Abstract
Background
Patients with both acute myocardial infarction (AMI) and chronic kidney disease (CKD) face a markedly poor prognosis, a key driver of which is insulin resistance (IR). This study aims to systematically evaluate and compare the predictive performance of four commonly used IR indices for major adverse cardiovascular events (MACE), and to assess their incremental value over the GRACE score in this patient group.
Methods
This retrospective cohort study analyzed 1,803 patients with AMI and CKD. Multivariable Cox regression determined associations between IR indices and MACE. Predictive performance was evaluated using C-statistics, continuous net reclassification improvement (cNRI), and integrated discrimination improvement (IDI).
Results
During a median follow-up of 28.2 months, 462 MACE occurred. Patients with MACE were older, had higher female proportion, elevated GRACE score, and increased diabetes prevalence (all p < 0.05). the triglyceride-glucose (TyG) index and the atherogenic index of plasma (AIP) demonstrated linear associations with MACE risk, whereas TyG-body mass index (TyG-BMI) and metabolic score for insulin resistance (METS-IR) exhibited U-shaped nonlinear relationships (p < 0.001). The Area Under the Curve (AUCs) for MACE prediction were: TyG index 0.62, AIP 0.57, TyG-BMI 0.58, and METS-IR 0.56. Incorporating IR indices significantly enhanced the GRACE score’s predictive capacity, with TyG index providing the greatest incremental improvement (cNRI = 0.137, IDI = 0.03).
Conclusion
IR indices predict outcomes in patients with AMI and CKD and enhance GRACE score prediction, with TyG index demonstrating superior performance.
Keywords: Acute myocardial infarction, chronic kidney disease, insulin resistance indices
GRAPHICAL ABSTRACT
1. Introduction
It is reported that approximately 30–50% of patients with acute myocardial infarction (AMI) have concomitant chronic kidney disease (CKD), imposing at least a twofold increased risk of cardiovascular events [1–3]. Patients with CKD exhibit a higher incidence of CVD due to factors such as anemia, inflammation, malnutrition, and sympathetic overactivity; this risk escalates progressively with worsening renal dysfunction. However, patients with advanced CKD are often excluded from AMI clinical trials due to their poor prognosis and high burden of comorbidities, resulting in insufficient evidence for current risk management strategies [2,4–6]. Therefore, identifying novel prognostic biomarkers holds critical importance for risk management in this high-risk yet understudied population of patients with AMI and CKD.
Patients with AMI and CKD frequently exhibit concomitant metabolic syndrome. In 2023, the American Heart Association (AHA) systematically designated this constellation of conditions as Cardiovascular-Kidney-Metabolic (CKM) syndrome, with insulin resistance (IR) implicated as a key driver of its progression [7]. IR, denoting a diminished sensitivity or responsiveness to the metabolic actions of insulin, not merely exacerbates renal impairment but also elevates the risk of cardiovascular complications [8–11]. IR is particularly relevant in CKD, as chronic inflammation, oxidative stress, and impaired insulin clearance in CKD exacerbate IR, while IR in turn accelerates renal and vascular injury. This bidirectional interaction links metabolic dysfunction to both renal and cardiovascular complications [9,12]. Compared to established measures of IR such as the Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) and the Quantitative Insulin Sensitivity Check Index (QUICKI), four alternative biomarkers offer greater simplicity and accessibility: TyG index, AIP, TyG-BMI, and METS-IR [13,14]. The TyG index is defined as the product of fasting triglyceride (TG) and glucose levels [15]. Building upon this, TyG-BMI represents a composite index derived by combining the TyG index with body mass index (BMI) [16]. In contrast, AIP was originally introduced in 2001 as a marker of atherosclerosis and cardiovascular disease [17]. METS-IR is calculated from various metabolic parameters, including fasting blood glucose (FBG), BMI, TG, and high-density lipoprotein cholesterol (HDL-C) [18].
Current research indicates that elevated levels of IR indices correlate closely with adverse outcomes in cardiovascular and cerebrovascular diseases, metabolic disorders, and malignancies [19–22]. A large prospective study reported that higher TyG levels were associated with both the incidence and mortality of AMI [23]. Another analysis of more than one thousand critically ill AMI patients found that TyG-BMI showed a U-shaped association with long-term mortality, supporting its prognostic utility in severe cardiac conditions [24]. In community-based cohorts, persistent elevations of the TyG index were linked to CKD development, while higher AIP values were related to short-term mortality among hospitalized cardiac patients [25,26]. Long-term data from a population-based study further showed that elevated METS-IR levels were independently associated with new-onset coronary heart disease and diabetes [27]. Collectively, these findings confirm the robustness of surrogate IR indices and provide a rationale for their application in high-risk patients with AMI and CKD. However, the association between different IR indices and clinical outcomes in patients with AMI and CKD remain poorly explored, particularly lacking comprehensive head-to-head comparisons in this high-risk population.
This study aimed to examine the associations between IR indices and major adverse cardiovascular events (MACE) in patients with AMI and CKD, and to evaluate and compare their value in enhancing the predictive capacity of the Global Registry of Acute Coronary Events (GRACE) risk score.
2. Methods
2.1. Study design
This retrospective observational cohort study was conducted at the First Affiliated Hospital of Wenzhou Medical University. The inclusion criteria is patients diagnosed with both AMI and CKD who were underwent percutaneous coronary intervention (PCI) between January 2012 and December 2022. Exclusion criteria comprised: (1) missing essential variables for IR index calculation; (2) age <18 or >90 years; and (3) incomplete follow-up data (Figure 1).
Figure 1.
Flowchart of the study design. AMI, acute myocardial infarction; CKD, chronic kidney disease; IR, insulin resistance.
The Institutional Review Board of the First Affiliated Hospital of Wenzhou Medical University granted ethical approval (No. YS2024-736) for this study, which was conducted in accordance with the Declaration of Helsinki. The retrospective design exempted the requirement for participant consent.
2.2. Data collection
Comprehensive baseline clinical data – including medical histories, demographic information, laboratory parameters, echocardiographic measurements, and medication regimens – were retrospectively extracted from the institutional electronic medical record system using standardized protocols.
All measurements were conducted during index hospitalization under rigorous protocols. Anthropometric data collection utilized calibrated instruments: body weight was measured to the nearest 0.1 kg using an RGZ-120 digital scale (Wuxi Weigher Factory, China) and height to the nearest 0.1 cm with a Seca 213 stadiometer (Hamburg, Germany), with participants in light clothing without footwear. BMI was derived as weight (kg)/height2 (m2).
Fasting peripheral venous blood samples were obtained within 24 h of admission after ≥8-h overnight fasting. Biochemical analyses (FBG, TG, HDL-C, serum creatinine, and BNP) were performed on a Cobas 8000 analyzer (Roche Diagnostics, Basel, Switzerland) using standardized enzymatic methods. All procedures adhered to institutional quality control protocols with regular calibration and validation.
2.3. Definitions
According to current guidelines [28], AMI comprises ST-segment elevation myocardial infarction (STEMI) and non-ST-segment elevation myocardial infarction (NSTEMI). STEMI is defined by typical chest discomfort or other ischemic symptoms accompanied by new ST-segment elevations in ≥2 contiguous leads or a new left bundle branch block, along with elevated cardiac biomarkers. NSTEMI is characterized by ischemic symptoms and elevated cardiac biomarkers, in the absence of ST-segment elevation on the electrocardiogram (ECG); it may exhibit ST-segment depression or T-wave inversion.
IR indices derived from fasting laboratory measurements included the TyG index calculated as ln[(TG, mg/dL) × (FBG, mg/dL)/2], TyG-BMI derived from TyG index × BMI, AIP expressed as log(TG/HDL-C), and METS-IR computed using (ln[2 × FBG (mg/dL) + TG (mg/dL)] × BMI)/(ln[HDL-C (mg/dL)]) [14–17].
CKD was defined as either an estimated glomerular filtration rate (eGFR) <60 mL/min/1.73m2 calculated via the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation or a documented clinical diagnosis of CKD in medical records [29,30].
