Association Between the Triglyceride-Glucose Index and Clinical Outcomes Modified by Diabetes Duration in Patients with Acute Coronary Syndrome and Type 2 Diabetes Mellitus Authors: Yixuan Duan, Miaohan Qiu, Weifeng Liu, Di Cao, Yang Li, Haiwei Liu, Kai Xu, Xiaozeng Wang, Jing Li, Yi Li, Yaling Han Published: 2026 ABSTRACT BackgroundAn elevated triglyceride-glucose (TyG) index is strongly linked to an increased risk of cardiovascular events in patients with acute coronary syndrome (ACS) and type 2 diabetes mellitus (T2DM). However, the potential impact of increasing diabetes duration on the association between the TyG index and cardiovascular risk remains unclear.MethodsConsecutive patients admitted for ACS with T2DM received PCI between March 2016 and March 2020 were enrolled in the analysis. A total of 7093 patients were categorized into tertiles according to the TyG index and T2DM duration. The primary outcome was ischemic events, composed of cardiac death, myocardial infarction, and/or stroke at 5 years. The main secondary outcomes were 5-year cardiac death and all-cause death. Interaction terms and non-linear associations of the TyG index with the clinical outcomes were examined across diabetes durations.ResultsOver 5 years, 753 (10.6%) experienced ischemic events, 375 (5.3%) cardiac death, and 636 (9.0%) all-cause death. Higher TyG index and longer diabetes duration independently associated with all outcomes. Compared with the lowest TyG tertile, the highest tertile had adjusted hazard ratios (HRs) of 1.55 (95% CI: 1.29–1.88) for ischemic events and 1.23 (1.10–1.38) for all-cause death. The association between TyG index and ischemic events was strongest in the shortest duration group (HR: 1.43), with significant interaction for ischemic and cardiac death (both p < 0.05).ConclusionsA higher TyG index was independently associated with an increased risk of ischemic events and mortality among patients with ACS and T2DM. This association appeared to be attenuated with longer diabetes duration, suggesting that the prognostic relevance of the TyG index may vary according to the chronicity of diabetes. Background Type 2 diabetes mellitus (T2DM) remains a common complication in patients with acute coronary syndrome (ACS).1 Given the significant overlap in cardiometabolic risk factors, concomitant T2DM substantially worsens clinical outcomes in ACS patients.2,3 Consequently, identifying high-risk individuals and implementing sufficient management strategies are crucial to improving their prognosis. The triglyceride-glucose (TyG) index, which incorporates fasting triglyceride and plasma glucose levels, has been demonstrated to serve as a marker for identifying patients with ACS and/or T2DM who are at high risk of cardiovascular adverse events.4 Clinically, patients exhibit dynamic changes in glucose and lipid metabolism as T2DM advances, with primary potential mechanisms likely involving β-cell dysfunction, insulin resistance, hormonal imbalances, and suboptimal therapeutic management.5 In light of these dynamic changes, current guidelines advocate for integrating comprehensive risk assessment into the routine management strategies of these patients.6 Current management guidelines for T2DM with ACS recommend