightweight clothes and no shoes. Medical history information included smoking history, drinking history, hypertension, diabetes, CVD, stroke, dyslipidemia, kidney disease, and cancer. Blood samples were collected from CHARLS participants at baseline after an overnight fast by professional staff, stored at − 20 °C, and transported to Beijing, where further measurements were conducted following standard procedures. Biochemical parameters included white blood cell (WBC), platelets (PLT), hemoglobin (Hb), fasting plasma glucose (FPG), blood urea nitrogen (BUN), uric acid (UA), serum creatinine, C-reactive protein (CRP), HbA1c, and lipid profiles such as total cholesterol (TC), triglyceride (TG), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C).

Hypertension was defined as follows: a self-reported hypertension based on physician diagnosis, and/or any use of antihypertensive drugs, and/or SBP/DBP ≥ 140/90 mmHg [17]. Diabetes was defined based on a self-reported physician diagnosis, use of hypoglycemic drugs, or FPG ≥ 126 mg/dL, 2-h plasma glucose ≥ 200 mg/dL, and/or an HbA1c level ≥ 6.5% at baseline [18]. The body mass index (BMI) was calculated as: BMI (kg/m2) = body mass (kg)/height2 (m2). A BMI of ≥ 24 kg/m2 is established as the diagnostic threshold for overweight classification according to standardized anthropometric criteria [19].

Ascertainment of exposure and endpoints

The exposure of this study was eGDR at baseline. The formula for calculating eGDR was as follows: eGDR (mg/kg/min) = 21.158−(0.09×WC)−(3.407×hypertension)−(0.551×HbA1c) [WC (cm), hypertension (yes = 1/no = 0), and HbA1c (%)].

The primary outcomes we are more concerned about are CVD events, including heart disease and stroke. Consistent with established precedents [16], CVD events were assessed by the following standardized questions: “Have you been told by a doctor that you have been diagnosed with a heart attack, coronary heart disease, angina, congestive heart failure, or other heart problems?” or “Have you been told by a doctor that you have been diagnosed with a stroke?” Participants who reported heart disease or stroke were defined as having CVD.

Statistical analysis

All statistical analyses were performed using IBM-SPSS (version 26.0, Chicago, IL, USA) and R (version 4.1.2, Vienna, Austria). A two-sides P value of less than 0.05 was considered to indicate statistical significance. Continuous variables were presented as mean ± standard deviation (SD) or median [interquartile range (IQR)] as appropriate. Comparisons of baseline data for normally and skewed distributed variables were performed using analysis of variance and the Kruskal-Wallis H test, respectively. Categorical variables were expressed as counts and percentages, with differences determined through chi-square tests. We conducted trend tests using the median value of each quartile of eGDR. All participants were stratified into four groups by eGDR quartiles: Q1 (eGDR < 9.14, n = 1600), Q2 (9.14 ≤ eGDR < 10.52, n = 1589), Q3 (10.52 ≤ eGDR < 11.29, n = 1614), and Q4 (eGDR ≥ 11.29, n = 1613).

The Boruta algorithm [20], a robust ensemble feature selection method demonstrating particular efficacy in high-dimensional data environments with nonlinear relationships, was used to determine the most critical features associated with incident CVD. Multivariate Cox proportional hazards regression models were applied to determine the associations of eGDR quartile groups with CVD incidence, and the results were expressed as hazard ratio (HR) and 95% confidence interval (CI) values. The proportional hazards assumption was verified using Schoenfeld residuals, and no significant violations were observed. Three Cox models were applied. Model 1 was a crude model; Model 2 adjusted for age, gender, BMI, SBP, and DBP; Model 3 further for adjusted for age, gender, BMI, SBP, DBP, smoking, diabetes, dyslipidemia, FPG, UA, TC, TG, LDL-C, and HDL-C. To investigate the dose-response relationship between eGDR and CVD incidence, restricted cubic spline (RCS) models with three knots based on the Model 3 were employed, with the eGDR value at HR = 1 treated as the reference.

We used calibration curve and χ2 likelihood ratio test to assess the goodness-of-fit of the model. The area under the receiver operating characteristic (ROC) curve (AUC) was used to assess the predictive value of eGDR on the incidence of CVD. To further estimate additional the predictive power beyond the basic models, the net reclassification improvement (NRI) was also computed. The decision curve analysis (DCA) was plotted to comprehensively evaluate the clinical efficacy of the model. Subgroup analyses were conducted to assess the effects of eGDR (continuous) on the incidence of ASCVD in several subgroups, including age (</≥65 years), gender (male/female), BMI (</≥24 kg/m2), smoking (yes/no), diabetes (yes/no), and dyslipidemia (yes/no).