The GRACE score is a guideline-recommended risk scoring system initially designed to predict 6-month mortality and cardiac events, calculated using a set of clinical variables outlined in published literature [31]. These variables include: age, heart rate, systolic blood pressure, serum creatinine, congestive heart failure, in-hospital PCI, in-hospital coronary aortic bypass grafting, history of MI, ST-segment depression, and elevated cardiac enzyme/marker levels [31].
Comorbidities and cardiovascular risk factors were defined through standardized criteria based on comprehensive medical record review. Current smoking status required self-reported active tobacco use within the 6 months preceding hospitalization. Hyperlipidemia was defined by documented clinical diagnosis, active lipid-lowering therapy, fasting low-density lipoprotein cholesterol (LDL-C) ≥130 mg/dL (3.36 mmol/L), or TC ≥200 mg/dL (5.17 mmol/L). Diabetes mellitus diagnosis required established clinical history, use of glucose-lowering agents, FBG ≥126 mg/dL (7.0 mmol/L), or HbA1c ≥6.5%. Hypertension was confirmed through clinical diagnosis documentation, antihypertensive medication use, or blood pressure measurements ≥140/90 mmHg on two separate occasions.
2.4. Outcomes and follow-up
The primary endpoint was MACE, comprising cardiovascular death, non-fatal reinfarction, or non-fatal ischemic stroke. Cardiovascular death encompassed fatalities attributable to cardiovascular causes. Reinfarction required recurrent elevation of cardiac biomarkers above the 99th percentile upper reference limit, accompanied by at least one of the following criteria: (1) recurrent ischemic symptoms lasting ≥20 min, (2) new ischemic electrocardiographic changes (ST-segment deviation ≥0.5 mm or T-wave inversion ≥1 mm in ≥2 contiguous leads), (3) development of pathological Q waves in ≥2 contiguous leads, or (4) angiographically confirmed coronary thrombus in epicardial vessels (≥2.0 mm diameter). Ischemic stroke was defined as a focal neurological deficit persisting >24 h with documented acute ischemic lesions on diffusion-weighted magnetic resonance imaging or computed tomography, excluding hemorrhagic or non-ischemic etiologies.
The median follow-up duration was 28.2 months (interquartile range: 12.6–44.0 months). Follow-up information was obtained through electronic medical records and standardized telephone interviews aligned with standard care after discharge. Patients who could not be contacted or whose survival status could not be verified through hospital or public records were considered lost to follow-up. These individuals were censored at the date of their last confirmed clinical contact. The observation period commenced at discharge and continued until death verification. Participants without MACE were censored at their last validated clinical encounter, which included either in-person consultations or structured telephone assessments.
2.5. Statistical analysis
Normality assessment employed the Shapiro-Wilk test supplemented by visual inspection of P-P plots and histograms. Continuous variables are presented as mean ± standard deviation (SD) or median (interquartile range [IQR]), while categorical variables are reported as frequencies (percentages). Comparative analyses utilized Student’s t-test for normally distributed continuous variables, Mann-Whitney U test for skewed continuous data, and Pearson’s chi-square test for categorical variables.
Event-free survival was analyzed using Kaplan-Meier curves with between-group differences assessed by log-rank tests. Multivariable Cox proportional hazards regression models examined independent associations between IR indices and MACE, incorporating clinically justified covariates: age, gender, diabetes, hypertension, STEMI, eGFR, dual antiplatelet therapy (DAPT), angiotensin-converting enzyme inhibitors/angiotensin receptor blockers (ACEI/ARB), beta-blockers, statins, current smoking status, left ventricular ejection fraction (LVEF), and GRACE risk score. Proportional hazards assumptions were verified via Schoenfeld residual analysis, with multicollinearity assessed by variance inflation factors (VIF >10 indicating significant collinearity). Restricted cubic splines (RCS) with knots at the 5th, 35th, 65th, and 95th percentiles evaluated potential nonlinear relationships. Missing covariate data (<10%) were addressed through multiple imputation using chained equations under the missing-at-random assumption [32].
Prespecified subgroup analyses assessed effect modification across clinically relevant strata (age, gender, BMI, LVEF, eGFR, AMI type, hypertension, and diabetes) by incorporating interaction terms into adjusted Cox models. The predictive performance of IR indices was quantified through: (1) receiver operating characteristic (ROC) curve analysis, (2) continuous net reclassification improvement (cNRI), and (3) integrated discrimination improvement (IDI) – evaluated both independently and for incremental value when combined with the GRACE score.
To further investigate the impact of albuminuria, a key cardiovascular risk factor, on our findings, we performed a sensitivity analysis in the subset of patients with available urine albumin-to-creatinine ratio (UACR) data (n = 1,190, representing 66.0% of the full cohort). Albuminuria was defined as a UACR ≥30 mg/g. Within this subset, patients were stratified by the presence or absence of albuminuria. Multivariable Cox proportional hazards models were constructed separately for each stratum to evaluate the associations between the IR indices and MACE. Additionally, an interaction term (IR index × albuminuria status) was incorporated into the models to formally test for potential effect modification by albuminuria on these associations.
All analyses used two-tailed tests with statistical significance defined as p < 0.05. Data management was performed using SPSS Statistics 25.0 (IBM Corp.), with advanced statistical modeling conducted in R 4.4.3 (R Foundation for Statistical Computing).
3. Results
3.1. Demographic characteristics
A total of 1803 patients [mean age (72.45 ± 10.24) years] were enrolled in this study and divided into two groups according to outcome: Non-MACE group(1341, 74.4%) and MACE group(462, 25.6%). The baseline characteristics of the study population are presented in Table 1. Compared with the Non-MACE group, the MACE group had older patients (73.45 ± 9.33 years), a lower proportion of male (64.07%), higher GRACE scores (128.03 ± 20.72), and a higher prevalence of diabetes (52.38%). Additionally, the MACE group had lower LVEF and eGFR, but higher FBG, TC, TG, rates of multivessel disease (50.43%) and left circumflex artery (LCX) lesions (69.05%), along with lower usage rates of statin (89.83%) and ACEI/ARB (30.09%). No significant differences were found in other baseline characteristics between the two groups (All p > 0.05).
Table 1.
Baseline characteristics.