individualized HbA1c targets based on comorbidities, life expectancy, and diabetes duration.7 Nevertheless, the stability of TyG index’s predictive value for cardiovascular events at varying diabetes durations requ8 es Supplementary validation. We therefore aimed to evaluate the connection of TyG index-diabetes duration with the incidence of prolonged cardiometabolic complications, as well as their potential interaction effects, in a large-sample, real-world cohort of ACS patients with concomitant T2DM. Methods Study Cohort and Research Design This study is a retrospective, high-volume, real-world observational cohort comprising sequentially patients who underwent PCI at the General Hospital of Northern Theater Command between March 2016 and March 2020 (see Supplementary Figure 1). Eligible participants were adults (≥18 years) diagnosed with ACS and T2DM who survived hospitalization. Patients were excluded if information on diabetes duration, serum triglyceride, or fasting glucose was missing, making TyG index calculation infeasible. Ethical approval for this observational study was granted by the Institutional Review Board of the General Hospital of Northern Theater Command, with an exemption from obtaining individual patient consent. All procedures complied with the principles outlined in the Declaration of Helsinki. Biochemical Measurements and Data Acquisition Upon hospital admission, venous blood specimens were collected and processed according to established laboratory procedures. Quantitative analysis of lipid profiles and fasting blood glucose concentrations was performed employing the Beckman Coulter AU5800 automated analyzer (Beckman Coulter Inc., California, USA). The triglyceride-glucose index was derived using the validated equation: Ln[fasting triglyceride concentration (mg/dL) × fasting plasma glucose (mg/dL)/2], as previously described.9 The enrolled patients were categorized into tertiles (T1-T3) based on TyG index and DM duration, respectively. 5-Year Outcomes and Follow-Up The primary outcome in our study was ischemic events, composited of cardiac death, myocardial infarction (MI), and/or stroke at 5 years after discharge. The secondary outcomes encompassed 5-year all-cause mortality and the individual components of ischemic events. A standardized follow-up protocol was implemented, featuring scheduled evaluations at quarterly intervals (3, 6, 9 months) during the first year, transitioning to yearly assessments thereafter until completion of the 5-year study period through telephone interviews or in-person visits by trained research staff. An independent clinical events committee performed blinded verification of all reported outcomes. Analytical Methods Categorical variables were expressed as counts (percentages) and assessed with either the χ2 test or Fisher’s exact test, depending on the data distribution. Time-to-event data were analyzed using Kaplan-Meier survival curves, with comparisons made through log-rank tests. To evaluate the associations between the Triglyceride-Glucose (TyG) index, duration of diabetes, and clinical outcomes, Cox proportional hazards models were employed to calculate the hazard ratio (HR) and 95% confidence interval (CI). Multivariable models were adjusted