Results

Baseline characteristics

The longitudinal study cohort comprised 17,708 participants at baseline (W1), with the sample expanding to 19,395 individuals at the fifth follow-up interval (W5). The 13,095 participants who took part in both W1 and W5 data investigation were included in this study. 5748 individuals were excluded because of the missing data including age, gender, BMI, diabetes history, hypertension history, WC, HbA1c, and CVD condition during the follow-up. Finally, 6416 individuals with reported CVD outcomes were recruited in this study (Fig. 1).

The baseline characteristics are presented in Table 1. Among the 6416 participants enrolled, the mean age was 56 years, 2848 (44.39%) participants were male, and 1339 (20.87%) participants had CVD events in W5. Overall, BMI, WC, SBP, DBP, proportion of diabetes and dyslipidemia, levels of WBC, Hb, FPG, UA, CRP, TC, TG, HbA1c, proportion of CVD in W5 all decreased with increasing eGDR (all P < 0.001). However, individuals with higher levels of eGDR tended to have smoking history (P < 0.001).

Fig. 1
figure 1

Flow chart for participants selection in CHARLS. *Incomplete information includes age, gender, BMI, diabetes history, hypertension history, WC, HbA1c, and CVD condition

Full size image
Table 1 Baseline characteristics of the study population stratified by quartiles of eGDR
Full size table

Feature selection

The Boruta algorithm verified the importance of sixteen variables that were most strongly related with the risk of CVD (Fig. 2). Although other relevant factors including smoking history and diabetes were excluded because of the low Z-value in comparison to the shadow feature, they were nonetheless included in the study based on earlier research and clinical experience [21]. Factors were selected for the final complete adjustment model when their Z-scores in the Boruta analysis were higher than the shadow features, or when added to the model, they had the largest matched effect among a group of biomarkers (max, mean, and min), or they were based on previous findings and clinical constraints.

Fig. 2
figure 2

Feature selection for CVD using the Boruta algorithm. A. The process of feature selection; B. The value evolution of Z-score in the screening process. The horizontal axis shows the name of each variable and the number of times the classifier is run in Fig. 2A and B, respectively. While the vertical axis represents the Z-value of each variable. The green boxes and lines represent confirmed variables, the yellow ones represent tentative attributes, and the red ones represent rejected variables in the model calculation. BMI Body mass index, WC Waist circumference, SBP Systolic blood pressure, DBP Diastolic blood pressure, WBC White blood cell, PLT Platelets, Hb Hemoglobin, FPG Fasting plasma glucose, BUN Blood urea nitrogen, UA Uric acid, CRP C-reactive protein, TC Total cholesterol, TG Triglyceride, LDL-C Low-density lipoprotein cholesterol, HDL-C High-density lipoprotein cholesterol, HbA1c Hemoglobin A1c, eGDR Estimated glucose disposal rate

Full size image

Association between eGDR and incident CVD

During a median follow-up period of 106.5 months, there were 1339 (20.87%) cases of incident CVD, including 1025 (15.96%) cases of heart disease and 439 (6.84%) case of stroke, were reported in W5. The dose-response curves between eGDR and CVD were presented in Fig. 3, RCS curves suggested a significant and linear relationship between eGDR and the incidence of CVD, heart disease, and stroke in with or without adjusting for covariates (all P for overall < 0.001 and P for non-linear > 0.05). The inflection point for the eGDR was calculated automatically during the analysis using RCS analysis. When the eGDR was 10.53, the risk of CVD was differentiated, and the HR value of the TyG index was near 1.

In the fully adjusted model (Model 3), participants with a unit increasement in eGDR was associated with a 21.2% (HR 0.788, 95% CI 0.669–0.929) lower risk of CVD, a 18.3% (HR 0.817, 95% CI 0.678–0.985) decreased risk of heart disease, and 39.5% (HR 0.705, 95% CI 0.539–0.923) lower risk of stroke (Table 2). Furthermore, we defined four categories of patients based on the quartiles of eGDR. The results of multivariate Cox proportional hazards regression models showed that, compared with the lowest quartile, the highest eGDR quartile was related to a decreased risk of CVD (HR 0.686, 95% CI 0.545–0.862), heart disease (HR 0.675, 95% CI 0.521–0.874), and stroke (HR 0.593, 95% CI 0.392–0.897). A trend test for decreasing CVD risk with increasing quartiles of eGDR was also significant in all three models (all P for trend < 0.05).