| Variables | Total (n = 1803) | Non-MACE (n = 1341) | MACE (n = 462) | p Value |
|---|---|---|---|---|
| Age (years) | 72.45 ± 10.24 | 72.10 ± 10.52 | 73.45 ± 9.33 | 0.01 |
| Male | 1241 (68.83) | 945 (70.47) | 296 (64.07) | 0.01 |
| BMI (kg/m2) | 23.54 ± 3.25 | 23.52 ± 3.23 | 23.62 ± 3.33 | 0.55 |
| SBP (mmHg) | 136.69 ± 26.27 | 136.67 ± 26.40 | 136.75 ± 25.94 | 0.954 |
| DBP (mmHg) | 74.34 ± 14.28 | 74.47 ± 14.23 | 73.96 ± 14.42 | 0.507 |
| STEMI | 964 (53.47) | 722 (53.84) | 242 (52.38) | 0.588 |
| Hypertension | 1385 (76.82) | 1017 (75.84) | 368 (79.65) | 0.094 |
| Diabetes | 839 (46.53) | 597 (44.52) | 242 (52.38) | 0.003 |
| Hyperlipidemia | 836 (46.37) | 615 (45.86) | 221 (47.84) | 0.463 |
| Current smoker | 651 (36.11) | 492 (36.69) | 159 (34.42) | 0.38 |
| GRACE score | 121.60 ± 22.19 | 119.39 ± 22.26 | 128.03 ± 20.72 | <0.001 |
| Laboratory data | ||||
| eGFR (mL/min/1.73m2) | 46.31 (28.98, 58.51) | 47.29 (30.77, 58.60) | 43.75 (20.90, 58.37) | 0.012 |
| FBG (mmol/L) | 6.90 (5.30, 9.40) | 6.80 (5.20, 8.90) | 7.30 (5.60, 10.90) | <0.001 |
| TC (mmol/L) | 4.55 (3.79, 5.29) | 4.49 (3.75, 5.29) | 4.60 (3.94, 5.29) | 0.027 |
| TG (mmol/L) | 1.41 (1.00, 2.05) | 1.34 (0.96, 1.95) | 1.65 (1.14, 2.10) | <0.001 |
| HDL-C (mmol/L) | 0.97 (0.81, 1.17) | 0.97 (0.82, 1.18) | 0.96 (0.81, 1.15) | 0.207 |
| LDL-C (mmol/L) | 2.55 (1.93, 3.24) | 2.54 (1.91, 3.25) | 2.57 (1.98, 3.23) | 0.633 |
| Echocardiographic and angiographic data | ||||
| LVEF (%) | 55.25 ± 11.67 | 55.88 ± 11.54 | 53.39 ± 11.86 | <0.001 |
| Multivessel disease | 767 (42.54) | 534 (39.82) | 233 (50.43) | <0.001 |
| LAD stenosis > = 50% | 1475 (81.81) | 1096 (81.73) | 379 (82.03) | 0.884 |
| LCX stenosis > = 50% | 1134 (62.90) | 815 (60.78) | 319 (69.05) | 0.002 |
| RCA stenosis > = 50% | 1156 (64.12) | 850 (63.39) | 306 (66.23) | 0.271 |
| Medical therapy | ||||
| DAPT | 1668 (92.51) | 1232 (91.87) | 436 (94.37) | 0.078 |
| Beta blocker | 957 (53.08) | 714 (53.24) | 243 (52.60) | 0.81 |
| ACEI/ARB | 620 (34.39) | 481 (35.87) | 139 (30.09) | 0.024 |
| Statin | 1660 (92.07) | 1245 (92.84) | 415 (89.83) | 0.039 |
| IR indices | ||||
| TyG index | 9.02 ± 0.69 | 8.95 ± 0.67 | 9.23 ± 0.73 | <0.001 |
| AIP | 0.51 ± 0.29 | 0.49 ± 0.28 | 0.57 ± 0.30 | <0.001 |
| TyG-BMI | 212.55 ± 34.48 | 210.58 ± 33.78 | 218.26 ± 35.85 | <0.001 |
| METS-IR | 39.06 ± 7.24 | 38.65 ± 6.92 | 40.23 ± 7.98 | <0.001 |
Abbreviations: TyG, triglyceride-glucose; T1, tertile 1; T2, tertile 2; T3, tertile 3; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; MI, myocardial infarction; PCI, percutaneous coronary intervention; CABG, coronary artery bypass grafting; ACS, acute coronary syndrome; NSTEMI, non–ST-segment elevation myocardial infarction; STEMI, ST-segment elevation myocardial infarction; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; FBG, fasting blood glucose; TC, total cholesterol; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; ALT, alanine aminotransferase; AST, aspartate aminotransferase; GRACE, The Global Registry of Acute Coronary Events; LVEF, left ventricular ejection fraction; LAD, left anterior descending coronary; LCX, left circumflex artery; RCA, right coronary artery; DAPT, dual antiplatelet therapy; ACEI/ARB, angiotensin-converting enzyme inhibitor/angiotensin II receptor blocker; TyG, triglyceride-glucose; TyG-BMI, triglyceride-glucose index with body mass index; AIP, atherogenic index of plasma; METS-IR, metabolic score for insulin resistance. The bold values in the tables indicate statistical significance, defined as a p-value < 0.05.
Overall, covariate missingness was modest (<10% across all variables), and no single variable had >10% missing values. A variable-level summary of observed and missing counts is provided in Supplementary Table S1, and the missingness pattern across variables is visualized in Supplementary Figure S1. Missing covariate data were addressed using multiple imputation with chained equations under a missing-at-random assumption, as prespecified in the Statistical Analysis section.
3.2. Associations between IR indices and MACE
During the median follow-up period of 28.2 months (12.6–44 months), MACE occurred in 462 patients (25.6%). As shown in the Kaplan-Meier curve (Figure 2), higher tertiles of all IR indices (TyG index, AIP, TyG-BMI, and METS-IR) were associated with increased incidence of MACE (All p < 0.05).
Figure 2.
Kaplan-Meier survival curves stratified by category of insulin resistance (IR) indexes. (A) Triglyceride-glucose (TyG) index; (B) Atherogenic index of plasma (AIP); (C) TyG index with body mass index (TyG-BMI); (D) Metabolic score for insulin resistance (METS-IR).
3.2.1. TyG index and risk of MACE
As shown in the RCS analysis (Figure 3A), a linear association was observed between the TyG index and MACE risk (P for nonlinear = 0.219). In Cox regression analysis (Table 2), the TyG index, when analyzed as a continuous variable, was significantly associated with MACE events both in unadjusted (HR: 1.61, 95% CI: 1.41–1.83, p < 0.001) and adjusted models (HR: 1.53, 95% CI: 1.33–1.77, p < 0.001). When categorized into tertiles, T3 of the TyG index remained significantly associated with MACE after adjustment (T3 adjusted HR: 1.94, 95% CI: 1.51–2.50; p < 0.001). Significant interactions emerged for age, while associations remained consistent across other subgroups (Supplementary Figure S2).
Figure 3.
RCS for the associations between the IR indexes and MACE. Red shadows and lines represent the 95% CI. TyG index (A), AIP (B), TyG-BMI (C), METS-IR (D). HR (95%CI) was adjusted according to the model 3. RCS, restricted cubic spline; IR, insulin resistance; MACE, major adverse cardiovascular event; CI, confidence interval; TyG, triglyceride-glucose; TyG-BMI, triglyceride-glucose index with body mass index; AIP, atherogenic index of plasma; METS-IR, metabolic score for insulin resistance.
Table 2.
Association between IR indices and MACE (cox regression).
| Unadjusted |
Adjusted* |
|||
|---|---|---|---|---|
| HR (95% CI) | p-Value | HR (95% CI) | p-Value | |
| TyG (continuous) | 1.61 (1.41–1.83) | <0.001 | 1.53 (1.33–1.77) | <0.001 |
| T1 | Ref | Ref | Ref | Ref |
| T2 | 1.29 (1.01–1.67) | 0.047 | 1.21 (0.93–1.56) | 0.149 |
| T3 | 2.20 (1.75–2.78) | <0.001 | 1.94 (1.51–2.50) | <0.001 |
| TyG-BMI (continuous) | 1.01 (1.01–1.01) | 0.002 | 1.01 (1.01–1.01) | <0.001 |
| T1 | Ref | Ref | Ref | Ref |
| T2 | 1.08 (0.85–1.37) | 0.524 | 1.15 (0.91–1.47) | 0.247 |
| T3 | 1.41 (1.13–1.77) | 0.003 | 1.45 (1.15–1.83) | 0.002 |
| AIP (continuous) | 2.47 (1.81–3.38) | <0.001 | 2.34 (1.69–3.23) | <0.001 |
| T1 | Ref | Ref | Ref | Ref |
| T2 | 1.25 (0.98–1.60) | 0.067 | 1.22 (0.96–1.56) | 0.108 |
| T3 | 1.79 (1.42–2.25) | <0.001 | 1.63 (1.29–2.05) | <0.001 |
| METS-IR (continuous) | 1.02 (1.01–1.03) | <0.001 | 1.02 (1.01–1.04) | <0.001 |
| T1 | Ref | Ref | Ref | Ref |
| T2 | 1.13 (0.89–1.44) | 0.302 | 1.10 (0.86–1.40) | 0.440 |
| T3 | 1.50 (1.20–1.89) | <0.001 | 1.44 (1.14–1.82) | 0.002 |
Adjusted for age, gender, diabetes, hypertension, STEMI, eGFR, dual antiplatelet therapy (DAPT), angiotensin-converting enzyme inhibitors/angiotensin receptor blockers (ACEI/ARB), beta-blockers, statins, current smoking status, left ventricular ejection fraction (LVEF), and GRACE risk score. IR, insulin resistance; HR, hazard ratio; CI, confidence interval. Other abbreviations as in Table 1. The bold values in the tables indicate statistical significance, defined as a p-value < 0.05.
3.2.2. AIP and risk of MACE
Figure 3B demonstrated a linear dose-response relationship between AIP and MACE risk (P for nonlinear = 0.384). Per 1-unit increase in AIP, MACE risk rose 134% (adjusted HR 2.34, 95% CI 1.69–3.23, p < 0.001). When treated as categorical variable (Table 2), T3 showed progressively elevated hazards versus T1 (T3 adjusted HR 1.63, 95% CI 1.29–2.05, p < 0.001). There were no significant differences observed within subgroups like diabetes and hypertension (Supplementary Figure S3).