for a range of potential confounders, including age, sex, hypertension, diabetes, history of myocardial infarction (MI), prior percutaneous coronary intervention (PCI), previous stroke, smoking status, ACS type, anemia, estimated glomerular filtration rate (eGFR), arterial access site, number of coronary arteries treated, stent count, and medications prescribed at discharge (aspirin, P2Y12 inhibitors, statins, ACE inhibitors/angiotensin receptor blockers, β-blockers, and proton pump inhibitors). Sensitivity analyses were performed to evaluate the robustness of the primary findings. First, a Fine-Gray competing-risk model was constructed with non-cardiovascular death treated as the competing event. Second, diabetes duration was additionally analyzed as a continuous variable to test its interaction with the TyG index. Incremental prognostic performance of the TyG index over the guideline-recommended GRACE risk score10 was evaluated by comparing the area under the receiver operating characteristic curve (AUC), continuous net reclassification improvement (NRI), and integrated discrimination improvement (IDI). Differences in AUCs were assessed using the DeLong test. To examine potential nonlinear relationships between the TyG index and clinical endpoints, including composite ischemic events, cardiac death, and all-cause mortality, restricted cubic spline (RCS) regression models were utilized. Effect modification analyses were conducted by introducing multiplicative interaction terms in proportional hazards models. The visreg package11 enabled graphical representation of interaction patterns, whereas the interactionRCS package12 provided quantitative estimates of duration-dependent TyG index effects. A significance threshold of P<0.05 (two-sided) was applied. Computational work utilized both SAS 9.4 (SAS Institute lnc., Cary, NC, USA) and R version 4.4.2 (http://www.rproject.org). Results Demographic and Clinical Characteristics Following the screening of 8277 consecutive individuals who fulfilled the inclusion criteria (see Supplementary Figure 1), a total of 7093 ACS patients with T2DM were included in the final analysis, with a follow-up rate of 92.1% at 5 years (median TyG index: 9.3 [IQR: 8.9–9.8]; median diabetes duration: 8.2 years [IQR: 3.4–13.1]). Baseline demographics, procedural characteristics, and discharge medications stratified by diabetes duration tertiles are summarized in Table 1. Patients in T3 group were older and less males, along with more proportion hypertension, history of stroke and MI. These patients also showed poorer renal function, as indicated by reduced eGFR. The procedural information was both comparable across T1-T3 groups of the TyG index and different durations.Table 1The Baseline Characteristics According to the T2DM DurationDuration of DiabetesP valueTertile 1<4.9 Years(N=2374)Tertile 24.9 to <10.5 Years(N=2363)Tertile 3≥ 10.5 Years(N=2356)Age, years59.3±10.461.6±9.664.5±8.6<0.0001Male, %1765 (74.3%)1664 (70.4%)1485 (63.0%)<0.0001Medical history, % Hypertension1625 (68.4%)1663 (70.4%)1712 (72.7%)0.0059 Previous MI506 (21.3%)544 (23.1%)576 (24.6%)0.0313 Previous stroke326 (13.7%)381 (16.2%)484 (20.6%)<0.0001 Previous PCI732 (30.8%)849 (35.9%)916 (38.9%)<0.0001Smoking, %<0.0001 