Fig. 3
figure 3

Restricted cubic spline curves for CVD according to the eGDR. Hazard ratios are indicated by solid lines and 95% CIs by shaded areas. The horizontal dotted line represents the hazard ratio of 1.0. The adjusted models adjusted for factors including age, gender, BMI, WC, SBP, DBP, smoking, hypertension, diabetes, dyslipidemia, FPG, UA, TC, TG, LDL-C, HDL-C, and HbA1c screened by the Boruta algorithm and clinical experience. CVD Cardiovascular disease, eGDR Estimated glucose disposal rate

Full size image
Table 2 Multivariate-adjusted hazard ratios (95% confidence intervals) of eGDR for CVD incidence
Full size table

Predictive utility test

Figure 4 illustrated the predictive utility of eGDR for CVD risk. ROC curve analysis demonstrated that the covariate-adjusted model incorporating the eGDR achieved an AUC of 0.712, indicating statistically superior discriminatory performance compared to reference model (AUC = 0.624) excluding this metabolic parameter. The calibration curve and the χ2 likelihood ratio test (χ2 = 4.876, P = 0.771) both suggested an excellent goodness-of-fit of the multivariate model. DCA was performed to examine the clinical utility of the model, revealing a favorable overall net benefit and clinical impact within most reasonable threshold probability of the model. Moreover, NRI analysis revealed that the comprehensive ability of the multivariate model for CVD identification is improved by 7% (NRI = 0.070, P = 0.002).

Fig. 4
figure 4

Predictive utility test of eGDR for CVD risk. A. The area under the receiver operating characteristic (ROC) curve (AUC); B. Calibration curve; C. Decision curve analysis (DCA); D. Scatter diagram of the net reclassification improvement (NRI). CVD Cardiovascular disease, eGDR Estimated glucose disposal rate

Full size image

Subgroup analysis

As shown in Table 3, the association between eGDR and CVD incidence was determined by subgroups analysis according to age, gender, BMI, smoking, diabetes, and dyslipidemia. The results were almost consistent with the major findings, and no finding suggested a potential interaction between these factors and eGDR in terms of the risk of CVD (all P > 0.05).

Table 3 Subgroup analysis of multivariate-adjusted hazard ratios (95% confidence intervals) for the association between eGDR and CVD incidence
Full size table

Discussion

In our population-based cohort study, we determined the association of eGDR with the risk of incident CVD (including heart disease and stroke) in 6416 individuals with a median follow-up period of 106.5 months. To sum up, we found that a higher eGDR level indicated a lower long-term CVD incidence risk in China, and this association also remained significant even after adjusting for all confounders. Individuals with a unit increasement in eGDR was related to a 21.2% lower risk of CVD, a 18.3% decreased risk of heart disease, and 39.5% lower risk of stroke in the fully adjusted model. Compared to the lowest quartile, the highest eGDR quartile was also related to a decreased risk of CVD in all three models, and the trend in incident CVD risk was similar with the main findings across subgroups. Dose-response analysis indicated linear relationships between eGDR and the incidence of CVD, heart disease, and stroke. Predictive utility test showed that the eGDR significantly enhanced the predictive power of risk modes for adverse cardiovascular endpoints. The findings highlight the importance of focusing on non-conventional risk factors such as IR, which may contribute considerably to the residual risk of CVD even when traditional risk variables are controlled. Incorporating eGDR or other surrogate markers of IR into routine clinical assessments may help strengthen the ability to identify high-risk individuals and develop targeted CVD prevention strategies.

Several previous studies have highlighted the evidence that IR is associated with diabetes, impaired lipid metabolism, and the elevated blood pressure, all of which are major risk factors for developing incident CVD risk [22,23,24]. Previous studies revealed that HOMA-IR, as a reliable surrogate marker of IR, is associated with higher risk of incident CVD in general populations or adults with or without diabetes [25,26,27]. The association between another surrogate indicator, triglyceride-glucose (TyG index), and CVD has been widely examined in the previous studies, and the results consistently showed elevated TyG index positively contributed to increased risk of CVD [28,29,30]. Overall, these findings imply that IR may be a novel and valuable biomarker for predicting incident CVD. However, HOMA-IR is calculated using fasting blood glucose and insulin measurements, which severely limits its use because fasting insulin is not regularly evaluated in general population. Furthermore, various factors, including as the use of insulin, insulin sensitizers, and insulin secretagogues, might interfere with the assessment of HOMA-IR, leading in misclassification [31, 32]. Data from 4861 participants showed that predictive performance of eGDR for CVD incidence was markedly superior to that of HOMA-IR and the TyG index [33]. Similarly, our findings, as expected, indicate that eGDR is a promising biomarker to predict incident CVD risk including heart disease and stroke in the Chinese population. This could be related to the incorporation of clinical and laboratory data into the calculation of eGDR, thus providing a more complete assessment of IR. Compared with the invasive and time-consuming conventional approaches for assessing IR status, was derived through WC, presence of hypertension, and HbA1c from a single sample at the same time, even in primary health care settings, enhancing the generalizability and practicality of eGDR in clinical applications and epidemiological studies. More importantly, eGDR has comparable accuracy to HIEG clamp technique and HOMA-IR in assessing the status of IR, presented an adequate capacity in estimating the incidence of CVD [34, 35]. Taking together, our results indicated eGDR may serve as a reliable and credible non-invasive predictor of IR, and has potential predictive value for CVD risk.