3.2.3. TyG-BMI and risk of MACE
A U-shaped relationship between TyG-BMI and MACE was observed in RCS (P for nonlinear = 0.029; Figure 3C). In Table 2, Cox analysis indicated 1% increased MACE hazard per unit rise (adjusted HR 1.01, 95% CI 1.01–1.01, p < 0.001). Tertiles comparisons showed T3 conferred 45% excess risk versus T1 (T3 adjusted HR 1.45, 95% CI 1.15–1.83, p = 0.002). Associations persisted uniformly across gender, BMI, eGFR and type of AMI except for age, diabetes, hypertension, and LVEF interactions (P for interaction < 0.05) (Supplementary Figure S4).
3.2.4. METS-IR and risk of MACE
RCS curves revealed significant U-shaped relationships between METS-IR and MACE (P for onlinear = 0.022, Figure 3D). Each unit increment conferred 2% excess MACE risk (adjusted HR 1.02, 95% CI 1.01–1.04, p < 0.001). Individuals in the highest TyG-BMI tertile faced 44% higher risk than the lowest-tertile reference (T3 adjusted HR 1.44, 95% CI 1.14–1.82, p = 0.002). Subgroup analysis (Supplementary Figure S5) confirmed robust METS-IR-MACE associations across age, gender, BMI, diabetes,and hypertension subgroups except LVEF.
3.3. Comparative analysis of IR indices for predicting MACE
Table 3 demonstrates that TyG index exhibits the strongest discriminative performance for MACE prediction among IR indices, with an AUC of 0.62 (95% CI 0.58–0.65), sensitivity of 0.65(0.62–0.68), and specificity of 0.54(0.49–0.58). Statistically significant improvements in AUC were observed for TyG compared to TyG-BMI (Difference = 0.099, p = 0.002), AIP (Difference = 0.043, p = 0.01), and METS-IR (Difference = 0.062, p = 0.002). TyG index also showed superior cNRI versus TyG-BMI (0.136, p < 0.001) and METS-IR (0.106, p < 0.001), along with greater IDI (0.017 and 0.016, p < 0.001). No significant differences in AUC emerged between TyG-BMI, AIP, and METS-IR (all p > 0.05). TyG-BMI demonstrated marginally higher specificity (0.57) than other indices (TyG index: 0.54, AIP: 0.56, METS-IR: 0.52).
Table 3.
Comparative analysis of IR indices for predicting MACE.
| Discrimination ability | AUC (95%CI) | Sensitivity (95%CI) | Specificity (95%CI) |
|---|---|---|---|
| TyG | 0.62 (0.58–0.65) | 0.65 (0.62–0.68) | 0.54 (0.49–0.58) |
| TyG-BMI | 0.57 (0.54–0.60) | 0.55 (0.52–0.58) | 0.57 (0.53–0.62) |
| AIP | 0.58 (0.55–0.61) | 0.58 (0.55–0.61) | 0.56 (0.52–0.61) |
| METS-IR | 0.56 (0.53–0.59) | 0.60 (0.58–0.63) | 0.52 (0.47–0.56) |
| Comparison | TyG vs TyG-BMI |
TyG vs AIP |
TyG vs METS-IR |
|||
|---|---|---|---|---|---|---|
| Difference | p Value | Difference | p Value | Difference | p Value | |
| AUC | 0.099 | 0.002 | 0.043 | 0.01 | 0.062 | 0.002 |
| cNRI | 0.136 | <0.001 | 0.086 | 0.159 | 0.106 | <0.001 |
| IDI | 0.017 | <0.001 | 0.008 | 0.159 | 0.016 | <0.001 |
The bold values in the tables indicate statistical significance, defined as a p-value < 0.05.
| Comparison | TyG-BMI vs AIP |
TyG-BMI vs METS-IR |
AIP vs METS-IR |
|||
|---|---|---|---|---|---|---|
| Difference | p Value | Difference | p Value | Difference | p Value | |
| AUC | −0.056 | 0.363 | −0.037 | 0.652 | 0.019 | 0.206 |
| cNRI | −0.126 | 0.027 | −0.044 | 0.465 | 0.116 | 0.027 |
| IDI | −0.009 | 0.013 | −0.005 | 0.106 | 0.009 | 0.007 |
3.4. Model performance after the addition of IR indices to the GRACE score for predicting
As shown in Table 4, incorporating TyG into the GRACE score significantly enhances MACE prediction, yielding the largest AUC improvement (0.657 vs 0.609, ΔAUC 0.048, p < 0.001) among IR indices. This combined model also achieved statistically significant IDI (0.03, p < 0.001) and cNRI (0.137, p < 0.001). While other IR indices showed marginal AUC gains (range: 0.632–0.634), their improvements were inconsistent: TyG-BMI provided nonsignificant IDI enhancement (0.003, p = 0.425) and AIP exhibited significant IDI but small cNRI improvements. METS-IR showed borderline nonsignificant cNRI (0.086, p = 0.053).
Table 4.
Model performance after the addition of IR indices to the GRACE score for predicting MACE.
| Model | AUC [95%CI] | p Value | IDI | p Value | cNRI | p Value |
|---|---|---|---|---|---|---|
| GRACE score | 0.609 [0.581, 0.637] | Ref | Ref | Ref | ||
| GRACE+TyG | 0.657 [0.625, 0.686] | <0.001 | 0.03 | <0.001 | 0.137 | <0.001 |
| GRACE+TyG-BMI | 0.634 [0.606, 0.661] | <0.001 | 0.003 | 0.425 | 0.076 | 0.033 |
| GRACE+AIP | 0.633 [0.605, 0.662] | 0.042 | 0.02 | <0.001 | 0.135 | <0.001 |
| GRACE+METS-IR | 0.632 [0.605, 0.660] | 0.01 | 0.008 | 0.02 | 0.086 | 0.053 |
Abbreviations as in Tables 1–3. The bold values in the tables indicate statistical significance, defined as a p-value < 0.05.
3.5. Sensitivity analyses
The associations between IR indices and MACE were consistent after excluding patients with diabetes (Supplementary Table S1). Similar results were observed when CKD severity was operationalized using alternative eGFR categories (>45, 30–45, <30 mL/min/1.73m2), with no meaningful heterogeneity across strata (Supplementary Table S2-S5).
Furthermore, to assess the potential influence of albuminuria on the observed associations, we performed a stratified analysis in the subset of patients with available UACR data (n = 1,190; Table 5). Although tests for interaction did not reach statistical significance (all P for interaction > 0.05), it is noteworthy that the HR for the associations between the IR indices and MACE showed numerical differences between patients with and without albuminuria. Specifically, for the continuous TyG index, a higher adjusted HR was observed in patients without albuminuria (adjusted HR: 2.00, 95% CI: 1.36–2.95, p < 0.001) compared to those with albuminuria (adjusted HR: 1.45, 95% CI: 1.16–1.82, p < 0.001). This trend was more pronounced when the TyG index was analyzed by tertiles: among patients without albuminuria, those in the highest tertile (T3) had a 4.56-fold higher risk of MACE (95% CI: 2.05–10.13) compared to those in the lowest tertile (T1), whereas the corresponding risk was 1.70 (95% CI: 1.18–2.46) in patients with albuminuria. A similar pattern was also observed for the AIP. These results indicate that the predictive value of the IR indices persists across different albuminuria statuses, with a trend toward stronger associations observed in the subgroup without albuminuria.
Table 5.