Never1026 (43.4%)1134 (48.1%)1308 (55.7%) Active986 (41.7%)846 (35.9%)644 (27.4%) Former353 (14.9%)379 (16.1%)398 (16.9%)Presentation, %<0.0001 UA1567 (66.0%)1599 (67.7%)1701 (72.2%) NSTEMI372 (15.7%)395 (16.7%)334 (14.2%) STEMI435 (18.3%)369 (15.6%)321 (13.6%) eGFR, mL/min/1.73m298.7±29.294.5±26.888.1±29.6<0.0001 TyG, mmol/L9.4±0.79.4±0.79.4±0.70.6484 TG, mmol/L2.1±1.82.0±1.71.8±1.3<0.0001 LDL-c, mmol/L2.3±0.82.2±0.82.2±0.90.0256 Glucose, mmol/L9.4±3.910.0±4.110.6±4.6<0.0001 Anemia, %852 (35.9%)1034 (43.9%)1252 (53.2%)<0.0001 LVEF, %58.0±8.758.0±9.157.7±9.00.4302Procedure information Transradial access, %2192 (92.6%)2169 (91.8%)2133 (90.7%)0.0558 Coronary arteries treated, % LM124 (5.2%)114 (4.8%)126 (5.3%)0.6957 LAD1231 (51.9%)1237 (52.3%)1184 (50.3%)0.3226 LCX627 (26.4%)612 (25.9%)626 (26.6%)0.8608 RCA906 (38.2%)938 (39.7%)972 (41.3%)0.0942 Number of stents1.6±0.91.6±0.91.6±0.90.4331 Total length of stents, mm42.4±28.342.6±27.443.7±28.30.2278 Average stent diameters, mm2.8±0.92.7±0.82.7±0.90.0104Medical treatment at discharge, % Aspirin2315 (97.5%)2287 (96.8%)2286 (97.0%)0.3108 P2Y12 inhibitor<0.0001 Clopidogrel1440 (60.7%)1576 (66.7%)1652 (70.1%) Ticagrelor934 (39.3%)787 (33.3%)704 (29.9%) Statins2174 (91.6%)2162 (91.5%)2173 (92.2%)0.5991 ACEI/ARB1503 (63.3%)1443 (61.1%)1452 (61.6%)0.2538 βblockers1659 (69.9%)1616 (68.4%)1626 (69.0%)0.5354 PPI827 (34.8%)865 (36.6%)809 (34.3%)0.2299Antidiabetic therapy, % Insulins377 (15.9%)1009 (42.7%)1590 (67.5%)<0.0001 Oral Medications1817 (76.5%)1869 (79.1%)1775 (75.3%)0.0074Abbreviations: ACEI, angiotensin converting enzyme inhibitors; ARB, angiotensin II receptor blocker; eGFR, estimated glomerular filtration rate; LAD, left anterior descending; LCX, left circumflex; LDL-c, low-density lipoprotein-c; LM, left main; LVEF, left ventricular ejection fraction; MI, myocardial infarction; NSTEMI, non-ST-elevation myocardial infarction; TG, total cholesterol; TyG, triglyceride-glucose; PCI, percutaneous coronary intervention; PPI, proton pump inhibitors; RCA, right coronary artery; STEMI, ST-elevation myocardial infarction; UA, unstable angina. Table 1The Baseline Characteristics According to the T2DM DurationDuration of DiabetesP valueTertile 1<4.9 Years(N=2374)Tertile 24.9 to <10.5 Years(N=2363)Tertile 3≥ 10.5 Years(N=2356)Age, years59.3±10.461.6±9.664.5±8.6<0.0001Male, %1765 (74.3%)1664 (70.4%)1485 (63.0%)<0.0001Medical history, % Hypertension1625 (68.4%)1663 (70.4%)1712 (72.7%)0.0059 Previous MI506 (21.3%)544 (23.1%)576 (24.6%)0.0313 Previous stroke326 (13.7%)381 (16.2%)484 (20.6%)<0.0001 Previous PCI732 (30.8%)849 (35.9%)916 (38.9%)<0.0001Smoking, %<0.0001 Never1026 (43.4%)1134 (48.1%)1308 (55.7%) Active986 (41.7%)846 (35.9%)644 (27.4%) Former353 (14.9%)379 (16.1%)398 (16.9%)Presentation, %<0.0001 UA1567 (66.0%)1599 (67.7%)1701 (72.2%) NSTEMI372 (15.7%)395 (16.7%)334 (14.2%) STEMI435 (18.3%)369 (15.6%)321 (13.6%) eGFR, mL/min/1.73m298.7±29.294.5±26.888.1±29.6<0.0001 TyG, mmol/L9.4±0.79.4±0.79.4±0.70.6484 TG, mmol/L2.1±1.82.0±1.71.8±1.3<0.0001 LDL-c, mmol/L2.3±0.82.2±0.82.2±0.90.0256 Glucose, mmol/L9.4±3.910.0±4.110.6±4.6<0.0001 Anemia, %852 (35.9%)1034 (43.9%)1252 (53.2%)<0.0001 LVEF, %58.0±8.758.0±9.157.7±9.00.4302Procedure information Transradial access, %2192 (92.6%)2169 (91.8%)2133 (90.7%)0.0558 Coronary arteries treated, % LM124 (5.2%)114 (4.8%)126 (5.3%)0.6957 LAD1231 (51.9%)1237 (52.3%)1184 (50.3%)0.3226 LCX627 (26.4%)612 (25.9%)626 (26.6%)0.8608 RCA906 (38.2%)938 (39.7%)972 (41.3%)0.0942 Number of stents1.6±0.91.6±0.91.6±0.90.4331 Total length of stents, mm42.4±28.342.6±27.443.7±28.30.2278 Average stent diameters, mm2.8±0.92.7±0.82.7±0.90.0104Medical treatment at discharge, % Aspirin2315 (97.5%)2287 (96.8%)2286 (97.0%)0.3108 P2Y12 inhibitor<0.0001 Clopidogrel1440 (60.7%)1576 (66.7%)1652 (70.1%) Ticagrelor934 (39.3%)787 (33.3%)704 (29.9%) Statins2174 (91.6%)2162 (91.5%)2173 (92.2%)0.5991 ACEI/ARB1503 (63.3%)1443 (61.1%)1452 (61.6%)0.2538 βblockers1659 (69.9%)1616 (68.4%)1626 (69.0%)0.5354 PPI827 (34.8%)865 (36.6%)809 (34.3%)0.2299Antidiabetic therapy, % Insulins377 (15.9%)1009 (42.7%)1590 (67.5%)<0.0001 Oral Medications1817 (76.5%)1869 (79.1%)1775 (75.3%)0.0074Abbreviations: ACEI, angiotensin converting enzyme inhibitors; ARB, angiotensin II receptor blocker; eGFR, estimated glomerular filtration rate; LAD, left anterior descending; LCX, left circumflex; LDL-c, low-density lipoprotein-c; LM, left main; LVEF, left ventricular ejection fraction; MI, myocardial infarction; NSTEMI, non-ST-elevation myocardial infarction; TG, total cholesterol; TyG, triglyceride-glucose; PCI, percutaneous coronary intervention; PPI, proton pump inhibitors; RCA, right coronary artery; STEMI, ST-elevation myocardial infarction; UA, unstable angina. The Baseline Characteristics According to the T2DM Duration Abbreviations: ACEI, angiotensin converting enzyme inhibitors; ARB, angiotensin II receptor blocker; eGFR, estimated glomerular filtration rate; LAD, left anterior descending; LCX, left circumflex; LDL-c, low-density lipoprotein-c; LM, left main; LVEF, left ventricular ejection fraction; MI, myocardial infarction; NSTEMI, non-ST-elevation myocardial infarction; TG, total cholesterol; TyG, triglyceride-glucose; PCI, percutaneous coronary intervention; PPI, proton pump inhibitors; RCA, right coronary artery; STEMI, ST-elevation myocardial infarction; UA, unstable angina. As TyG tertiles increased (Supplementary Table 1), patients tended to be younger, more frequently female, current smokers, and had a higher incidence of acute myocardial infarction. Moreover, those in the highest tertile (T3) exhibited lower LVEF, a lower frequency of prior PCI, and a reduced prevalence of anemia (all p < 0.001). Notably, insulin use increased progressively across TyG tertiles, whereas the use of oral hypoglycemic agents did not differ significantly among the groups. Association Between TyG Tertiles and Clinical Outcomes The primary outcome of ischemic events occurred in 753 (10.6%) patients, of whom 375 (5.3%) had cardiac death, 185 (2.6%) had myocardial infarction, and 273 (3.8%) had stroke. The incidence of ischemic events was higher in the T3 TyG index group, compared with patients with other two groups (T1 vs T2 vs T3, 8.4% vs 11.5% vs 12.0%, respectively; p < 0.001) (Table 2, Supplementary Figure 2A). A similar trend was observed in the cardiac death (6.1% vs 8.3% vs 12.5%; P < 0.001) and all-cause death (7.8% vs 8.7% vs 10.3%; p = 0.016) (Table 2, Supplementary Figure 2B and 2C). Compared