Substantial studies have confirmed that a decreased eGDR level is closely linked with an increased risk of CVD. Data from 11,656 US individuals with a median follow-up of 12.8 years revealed that the association between reduced eGDR and all-cause and cardiovascular mortality was independently significant, contributing to the identification of individuals at high risk for different levels of glucose tolerances [36]. Consistent with the results, findings from a cross-sectional study of 10,895 participants conducted in southeastern China supported that each SD increase of eGDR brought a 25.9% (HR 0.741, 95% CI 0.636–0.864) risk reduction for prevalent ischemic heart disease and implicated the optimization of eGDR in identifying ischemic heart disease [37]. Similarly, the China National Stroke Registry III study demonstrated that patients with higher eGDR were significantly associated with higher incidence of excellent functional outcome among patients with acute ischemic stroke, independent of traditional cardiovascular predictors [38]. It is worth noting that both WC, hypertension and HbA1c utilized in developing the eGDR are all regarded as time-varying exposures. However, the evidence presented above cannot be ignored for the possibility of regression dilution bias caused by a single measurement of eGDR [39]. Several studies have attempted to address these limitations and methodological deficiencies by evaluating the influence of eGDR change on the study outcomes. Based on two repeated measurements of eGDR, results from two large prospective cohorts in Europe and Asia of 16,656 participants demonstrated that individuals with increasing and persistently low level of eGDR displayed the increased risk of incident CVD, emphasizing the significant importance of dynamic monitoring of eGDR level for the CVD prevention and treatment in general population [40].

Our study also found that the trend in incident CVD risk were broadly consistent with the main results across subgroups, indicating the sensitivity was not greatly influenced by different population characteristics and eGDR probably did not need adjust by age, gender, BMI, smoking, diabetes, and dyslipidemia conditions. It is worth mentioning that in our study, hypertension was not included as a factor for subgroup analysis because there was no hypertension in the Q3 and Q4 groups. As we are aware, hypertension is one of the most significant risk factors for CVD and plays a vital role in the occurrence and development of CVD [41]. As shown in a previous study by our group involving 810 hypertensive patients, IR was associated with adverse cardiovascular outcomes and those with the most severe IR had a 47.0% increased risk of CVD at 1-year follow-up [28]. Additionally, considering the bidirectional relationship between IR and obesity, previous studies have explored whether obesity mediated the association of eGDR and CVD. A prospective cohort study of 5512 participants suggested that obesity partly mediated the relationship, therefore controlling body mass may alleviate the unfavorable influencing of IR on circulatory system [42]. Interestingly, our findings suggested that eGDR seemed to demonstrate greater sensitivity in predicting CVD among non-diabetic individuals (Q4 vs. Q1: HR 0.632, 95% CI 0.504–0.794 in non-diabetic individuals; HR 0.740, 95% CI 0.388–1.414 in diabetic individuals). Earlier studies indicated that individuals with diabetes may demonstrate diminished sensitivity to predictive indicators attributed to the presence of additional risk factors, a phenomenon commonly referred to as the ceiling effect [43]. Identifying risk variables in general populations may lead to earlier intervention attempts, considerably reducing disease burden. Consequently, we aimed to explore the relationship between eGDR and incident CVD in community-dwelling populations without pre-existing CVD in a nationwide prospective cohort study. Our findings may help to extend the understanding of the relationship between IR and CVD incidence, as well as aid to risk stratification and the improvement of CVD prevention efforts.