Association between IR indices and MACE (cox regression) after excluding patients without UACR data.
| Patients with albuminuria (n = 852) |
Patients without albuminuria (n = 338) |
||||
|---|---|---|---|---|---|
| HR (95% CI) | p Value | HR (95% CI) | p Value | P for interaction | |
| TyG (continuous) | 1.45 (1.16–1.82) | <0.001 | 2.00 (1.36–2.95) | <0.001 | 0.382 |
| T1 | Ref | Ref | Ref | Ref | |
| T2 | 0.89 (0.60–1.33) | 0.577 | 2.34 (1.07–5.13) | 0.033 | |
| T3 | 1.70 (1.18–2.46) | 0.005 | 4.56 (2.05–10.13) | <0.001 | |
| TyG-BMI (continuous) | 1.01 (1.01–1.01) | 0.014 | 1.01 (1.00–1.02) | 0.068 | 0.701 |
| T1 | Ref | Ref | Ref | Ref | |
| T2 | 1.10 (0.76–1.61) | 0.613 | 0.90 (0.47–1.72) | 0.75 | |
| T3 | 1.49 (1.05–2.13) | 0.027 | 1.35 (0.72–2.54) | 0.35 | |
| AIP (continuous) | 2.56 (1.54–4.25) | <0.001 | 4.32 (1.77–10.55) | <0.001 | 0.561 |
| T1 | Ref | Ref | Ref | Ref | |
| T2 | 1.02 (0.70–1.48) | 0.931 | 2.73 (1.32–5.66) | 0.007 | |
| T3 | 1.56 (1.11–2.19) | 0.011 | 4.04 (1.96–8.33) | <0.001 | |
| METS-IR (continuous) | 1.04 (1.02–1.06) | <0.001 | 1.03 (0.99–1.07) | 0.182 | 0.507 |
| T1 | Ref | Ref | Ref | Ref | |
| T2 | 1.10 (0.75–1.61) | 0.631 | 1.25 (0.65–2.43) | 0.500 | |
| T3 | 1.83 (1.27–2.62) | 0.001 | 1.21 (0.64–2.28) | 0.553 | |
Adjusted for age, gender, diabetes, hypertension, STEMI, eGFR, dual antiplatelet therapy (DAPT), angiotensin-converting enzyme inhibitors/angiotensin receptor blockers (ACEI/ARB), beta-blockers, statins, current smoking status, left ventricular ejection fraction (LVEF), and GRACE risk score. UACR, Urine Albumin-to-Creatinine Ratio; IR, insulin resistance; HR, hazard ratio; CI, confidence interval. Other abbreviations as in Table 1. The bold values in the tables indicate statistical significance, defined as a p-value < 0.05.
4. Discussion
This retrospective cohort study of 1,803 patients with AMI and CKD demonstrates that surrogate indices of IR are independently associated with an increased risk of MACE and provide significant incremental prognostic value beyond the established GRACE risk score. Specifically, the TyG index exhibited the strongest predictive performance among the evaluated IR indices (TyG index, AIP, TyG-BMI, METS-IR), both as a standalone predictor (AUC 0.62) and, most notably, when integrated with the GRACE score, yielding the largest improvements in discrimination (ΔAUC 0.048), reclassification (cNRI = 0.137), and integrated discrimination (IDI = 0.03). Although the incremental improvements in AUC, cNRI, and IDI were statistically significant, their magnitude was modest. This finding highlights that while IR indices, particularly the TyG index, provide added prognostic value beyond the GRACE score, their clinical utility should be interpreted as complementary rather than standalone risk predictors in patients with AMI and CKD. Notably, while the TyG index and AIP showed linear associations with MACE risk, TyG-BMI and METS-IR demonstrated U-shaped relationships, highlighting potential differences in how these composite indices reflect underlying pathophysiology in this complex patient population.
Our findings extend prior evidence linking IR indices to adverse cardiovascular outcomes by specifically validating their prognostic utility and enabling head-to-head comparison in the high-risk, understudied cohort of patients suffering from both AMI and CKD, a population increasingly recognized under the CKM syndrome paradigm where IR plays a pivotal role.
Insulin resistance: roots and effects on heart and kidney health
IR results from the complex interplay of genetic and acquired factors, with genetic predisposition primarily driven by gene mutations and hereditary susceptibility: studies indicate that IR in SHORT syndrome patients frequently associates with aberrant insulin signaling due to loss-of-function mutations in the PI3KR3 gene, while a subset of metabolic syndrome cases also demonstrates significant genetic links to IR development [33,34]. Among acquired contributors, obesity – particularly visceral adiposity – serves as a pivotal instigator, impairing insulin sensitivity and glucose homeostasis across multiple tissues (most notably the liver) through mechanisms including chronic inflammation, lipotoxicity, mitochondrial dysfunction, reactive oxygen species (ROS) accumulation, and lysosomal dysfunction [35,36]. Furthermore, gut microbiota dysbiosis has emerged as a novel mechanism contributing to IR, potentially via increased bacterial endotoxin release [37]. Notably, recent work led by Tadashi Takeuchi revealed that variations in gut microbial carbohydrate metabolism patterns are associated with insulin sensitivity, suggesting a potential link that may involve specific metabolic pathways [38].
Insulin acts not only on metabolic tissues but also directly regulates cardiovascular and renal functions, mediating complex pathophysiological alterations. Under IR, impairment of the PI3K pathway – which physiologically enhances nitric oxide (NO) production – shifts signaling toward the MAPK pathway. This aberrant signaling activation promotes vascular pathology, including vasoconstriction, stiffening, and ultimately, atherosclerotic plaque formation [39,40]. In vitro experiments by Ruth M. Foutz’s team demonstrated that MAPK pathway activation mediates renal mesangial cell hyperplasia through modulation of large-conductance Ca2+-activated K+ channels (BK channels) [41]. Interestingly, evidence indicates that insulin resistance may compromise podocyte structural integrity [42,43]. Functionally, human podocytes display specialized insulin sensitivity within the renal filtration barrier, whereas those isolated from diabetic model mice exhibit impaired insulin responsiveness in glucose metabolism. Collectively, these findings suggest that podocyte-specific insulin resistance potentially underlies renal injury [42,44]. Furthermore, excessive activation of the epithelial sodium channel (EnNaC), demonstrated in mouse models, represents another contributory factor linking IR to cardiovascular dysfunction [11,45].
In clinical practice, the hyperinsulinemic-euglycemic clamp, while gold-standard for IR assessment, faces clinical limitations due to cost and invasiveness [13]. Recent studies validate non-insulin-based indices (TyG, TyG-BMI, AIP, METS-IR) as practical alternatives, demonstrating both reliability and operational simplicity for widespread research and clinical use [15–18].
Comparative MACE predictive efficacy of IR indices
The TyG index, a novel and validated surrogate for IR, demonstrates robust predictive capacity across multiple disease states. In a large-scale cohort of 98,849 individuals, elevated TyG levels were significantly associated with increased MI risk (1,555 incident cases) during a median follow-up of 11.03 years [46]. Furthermore, a 2023 study identified TyG as a predictor of both in-hospital and 1-year mortality among critical ill patients with coronary artery disease and CKD, indicating that increased TyG values within a defined range were associated with higher in-hospital mortality [25]. Our findings align with these observations: RCS analysis revealed a linear increasing relationship between TyG elevation and MACE risk. Notably, TyG provided incremental predictive value to the GRACE risk score.
Similarly, the TyG-BMI index – derived from the TyG index – has also been linked to cardiovascular risk. A retrospective analysis by Yang Cheng’s team demonstrated a proportional increase in MACCE incidence with elevated TyG-BMI levels among elderly and female patients [47]. However, their study reported no significant improvement in MACE prediction by TyG-BMI in the elderly cohort, whereas our cohort analysis confirmed its capacity to enhance risk stratification using the GRACE score. Although METS-IR was reported to outperform both TyG and TyG-BMI for predicting post-ablation atrial fibrillation recurrence in one trial [48], this study found that it not only demonstrated limited prognostic value but also lacked robustness in subgroup analyses. These findings suggest its clinical application requires careful consideration across different disease. Across several studies, TyG-BMI has shown a U-shaped association with adverse outcomes, whereas comparable patterns for METS-IR have not been consistently observed [49,50]. A plausible explanation is that both indices incorporate BMI, a parameter closely linked to the so-called ‘obesity paradox’ observed in cardiovascular and renal populations, in which higher or intermediate BMI is paradoxically associated with lower mortality – possibly reflecting better nutritional reserves, muscle mass preservation, and greater tolerance to acute physiological stress [51]. Prior ICU data in patients with stage 3–5 CKD also demonstrated a U-shaped relationship between BMI and mortality, with both underweight and extreme obesity associated with poorer outcomes [50]. Evidence from Japanese cohorts further suggests that the impact of overweight or obesity on prognosis after acute myocardial infarction may depend on disease severity and hemodynamic stability [52]. Mechanistically, this U-shaped pattern may reflect the competing influences of malnutrition, inflammation, and metabolic overload: low TyG-BMI or METS-IR values may indicate catabolic or frail states with reduced metabolic reserve, while excessively high levels suggest metabolic stress, IR, and endothelial dysfunction, all contributing to increased cardiovascular risk.