with T1 patients, those in T3 group had significantly increased risks of adverse outcomes. The adjusted HRs and 95% CIs for T3 vs T1 were 1.55 (1.29–1.88) for ischemic events (p < 0.0001), 1.52 (1.17–1.98) for cardiac death (p = 0.0018), and 1.40 (1.15–1.71) for all-cause death (p = 0.0010) (Table 2). When the TyG index was evaluated as a continuous variable, each 1-unit increase was independently associated with higher risks of ischemic events (adjusted HR 1.27; 95% CI, 1.15–1.41; p < 0.0001), cardiac death (adjusted HR 1.26; 95% CI, 1.09–1.46; p = 0.0022), and all-cause death (adjusted HR 1.23; 95% CI, 1.10–1.38; p = 0.0004) (Table 2).Table 2Cox Regression Analyses of TyG Index and Duration for Stratificating Clinical OutcomesTyG IndexDiabetes DurationEvents (%)Adjusted HR (95%CI)p valueEvents (%)Adjusted HR (95%CI)p valueIschemic EventsIschemic EventsTertiles 1198 (8.4%)Ref.-Tertiles 1200 (8.4%)Ref.-Tertiles 2271 (11.5%)1.38 (1.14–1.67)0.0009Tertiles 2230 (9.7%)1.04 (0.86–1.27)0.6787Tertiles 3284 (12.0%)1.55 (1.29–1.88)<0.0001Tertiles 3323 (13.7%)1.33 (1.11–1.61)0.0024TyG index per 1 unit increase-1.27 (1.15–1.41)<0.0001Duration per 1 year increase-1.02 (1.01–1.03)0.0001Cardiac DeathCardiac DeathTertiles 1102 (4.3%)Ref.-Tertiles 187 (3.7%)Ref.-Tertiles 2122 (5.2%)1.16 (0.88–1.53)0.2809Tertiles 2106 (4.5%)1.02 (0.76–1.36)0.9044Tertiles 3151 (6.4%)1.52 (1.17–1.98)0.0018Tertiles 3182 (7.7%)1.34 (1.02–1.75)0.0328TyG index per 1 unit increase-1.26 (1.09–1.46)0.0022Duration per 1 year increase-1.02 (1.01–1.04)0.0006MIMITertiles 144 (1.9%)Ref.-Tertiles 143 (1.8%)Ref.-Tertiles 269 (2.9%)1.49 (1.01–2.21)0.0467Tertiles 265 (2.8%)1.43 (0.96–2.13)0.0798Tertiles 372 (3.0%)1.72 (1.16–2.54)0.0067Tertiles 377 (3.3%)1.68 (1.14–2.50)0.0094TyG index per 1 unit increase-1.25 (1.02–1.53)0.0331Duration per 1 year increase-1.03 (1.00–1.05)0.0203StrokeStrokeTertiles 172 (3.0%)Ref.-Tertiles 183 (3.5%)Ref.-Tertiles 2110 (4.7%)1.54 (1.13–2.10)0.0059Tertiles 288 (3.7%)1.03 (0.75–1.40)0.8764Tertiles 391 (3.8%)1.38 (1.00–1.90)0.0523Tertiles 3102 (4.3%)1.23 (0.91–1.67)0.1849TyG index per 1 unit increase-1.20 (1.01–1.42)0.0391Duration per 1 year increase-1.01 (0.99–1.03)0.178All-cause DeathAll-cause DeathTertiles 1185 (7.8%)Ref.-Tertiles 1146 (6.1%)Ref.-Tertiles 2206 (8.7%)1.09 (0.89–1.34)0.4129Tertiles 2196 (8.3%)1.11 (0.89–1.38)0.336Tertiles 3245 (10.3%)1.40 (1.15–1.71)0.0010Tertiles 3294 (12.5%)1.34 (1.09–1.65)0.0058TyG index per 1 unit increase-1.23 (1.10–1.38)0.0004Duration per 1 year increase-1.02 (1.01–1.03)0.0011Notes: Model adjusted for age, gender, hypertension, diabetes, previous myocardial infarction, previous percutaneous coronary intervention, previous stroke, smoking, type of ACS, anemia, eGFR, arterial access, coronary arteries treated, number of stents, and medical treatment (aspirin, P2Y12 inhibitor, statin, ACEI/ARB, β blockers, PPI and antidiabetic therapy).Abbreviations: CI, confidence interval; HR: hazard ratio; MI, myocardial infarction. TyG, triglyceride-glucose. Table 2Cox Regression Analyses of TyG Index and Duration for Stratificating Clinical OutcomesTyG IndexDiabetes DurationEvents (%)Adjusted HR (95%CI)p valueEvents (%)Adjusted HR (95%CI)p valueIschemic EventsIschemic EventsTertiles 1198 (8.4%)Ref.