Mechanisms underlying the relationship between IR and CVD have been described. In the IR state, β-cells detect the severity of IR and increase insulin secretion to compensate for the defective insulin action [44]. Substantial studies have shown that high concentrations of insulin can accelerate the atherosclerotic process through various mechanisms, such as (i) facilitating smooth muscle cell proliferation, collagen cross-linking, and collagen deposition [45], (ii) exacerbating inflammation and oxidative stress [46], (iii) causing platelet overactivity and abnormal adhesion induction [47], and (iv) enhancing LDL-C transport into arterial smooth muscle cells [48]. The cumulative effect of pathological changes may cause a number of issues such as endothelial dysfunction and increased vascular stiffness, ultimately resulting in heart structural and functional abnormalities [49, 50].

The main strengths of the study include that we utilized data from CHARLS, a national survey conducted in China that is widely recognized for its reliability and representativeness. Moreover, our study specifically focuses on the predictive utility of IR assessed by eGDR for incident CVD risk. This novel approach may contribute to the existing theoretical research and clinical management on the CVD and provides valuable insights for policy makers. Nevertheless, several potential limitations of this study should be acknowledged. Firstly, the CHARLS project only included Chinese individuals aged 45 and above, which may affect the generalizability of our conclusions; thus, further studies are urgently required to confirm our results through broader population samples and diverse demographics. Secondly, the study’s primary endpoint was ascertained through subject-reported physician-diagnosed CVD events, introducing potential recall bias. Given the absence of mortality data, interpretations of the epidemiological association between eGDR and CVD risk require cautious consideration of these methodological constraints. Thirdly, limited by the nature of observational studies, we are unable to establish a very clear causality between eGDR and incident CVD risk. However, this appears to be minimal, as the results remained stable even when we excluded individuals who experienced the endpoint within the first two years. Meanwhile, although our model adjusts for many covariates, residual confounding still cannot be completely eliminated, which is a common problem in observational studies.

Conclusion

Overall, our study shows that individuals with lower eGDR levels had higher risk of CVD in the future, and this relationship remained significant after adjustment for other confounders. For individuals at high risk of CVD, eGDR might be a valuable tool for risk categorization and management. Moreover, incorporating eGDR into the multivariate risk model significantly enhances the predictive performance for CVD. Importantly, further studies are also required to confirm our findings and identify the potential mechanisms behind the association between eGDR and incident CVD risk.

Data availability

No datasets were generated or analysed during the current study.

Abbreviations

eGDR:

Estimated glucose disposal rate

IR:

Insulin resistance

CVD:

Cardiovascular disease

CHARLS:

China Health and Retirement Longitudinal Study

RCS:

Restricted cubic spline

ROC:

Receiver operating characteristic

NRI:

Net reclassification improvement

DCA:

Decision curve analysis

HIEG:

Hyperinsulinemic–euglycemic

HOMA-IR:

Homeostasis model assessment for IR

References

  1. Collaborators GBDCoD. Global burden of 288 causes of death and life expectancy decomposition in 204 countries and territories and 811 subnational locations, 1990–2021: a systematic analysis for the global burden of disease study 2021. Lancet. 2024;403(10440): 2100–32.

    Article  Google Scholar 

  2. Diseases GBD, Injuries C. Global incidence, prevalence, years lived with disability (YLDs), disability-adjusted life-years (DALYs), and healthy life expectancy (HALE) for 371 diseases and injuries in 204 countries and territories and 811 subnational locations, 1990–2021: a systematic analysis for the global burden of disease study 2021. Lancet. 2024;403(10440): 2133–61.

    Article  Google Scholar 

  3. Zou Z, Cini K, Dong B, Ma Y, Ma J, Burgner DP, et al. Time trends in cardiovascular disease mortality across the BRICS: an age-period-cohort analysis of key nations with emerging economies using the global burden of disease study 2017. Circulation. 2020;141(10):790–9.

    Article  PubMed  Google Scholar 

  4. Sun J, Lyu S, Dai Z. The impacts of socioeconomic status and lifestyle on health status of residents: evidence from Chinese general social survey data. Int J Health Plann Manag. 2019;34(4):1097–108.

    Article  Google Scholar 

  5. Zhou M, Wang H, Zhu J, Chen W, Wang L, Liu S, et al. Cause-specific mortality for 240 causes in China during 1990–2013: a systematic subnational analysis for the global burden of disease study 2013. Lancet. 2016;387(10015):251–72.

    Article  PubMed  Google Scholar 

  6. Report on cardiovascular health and diseases in China. 2023: an updated summary. Chin Circulation J. 2024;39(07):625–60.

    Google Scholar 

  7. Wu Y, Benjamin EJ, MacMahon S. Prevention and control of cardiovascular disease in the rapidly changing economy of China. Circulation. 2016;133(24):2545–60.