AIP is a validated predictor of cardiovascular risk. Its association with plaque vulnerability in AMI patients has been confirmed in a study using optical coherence tomography technology [53]. A 2024 meta-analysis further demonstrated that elevated AIP levels significantly correlate with increased MACE risk among elderly AMI patients [54]. Beyond cardiovascular implications, AIP levels may exhibit a nonlinear relationship with CKD incidence, potentially mediated by malnutrition, systemic inflammation, and adipose deposition [55]. In this study, AIP demonstrated superior predictive performance characterized by high specificity and sensitivity, while substantially enhancing the risk stratification capacity of the GRACE score. When compared with the TyG index, however, AIP demonstrates slightly inferior predictive performance. Although equally significant for prognostic assessment in stroke patients, the TyG index exhibits broader predictive scope for cardiovascular diseases regardless of the glucose metabolism status [56–58]. Consistent with this pattern, our study confirmed that while both TyG and AIP show robust predictive capabilities, the TyG index provides superior enhancement to the GRACE score’s risk stratification efficacy.
In the sensitivity analysis, although the test for interaction did not reach statistical significance, we observed a phenomenon worthy of further exploration: across all evaluated IR indices, particularly the TyG index and AIP, the strength of their associations with MACE – as measured by HR – was numerically higher in patients without albuminuria. This finding may have important clinical implications. First, it confirms that even among patients with normal albuminuria, IR indicators such as the TyG index remain effective in identifying individuals at high risk of MACE, and their discriminative ability may even be stronger. This may be because, in patients with albuminuria, cardiovascular risk is already substantially elevated due to renal injury itself [57], leaving relatively limited room for incremental prediction by other risk factors such as IR. Second, this supports the independence of IR as a core driver of CKM syndrome. Our results suggest that metabolic disturbances related to IR constitute an independent and significant cardiovascular threat even in the absence of overt glomerular injury, as indicated by albuminuria. This provides a rationale for widespread screening of IR status in CKD patients with reduced eGFR but without albuminuria, and even in the broader population at high cardiovascular risk. As an easily accessible metric, the TyG index may hold particular value for refined risk stratification in such patients.
Mechanisms linking IR to cardio-renal injury
IR exerts multiple deleterious effects beyond metabolic dysregulation, extending to vascular and renal injury. Compensatory hyperinsulinemia accompanying IR initiates a cascade of hemodynamic and inflammatory alterations. Specifically, it contributes to endothelial dysfunction, activates the renin–angiotensin–aldosterone system (RAAS) and sympathetic nervous system, and enhances proximal tubular sodium reabsorption, collectively promoting hypertension and low-grade inflammation [59].
Furthermore, emerging evidence suggests that hyperinsulinemia itself acts as an independent risk factor for albuminuria, even in the absence of overt hyperglycemia, likely through glomerular endothelial injury and altered tubular sodium handling [60,61].
Importantly, the relationship between IR and CKD is bidirectional. CKD can in turn aggravate IR – independent of obesity or diabetes – through the accumulation of uremic toxins, chronic inflammation, and impaired insulin signaling in skeletal muscle and hepatic tissues [59,62,63]. In parallel, activation of mineralocorticoid pathways in CKD, characterized by elevated aldosterone levels, further exacerbates systemic inflammation, hypertension, fibrosis, and left ventricular hypertrophy, thereby linking renal dysfunction to adverse cardiovascular outcomes [59,64].
Collectively, these interrelated mechanisms highlight the complex pathophysiological interplay between IR, hyperinsulinemia, and cardiorenal dysfunction.
Strengths and limitations
This study possesses several notable strengths. First, it represents one of the largest cohorts (n = 1,803) specifically investigating the prognostic value of IR indices in the high-risk, clinically challenging population of patients with concomitant AMI and CKD, a group often underrepresented in clinical trials. Second, we conducted a comprehensive head-to-head comparison of four distinct, readily calculable IR surrogate indices (TyG index, AIP, TyG-BMI, METS-IR), providing novel insights into their relative predictive performance and differential associations (linear vs. U-shaped) with MACE risk in this population. Third, the rigorous statistical methodology employed multivariable Cox regression with extensive adjustment for established confounders and GRACE score, restricted cubic splines to model nonlinear relationships, and robust metrics (AUC, cNRI, IDI) to evaluate both standalone and incremental predictive value. Fourth, the findings demonstrate clear clinical utility, identifying the TyG index as a simple, accessible biomarker that significantly enhances risk stratification beyond the widely used GRACE score, potentially informing targeted risk management strategies within the emerging CKM syndrome paradigm.
However, several limitations warrant consideration. First, the retrospective single-center design and exclusion of screened patients may introduce selection bias, while unmeasured confounders could potentially influence outcomes. In particular, patients older than 90 years were excluded because their frailty, multiple comorbidities, and age-related physiological changes could markedly affect the incidence of MACE, potentially confounding the associations observed in this study. Furthermore, anthropometric and biochemical parameters such as BMI and lipid profiles may not accurately reflect metabolic status in this very elderly population. Second, IR indices were derived solely from baseline measurements, precluding assessment of metabolic profile dynamics during follow-up. Third, the absence of gold-standard insulin sensitivity measures limits mechanistic interpretation. Because albuminuria data were unavailable, CKD definition relied solely on eGFR criteria, which might have underestimated CKD prevalence, potentially excluding early-stage CKD cases with preserved filtration but subclinical albuminuria.
Conclusion
The findings demonstrate that IR indices are associated with MACE in patients with AMI and CKD and improve the predictive capacity of the GRACE score, with the TyG index showing the best performance. Given the clinical accessibility and computational convenience of IR indices, they enable patient risk stratification, thus guiding personalized treatment.
Supplementary Material
Acknowledgements
We thank all the participants and colleagues who contributed to this study. Weicheng Ni, Qingwei Ni, and Ruihao Jiang contributed equally to this work as co-first authors. They were responsible for conceptualization, data curation, formal analysis, investigation, methodology, and original draft preparation. Xuliang Ying, Zhongda Zhu, and Jing Chen contributed to data curation, resources, and validation. Yuanzhen Lin and Shanhu Cao provided resources, project administration, and validation support. Changxi Chen participated in methodology development and supervision. Corresponding authors Xi Zhou and Hao Zhou led the study conceptualization, secured funding, supervised project administration and resources, and edited the manuscript.
Funding Statement
This work was supported by the National Natural Science Foundation of China [Grant Number 82271620] and the Zhejiang Provincial Natural Science Foundation of China [Grant Number LY22H020011].
Ethical approval
Ethical approval for this retrospective study was granted by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University (YS2024-736); the research complies with the Declaration of Helsinki. In line with standard practice for such studies, the Institutional Review Board waived the need for informed consent. All analytical procedures were performed on fully anonymized data.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Data availability statement
All research data can be obtained from the corresponding author upon reasonable request.