    Article  PubMed  PubMed Central  Google Scholar 

  8. Barber SL, Huang B, Santoso B, Laing R, Paris V, Wu C. The reform of the essential medicines system in China: a comprehensive approach to universal coverage. J Glob Health. 2013;3(1):010303.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Caturano A, Galiero R, Vetrano E, Sardu C, Rinaldi L, Russo V et al. Insulin-Heart axis: bridging physiology to insulin resistance. Int J Mol Sci. 2024;25(15).

  10. Ebert T, Anker SD, Ruilope LM, Fioretto P, Fonseca V, Umpierrez GE, et al. Outcomes with finerenone in patients with chronic kidney disease and type 2 diabetes by baseline insulin resistance. Diabetes Care. 2024;47(3):362–70.

    Article  CAS  PubMed  Google Scholar 

  11. Muniyappa R, Lee S, Chen H, Quon MJ. Current approaches for assessing insulin sensitivity and resistance in vivo: advantages, limitations, and appropriate usage. Am J Physiol Endocrinol Metab. 2008;294(1):E15–26.

    Article  CAS  PubMed  Google Scholar 

  12. Epstein EJ, Osman JL, Cohen HW, Rajpathak SN, Lewis O, Crandall JP. Use of the estimated glucose disposal rate as a measure of insulin resistance in an urban multiethnic population with type 1 diabetes. Diabetes Care. 2013;36(8):2280–5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Zabala A, Darsalia V, Lind M, Svensson AM, Franzen S, Eliasson B, et al. Estimated glucose disposal rate and risk of stroke and mortality in type 2 diabetes: a nationwide cohort study. Cardiovasc Diabetol. 2021;20(1):202.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Peng J, Li A, Yin L, Yang Q, Pan J, Yi B. Estimated glucose disposal rate predicts renal progression in type 2 diabetes mellitus: a retrospective cohort study. J Endocr Soc. 2023;7(7):bvad069.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Garofolo M, Gualdani E, Scarale MG, Bianchi C, Aragona M, Campi F, et al. Insulin resistance and risk of major vascular events and all-cause mortality in type 1 diabetes: a 10-year follow-up study. Diabetes Care. 2020;43(10):e139–41.

    Article  PubMed  PubMed Central  Google Scholar 

  16. Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China health and retirement longitudinal study (CHARLS). Int J Epidemiol. 2014;43(1):61–8.

    Article  PubMed  Google Scholar 

  17. McEvoy JW, McCarthy CP, Bruno RM, Brouwers S, Canavan MD, Ceconi C, et al. 2024 ESC guidelines for the management of elevated blood pressure and hypertension. Eur Heart J. 2024;45(38):3912–4018.

    Article  PubMed  Google Scholar 

  18. American Diabetes Association Professional Practice C. 2. Classification and diagnosis of diabetes: standards of medical care in Diabetes-2022. Diabetes Care. 2022;45(Suppl 1):S17–38.

    Article  Google Scholar 

  19. Guideline for diagnosis and treatment of obesity. (2024 edition). Chin J Dig Surg. 2024;23(10):1237-60.

  20. Kursa M, Rudnicki W. Feature selection with the Boruta package. J Stat Softw. 2010;36(11):1–13.

    Article  Google Scholar 

  21. Kahn SE, Cooper ME, Del Prato S. Pathophysiology and treatment of type 2 diabetes: perspectives on the past, present, and future. Lancet. 2014;383(9922):1068–83.

    Article  CAS  PubMed  Google Scholar 

  22. Abdul-Ghani M, Maffei P, DeFronzo RA. Managing insulin resistance: the forgotten pathophysiological component of type 2 diabetes. Lancet Diabetes Endocrinol. 2024;12(9):674–80.

    Article  CAS  PubMed  Google Scholar 

  23. Colantoni A, Bucci T, Cocomello N, Angelico F, Ettorre E, Pastori D, et al. Lipid-based insulin-resistance markers predict cardiovascular events in metabolic dysfunction associated steatotic liver disease. Cardiovasc Diabetol. 2024;23(1):175.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. da Silva AA, do Carmo JM, Li X, Wang Z, Mouton AJ, Hall JE. Role of hyperinsulinemia and insulin resistance in hypertension: metabolic syndrome revisited. Can J Cardiol. 2020;36(5):671–82.

    Article  PubMed  Google Scholar 

  25. Bonora E, Kiechl S, Willeit J, Oberhollenzer F, Egger G, Meigs JB, et al. Insulin resistance as estimated by homeostasis model assessment predicts incident symptomatic cardiovascular disease in Caucasian subjects from the general population: the Bruneck study. Diabetes Care. 2007;30(2):318–24.