References
- 1.Bergmark BA, Mathenge N, Merlini PA, et al. Acute coronary syndromes. Lancet. 2022;399(10332):1347–1358. doi: 10.1016/s0140-6736(21)02391-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lai AC, Bienstock SW, Sharma R, et al. A personalized approach to chronic kidney disease and cardiovascular disease. J Am Coll Cardiol. 2021;77(11):1470–1479. doi: 10.1016/j.jacc.2021.01.028. [DOI] [PubMed] [Google Scholar]
- 3.Ruff CT, Braunwald E.. The evolving epidemiology of acute coronary syndromes. Nat Rev Cardiol. 2011;8(3):140–147. doi: 10.1038/nrcardio.2010.199. [DOI] [PubMed] [Google Scholar]
- 4.Charytan D, Kuntz R.. The exclusion of patients with chronic kidney disease from clinical trials in coronary artery disease. Kidney Int. 2006;70(11):2021–2030. doi: 10.1038/sj.ki.5001934. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Sarnak MJ, Amann K, Bangalore S, et al. Chronic kidney disease and coronary artery disease. J Am Coll Cardiol. 2019;74(14):1823–1838. doi: 10.1016/j.jacc.2019.08.1017. [DOI] [PubMed] [Google Scholar]
- 6.Gansevoort RT, Correa-Rotter R, Hemmelgarn BR, et al. Chronic kidney disease and cardiovascular risk: epidemiology, mechanisms, and prevention. Lancet. 2013;382(9889):339–352. doi: 10.1016/S0140-6736(13)60595-4. [DOI] [PubMed] [Google Scholar]
- 7.Ndumele CE, Neeland IJ, Tuttle KR, et al. A synopsis of the evidence for the science and clinical management of cardiovascular-kidney-metabolic (CKM) syndrome: a scientific statement from the American Heart Association. Circulation. 2023;148(20):1636–1664. doi: 10.1161/CIR.0000000000001186. [DOI] [PubMed] [Google Scholar]
- 8.Artunc F, Schleicher E, Weigert C, et al. The impact of insulin resistance on the kidney and vasculature. Nat Rev Nephrol. 2016;12(12):721–737. doi: 10.1038/nrneph.2016.145. [DOI] [PubMed] [Google Scholar]
- 9.Laakso M, Kuusisto J.. Insulin resistance and hyperglycaemia in cardiovascular disease development. Nat Rev Endocrinol. 2014;10(5):293–302. doi: 10.1038/nrendo.2014.29. [DOI] [PubMed] [Google Scholar]
- 10.Devesa A, Fuster V, Vazirani R, et al. Cardiac insulin resistance in subjects with metabolic syndrome traits and early subclinical atherosclerosis. Diabetes Care. 2023;46(11):2050–2057. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Hill MA, Yang Y, Zhang L, et al. Insulin resistance, cardiovascular stiffening and cardiovascular disease. Metabolism. 2021;119:154766. doi: 10.1016/j.metabol.2021.154766. [DOI] [PubMed] [Google Scholar]
- 12.De Cosmo S, Menzaghi C, Prudente S, et al. Role of insulin resistance in kidney dysfunction: insights into the mechanism and epidemiological evidence. Nephrol Dial Transplant. 2013;28(1):29–36. doi: 10.1093/ndt/gfs290. [DOI] [PubMed] [Google Scholar]
- 13.DeFronzo RA, Tobin JD, Andres R.. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol. 1979;237(3):E214–E223. doi: 10.1152/ajpendo.1979.237.3.e214. [DOI] [PubMed] [Google Scholar]
- 14.Del Prato S. Measurement of insulin resistance in vivo. Drugs. 1999;58(Supplement 1):3–6. doi: 10.2165/00003495-199958001-00002. [DOI] [PubMed] [Google Scholar]
- 15.Simental-Mendía LE, Rodríguez-Morán M, Guerrero-Romero F.. The product of fasting glucose and triglycerides as surrogate for identifying insulin resistance in apparently healthy subjects. Metab Syndr Relat Disord. 2008;6(4):299–304. doi: 10.1089/met.2008.0034. [DOI] [PubMed] [Google Scholar]
- 16.Ramírez-Vélez R, Pérez-Sousa M, González-Ruíz K, et al. Obesity- and lipid-related parameters in the identification of older adults with a high risk of prediabetes according to the American Diabetes Association: an Analysis of the 2015 Health, Well-Being, and Aging Study. Nutrients. 2019;11(11):2654. doi: 10.3390/nu11112654. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Dobiášová M, Frohlich J.. The plasma parameter log (TG/HDL-C) as an atherogenic index: correlation with lipoprotein particle size and esterification rate inapob-lipoprotein-depleted plasma (FERHDL). Clin Biochem. 2001;34(7):583–588. doi: 10.1016/s0009-9120(01)00263-6. [DOI] [PubMed] [Google Scholar]
- 18.Bello-Chavolla OY, Almeda-Valdes P, Gomez-Velasco D, et al. METS-IR, a novel score to evaluate insulin sensitivity, is predictive of visceral adiposity and incident type 2 diabetes. Eur J Endocrinol. 2018;178(5):533–544. doi: 10.1530/eje-17-0883. [DOI] [PubMed] [Google Scholar]
- 19.Tao LC, Xu J N, Wang T, et al. 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. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Lopez-Jaramillo P, Gomez-Arbelaez D, Martinez-Bello D, et al. Association of the triglyceride glucose index as a measure of insulin resistance with mortality and cardiovascular disease in populations from five continents (PURE study): a prospective cohort study. Lancet Healthy Longev. 2023;4(1):e23–e33. doi: 10.1016/S2666-7568(22)00247-1. [DOI] [PubMed] [Google Scholar]
- 21.Wei X, Min Y, Song G, et al. Association between triglyceride-glucose related indices with the all-cause and cause-specific mortality among the population with metabolic syndrome. Cardiovasc Diabetol. 2024;23(1):134. doi: 10.1186/s12933-024-02215-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kim YM, Kim JH, Park JS, et al. Association between triglyceride-glucose index and gastric carcinogenesis: a health checkup cohort study. Gastric Cancer. 2022;25(1):33–41. doi: 10.1007/s10120-021-01222-4. [DOI] [PubMed] [Google Scholar]
- 23.Chen N, Ma LL, Zhang Y, et al. Association of long-term triglyceride-glucose index patterns with the incidence of chronic kidney disease among non-diabetic population: evidence from a functional community cohort. Cardiovasc Diabetol. 2024;23(1):7. doi: 10.1186/s12933-023-02098-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Luo C, Li Q, Wang Z, et al. Association between triglyceride glucose-body mass index and all-cause mortality in critically ill patients with acute myocardial infarction: retrospective analysis of the MIMIC-IV database. Front Nutr. 2024;11:1399969. doi: 10.3389/fnut.2024.1399969. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ye Z, An S, Gao Y, et al. Association between the triglyceride glucose index and in-hospital and 1-year mortality in patients with chronic kidney disease and coronary artery disease in the intensive care unit. Cardiovasc Diabetol. 2023;22(1):110. doi: 10.1186/s12933-023-01843-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Wang Y, Zhang HF.. Associations of the atherogenic index of plasma with 28-day in-hospital mortality in patients with acute myocardial infarction: a retrospective cohort study from the eICU. Lipids Health Dis. 2025;24(1):202. doi: 10.1186/s12944-025-02630-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Tamehri Zadeh SS, Cheraghloo N, Masrouri S, et al. Association between metabolic score for insulin resistance and clinical outcomes: insights from the Tehran lipid and glucose study. Nutr Metab. 2024;21(1):34. doi: 10.1186/s12986-024-00808-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Thygesen K, Alpert JS, Jaffe AS, et al. Fourth universal definition of myocardial infarction. J Am Coll Cardiol. 2018;72(18):2231–2264. doi: 10.1016/j.jacc.2018.08.1038. [DOI] [PubMed] [Google Scholar]
- 29.Andrassy KM. Comments on ‘KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease’. Kidney Int. 2013;84(3):622–623. doi: 10.1038/ki.2013.243. [DOI] [PubMed] [Google Scholar]
- 30.Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604–612. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Eagle KA, Lim MJ, Dabbous OH, et al. A validated prediction model for all forms of acute coronary syndrome: estimating the risk of 6-month postdischarge death in an international registry. JAMA. 2004;291(22):2727–2733. doi: 10.1001/jama.291.22.2727. [DOI] [PubMed] [Google Scholar]
- 32.White IR, Royston P, Wood AM.. Multiple imputation using chained equations: issues and guidance for practice. Stat Med. 2011;30(4):377–399. doi: 10.1002/sim.4067. [DOI] [PubMed] [Google Scholar]
- 33.Melvin A, O’Rahilly S, Savage DB.. Genetic syndromes of severe insulin resistance. Curr Opin Genet Dev. 2018;50:60–67. doi: 10.1016/j.gde.2018.02.002. [DOI] [PubMed] [Google Scholar]
- 34.Brown AE, Walker M.. Genetics of insulin resistance and the metabolic syndrome. Curr Cardiol Rep. 2016;18(8):75. doi: 10.1007/s11886-016-0755-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Ahmed B, Sultana R, Greene MW.. Adipose tissue and insulin resistance in obese. Biomed Pharmacother. 2021;137:111315. doi: 10.1016/j.biopha.2021.111315. [DOI] [PubMed] [Google Scholar]
- 36.Jiang J, Cai X, Pan Y, et al. Relationship of obesity to adipose tissue insulin resistance. BMJ Open Diabetes Res Care. 2020;8(1):e000741. doi: 10.1136/bmjdrc-2019-000741. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Saad MJA, Santos A, Prada PO.. Linking gut microbiota and inflammation to obesity and insulin resistance. Physiology. 2016;31(4):283–293. doi: 10.1152/physiol.00041.2015. [DOI] [PubMed] [Google Scholar]
- 38.Takeuchi T, Kubota T, Nakanishi Y, et al. Gut microbial carbohydrate metabolism contributes to insulin resistance. Nature. 2023;621(7978):389–395. doi: 10.1038/s41586-023-06466-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Cusi K, Maezono K, Osman A, et al. Insulin resistance differentially affects the PI 3-kinase– and MAP kinase–mediated signaling in human muscle. J Clin Invest. 2000;105(3):311–320. doi: 10.1172/jci7535. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Steinberg HO, Chaker H, Leaming R, et al. Obesity/insulin resistance is associated with endothelial dysfunction. Implications for the syndrome of insulin resistance. J Clin Invest. 1996;97(11):2601–2610. doi: 10.1172/jci118709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Foutz RM, Grimm PR, Sansom SC.. Insulin increases the activity of mesangial BK channels through MAPK signaling. Am J Physiol Renal Physiol. 2008;294(6):F1465–F1472. doi: 10.1152/ajprenal.00012.2008. [DOI] [PubMed] [Google Scholar]
- 42.Coward RJM, Welsh GI, Yang J, et al. The human glomerular podocyte is a novel target for insulin action. Diabetes. 2005;54(11):3095–3102. doi: 10.2337/diabetes.54.11.3095. [DOI] [PubMed] [Google Scholar]
- 43.Welsh GI, Hale LJ, Eremina V, et al. Insulin signaling to the glomerular podocyte is critical for normal kidney function. Cell Metab. 2010;12(4):329–340. doi: 10.1016/j.cmet.2010.08.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Tejada T, Catanuto P, Ijaz A, et al. Failure to phosphorylate AKT in podocytes from mice with early diabetic nephropathy promotes cell death. Kidney Int. 2008;73(12):1385–1393. doi: 10.1038/ki.2008.109. [DOI] [PubMed] [Google Scholar]
- 45.Sowers JR, Habibi J, Aroor AR, et al. Epithelial sodium channels in endothelial cells mediate diet-induced endothelium stiffness and impaired vascular relaxation in obese female mice. Metabolism. 2019;99:57–66. doi: 10.1016/j.metabol.2019.153946. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Tian X, Zuo Y, Chen S, et al. Triglyceride–glucose index is associated with the risk of myocardial infarction: an 11-year prospective study in the Kailuan cohort. Cardiovasc Diabetol. 2021;20(1):19. doi: 10.1186/s12933-020-01210-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Cheng Y, Fang Z, Zhang X, et al. Association between triglyceride glucose-body mass index and cardiovascular outcomes in patients undergoing percutaneous coronary intervention: a retrospective study. Cardiovasc Diabetol. 2023;22(1):75. doi: 10.1186/s12933-023-01794-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Wang Z, He H, Xie Y, et al. Non-insulin-based insulin resistance indexes in predicting atrial fibrillation recurrence following ablation: a retrospective study. Cardiovasc Diabetol. 2024;23(1):87. doi: 10.1186/s12933-024-02158-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Shen R, Lin L, Bin Z, et al. The U-shape relationship between insulin resistance-related indexes and chronic kidney disease: a retrospective cohort study from National Health and Nutrition Examination Survey 2007–2016. Diabetol Metab Syndr. 2024;16(1):168. doi: 10.1186/s13098-024-01408-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Druml W, Zajic P, Winnicki W, et al. Association of body mass index and outcome in acutely ill patients with chronic kidney disease requiring intensive care therapy. J Ren Nutr. 2020;30(4):305–312. doi: 10.1053/j.jrn.2019.09.006. [DOI] [PubMed] [Google Scholar]
- 51.Kalantar-Zadeh K, Rhee CM, Chou J, et al. The obesity paradox in kidney disease: how to reconcile it with obesity management. Kidney Int Rep. 2017;2(2):271–281. doi: 10.1016/j.ekir.2017.01.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Matsushita K, Kojima S, Hirakawa K, et al. Differential effects of overweight/obesity depending on the severity of heart failure complicating acute myocardial infarction in Japan. Prog Cardiovasc Dis. 2023;78:49–57. doi: 10.1016/j.pcad.2022.11.020. [DOI] [PubMed] [Google Scholar]
- 53.Wu S, Gao Y, Liu W, et al. The relationship between atherogenic index of plasma and plaque vulnerabilities: an optical coherence tomography study. Cardiovasc Diabetol. 2024;23(1):442. doi: 10.1186/s12933-024-02532-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Jiang S, Liu S, Xiao G, et al. Atherogenic index of plasma and the clinical outcome of patients with acute coronary syndrome: a meta-analysis. Ann Med. 2025;57(1):2442532. doi: 10.1080/07853890.2024.2442532. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Wang B, Jiang C, Qu Y, et al. Nonlinear association between atherogenic index of plasma and chronic kidney disease: a nationwide cross-sectional study. Lipids Health Dis. 2024;23(1):312. doi: 10.1186/s12944-024-02288-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Qian J, Chi Q, Qian C, et al. Atherogenic index of plasma and triglyceride-glucose index mediate the association between stroke and all-cause mortality: insights from the lipid paradox. Lipids Health Dis. 2025;24(1):173. doi: 10.1186/s12944-025-02586-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.Scirica BM, Mosenzon O, Bhatt DL, et al. Cardiovascular outcomes according to urinary albumin and kidney disease in patients with type 2 diabetes at high cardiovascular risk: observations from the SAVOR-TIMI 53 trial. JAMA Cardiol. 2018;3(2):155–163. doi: 10.1001/jamacardio.2017.4228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58.Wu X, Qiu W, Yang H, et al. Associations of the triglyceride-glucose index and atherogenic index of plasma with the severity of new-onset coronary artery disease in different glucose metabolic states. Cardiovasc Diabetol. 2024;23(1):76. doi: 10.1186/s12933-024-02163-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59.Whaley-Connell A, Sowers JR.. Insulin resistance in kidney disease: is there a distinct role separate from that of diabetes or obesity? Cardiorenal Med. 2017;8(1):41–49. doi: 10.1159/000479801. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 60.Barzilay JI, Farag YMK, Durthaler J.. Albuminuria: an underappreciated risk factor for cardiovascular disease. J Am Heart Assoc. 2024;13(2):e030131. doi: 10.1161/JAHA.123.030131. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61.Lian H, Wu H, Ning J, et al. The risk threshold for hemoglobin A1c associated with albuminuria: a population-based study in China. Front Endocrinol. 2021;12:673976. doi: 10.3389/fendo.2021.673976. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62.Ensling M, Steinmann W, Whaley-Connell A.. Hypoglycemia: a possible link between insulin resistance, metabolic dyslipidemia, and heart and kidney disease (the cardiorenal syndrome). Cardiorenal Med. 2011;1(1):67–74. doi: 10.1159/000322886. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 63.Janus A, Szahidewicz-Krupska E, Mazur G, et al. Insulin resistance and endothelial dysfunction constitute a common therapeutic target in cardiometabolic disorders. Mediators Inflamm. 2016;2016:3634948. doi: 10.1155/2016/3634948. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 64.Brown NJ. Contribution of aldosterone to cardiovascular and renal inflammation and fibrosis. Nat Rev Nephrol. 2013;9(8):459–469. doi: 10.1038/nrneph.2013.110. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All research data can be obtained from the corresponding author upon reasonable request.