    Article  PubMed  Google Scholar 

  26. Bonora E, Formentini G, Calcaterra F, Lombardi S, Marini F, Zenari L, et al. HOMA-estimated insulin resistance is an independent predictor of cardiovascular disease in type 2 diabetic subjects: prospective data from the Verona diabetes complications study. Diabetes Care. 2002;25(7):1135–41.

    Article  PubMed  Google Scholar 

  27. Effoe VS, Wagenknecht LE, Echouffo Tcheugui JB, Chen H, Joseph JJ, Kalyani RR et al. Sex differences in the association between insulin resistance and incident coronary heart disease and stroke among Blacks without diabetes mellitus: the Jackson heart study. J Am Heart Assoc. 2017;6(2).

  28. Tao S, Yu L, Li J, Huang L, Huang X, Zhang W, et al. Association between the triglyceride-glucose index and 1-year major adverse cardiovascular events in patients with coronary heart disease and hypertension. Cardiovasc Diabetol. 2023;22(1):305.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Zhang X, Du Y, Zhang T, Zhao Z, Guo Q, Ma X, et al. Prognostic significance of triglyceride-glucose index in acute coronary syndrome patients without standard modifiable cardiovascular risk factors. Cardiovasc Diabetol. 2024;23(1):270.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Tao S, Yu L, Li J, Xie Z, Huang L, Yang D, et al. Prognostic value of triglyceride-glucose index in patients with chronic coronary syndrome undergoing percutaneous coronary intervention. Cardiovasc Diabetol. 2023;22(1):322.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Fonseca V, Bakris GL, Bell DS, McGill JB, Raskin P, Messerli FH, et al. Differential effect of beta-blocker therapy on insulin resistance as a function of insulin sensitizer use: results from GEMINI. Diabet Med. 2007;24(7):759–63.

    Article  CAS  PubMed  Google Scholar 

  32. van der Aa MP, Elst MA, van de Garde EM, van Mil EG, Knibbe CA, van der Vorst MM. Long-term treatment with Metformin in obese, insulin-resistant adolescents: results of a randomized double-blinded placebo-controlled trial. Nutr Diabetes. 2016;6(8):e228.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Song J, Ma R, Yin L. Associations between estimated glucose disposal rate and arterial stiffness and mortality among US adults with non-alcoholic fatty liver disease. Front Endocrinol (Lausanne). 2024;15:1398265.

    Article  PubMed  Google Scholar 

  34. Mertens J, Roosens L, Braspenning R, Vandebeeck J, Francque S, De Block C. The 13 C glucose breath test accurately identifies insulin resistance in people with type 1 diabetes. J Clin Endocrinol Metab. 2024.

  35. Penno G, Solini A, Orsi E, Bonora E, Fondelli C, Trevisan R, et al. Insulin resistance, diabetic kidney disease, and all-cause mortality in individuals with type 2 diabetes: a prospective cohort study. BMC Med. 2021;19(1):66.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Guo R, Tong J, Cao Y, Zhao W. Association between estimated glucose disposal rate and cardiovascular mortality across the spectrum of glucose tolerance in the US population. Diabetes Obes Metab. 2024;26(12):5827–35.

    Article  CAS  PubMed  Google Scholar 

  37. Xuan J, Juan D, Yuyu N, Anjing J. Impact of estimated glucose disposal rate for identifying prevalent ischemic heart disease: findings from a cross-sectional study. BMC Cardiovasc Disord. 2022;22(1):378.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Lu Z, Xiong Y, Feng X, Yang K, Gu H, Zhao X, et al. Insulin resistance estimated by estimated glucose disposal rate predicts outcomes in acute ischemic stroke patients. Cardiovasc Diabetol. 2023;22(1):225.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. VanderWeele TJ, Jackson JW, Li S. Causal inference and longitudinal data: a case study of religion and mental health. Soc Psychiatry Psychiatr Epidemiol. 2016;51(11):1457–66.

    Article  PubMed  Google Scholar 

  40. Zheng X, Han W, Li Y, Jiang M, Ren X, Yang P, et al. Changes in the estimated glucose disposal rate and incident cardiovascular disease: two large prospective cohorts in Europe and Asia. Cardiovasc Diabetol. 2024;23(1):403.

    Article  PubMed  PubMed Central  Google Scholar 

  41. McManus R, Constanti M, Floyd CN, Glover M, Wierzbicki AS, National Institute for H. Managing cardiovascular disease risk in hypertension. Lancet. 2020;395(10227):869–70.

    Article  PubMed  Google Scholar 

  42. Zhang Z, Zhao L, Lu Y, Xiao Y, Zhou X. Insulin resistance assessed by estimated glucose disposal rate and risk of incident cardiovascular diseases among individuals without diabetes: findings from a nationwide, population based, prospective cohort study. Cardiovasc Diabetol. 2024;23(1):194.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Wang T, Lu J, Su Q, Chen Y, Bi Y, Mu Y, et al. Ideal cardiovascular health metrics and major cardiovascular events in patients with prediabetes and diabetes. JAMA Cardiol. 2019;4(9):874–83.

    Article  PubMed  PubMed Central  Google Scholar 

  44. DeFronzo RA. Insulin resistance, lipotoxicity, type 2 diabetes and atherosclerosis: the missing links. The Claude Bernard lecture 2009. Diabetologia. 2010;53(7):1270–87.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Mellor KM, Bell JR, Young MJ, Ritchie RH, Delbridge LM. Myocardial autophagy activation and suppressed survival signaling is associated with insulin resistance in fructose-fed mice. J Mol Cell Cardiol. 2011;50(6):1035–43.

    Article  CAS  PubMed  Google Scholar 

  46. Bahramzadeh A, Samavarchi Tehrani S, Goodarzi G, Seyyedebrahimi S, Meshkani R. Combination therapy of Metformin and Morin attenuates insulin resistance, inflammation, and oxidative stress in skeletal muscle of high-fat diet-fed mice. Phytother Res. 2024;38(2):912–24.

    Article  CAS  PubMed  Google Scholar 

  47. Gerrits AJ, Koekman CA, van Haeften TW, Akkerman JW. Platelet tissue factor synthesis in type 2 diabetic patients is resistant to Inhibition by insulin. Diabetes. 2010;59(6):1487–95.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Porter KE, Riches K. The vascular smooth muscle cell: a therapeutic target in type 2 diabetes? Clin Sci (Lond). 2013;125(4):167–82.

    Article  CAS  PubMed  Google Scholar 

  49. Perticone M, Maio R, Gigliotti S, Shehaj E, Toscani AF, Capomolla A et al. Mutual effect modification between insulin resistance and endothelial dysfunction in predicting incident heart failure in hypertensives. Biomedicines. 2023;11(8).

  50. Hill MA, Yang Y, Zhang L, Sun Z, Jia G, Parrish AR, et al. Insulin resistance, cardiovascular stiffening and cardiovascular disease. Metabolism. 2021;119:154766.

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

The present study utilized data from the CHARLS. We appreciate CHARLS research team and individuals included in the study.

Funding

This work was supported by the High Level Chinese Medical Hospital Promotion Project (No. HLCMHPP2023065), the National Natural Science Foundation of China (No. 82474494), and the National Key Research and Development Program of China (No. 2022YFC3500102).

Author information

Authors and Affiliations

  1. Department of Cardiology, Guang’anmen Hospital, China Academy of Chinese Medical Sciences, Beijing, 100053, China

    Shiyi Tao, Jun Li, Ji Wu, Xuanchun Huang, Zicong Xie, Tiantian Xue, Yonghao Li & Lilan Su

  2. Graduate School, Beijing University of Chinese Medicine, Beijing, 100029, China

    Shiyi Tao, Lintong Yu & Yonghao Li

Authors
  1. Shiyi Tao
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Lintong Yu
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Jun Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Ji Wu
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Xuanchun Huang
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Zicong Xie
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Tiantian Xue
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Yonghao Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Lilan Su
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

JL and ST conceived the concept and designed the study. ST and LY performed the data collection, data analyses and wrote the original manuscript. JL took responsibility for the integrity of the data and the accuracy of the data analysis. JW, XH, and ZX contributed to data collection and figure mapping. TX, YL, and LS were involved in statistical analyses and data checking.

Corresponding author

Correspondence to Jun Li.

Ethics declarations

Ethics approval and consent to participate

The CHARLS project was approved by the Institutional Review Board of Peking University (approval number: IRB00001052-11015 for household survey and IRB00001052-11014 for blood sample), and all participants voluntarily participated and signed an informed consent form.

Consent for publication

All authors have consent for publication.

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.

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

Tao, S., Yu, L., Li, J. et al. Insulin resistance quantified by estimated glucose disposal rate predicts cardiovascular disease incidence: a nationwide prospective cohort study. Cardiovasc Diabetol 24, 161 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12933-025-02672-1

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12933-025-02672-1

Keywords

Cardiovascular Diabetology

ISSN: 1475-2840