-Tertiles 1200 (8.4%)Ref.-Tertiles 2271 (11.5%)1.38 (1.14–1.67)0.0009Tertiles 2230 (9.7%)1.04 (0.86–1.27)0.6787Tertiles 3284 (12.0%)1.55 (1.29–1.88)<0.0001Tertiles 3323 (13.7%)1.33 (1.11–1.61)0.0024TyG index per 1 unit increase-1.27 (1.15–1.41)<0.0001Duration per 1 year increase-1.02 (1.01–1.03)0.0001Cardiac DeathCardiac DeathTertiles 1102 (4.3%)Ref.-Tertiles 187 (3.7%)Ref.-Tertiles 2122 (5.2%)1.16 (0.88–1.53)0.2809Tertiles 2106 (4.5%)1.02 (0.76–1.36)0.9044Tertiles 3151 (6.4%)1.52 (1.17–1.98)0.0018Tertiles 3182 (7.7%)1.34 (1.02–1.75)0.0328TyG index per 1 unit increase-1.26 (1.09–1.46)0.0022Duration per 1 year increase-1.02 (1.01–1.04)0.0006MIMITertiles 144 (1.9%)Ref.-Tertiles 143 (1.8%)Ref.-Tertiles 269 (2.9%)1.49 (1.01–2.21)0.0467Tertiles 265 (2.8%)1.43 (0.96–2.13)0.0798Tertiles 372 (3.0%)1.72 (1.16–2.54)0.0067Tertiles 377 (3.3%)1.68 (1.14–2.50)0.0094TyG index per 1 unit increase-1.25 (1.02–1.53)0.0331Duration per 1 year increase-1.03 (1.00–1.05)0.0203StrokeStrokeTertiles 172 (3.0%)Ref.-Tertiles 183 (3.5%)Ref.-Tertiles 2110 (4.7%)1.54 (1.13–2.10)0.0059Tertiles 288 (3.7%)1.03 (0.75–1.40)0.8764Tertiles 391 (3.8%)1.38 (1.00–1.90)0.0523Tertiles 3102 (4.3%)1.23 (0.91–1.67)0.1849TyG index per 1 unit increase-1.20 (1.01–1.42)0.0391Duration per 1 year increase-1.01 (0.99–1.03)0.178All-cause DeathAll-cause DeathTertiles 1185 (7.8%)Ref.-Tertiles 1146 (6.1%)Ref.-Tertiles 2206 (8.7%)1.09 (0.89–1.34)0.4129Tertiles 2196 (8.3%)1.11 (0.89–1.38)0.336Tertiles 3245 (10.3%)1.40 (1.15–1.71)0.0010Tertiles 3294 (12.5%)1.34 (1.09–1.65)0.0058TyG index per 1 unit increase-1.23 (1.10–1.38)0.0004Duration per 1 year increase-1.02 (1.01–1.03)0.0011Notes: Model adjusted for age, gender, hypertension, diabetes, previous myocardial infarction, previous percutaneous coronary intervention, previous stroke, smoking, type of ACS, anemia, eGFR, arterial access, coronary arteries treated, number of stents, and medical treatment (aspirin, P2Y12 inhibitor, statin, ACEI/ARB, β blockers, PPI and antidiabetic therapy).Abbreviations: CI, confidence interval; HR: hazard ratio; MI, myocardial infarction. TyG, triglyceride-glucose. Cox Regression Analyses of TyG Index and Duration for Stratificating Clinical Outcomes Notes: Model adjusted for age, gender, hypertension, diabetes, previous myocardial infarction, previous percutaneous coronary intervention, previous stroke, smoking, type of ACS, anemia, eGFR, arterial access, coronary arteries treated, number of stents, and medical treatment (aspirin, P2Y12 inhibitor, statin, ACEI/ARB, β blockers, PPI and antidiabetic therapy). Abbreviations: CI, confidence interval; HR: hazard ratio; MI, myocardial infarction. TyG, triglyceride-glucose. Association Between Diabetes Duration Tertiles and Clinical Outcomes Patients with higher diabetes duration had a relatively higher risk of ischemic events (T3 vs T1: adjusted HR, 1.33, 95% CI, 1.11–1.61, p =0.0024) (Table 2). The same trend can be found in cardiac death and all-cause death (T3 vs T1: adjusted HR, 1.34, 95% CI,1.02-1.75, p =0.0328; T3 vs T1: adjusted HR,1.34, 95% CI 1.09–1.65, p =0.0058, respectively). When analyzed as a continuous variable, as the per 1 unit of diabetes duration increase, the risk of ischemic events (adjusted HR, 1.02, 95%CI, 1.01–1.03, p =0.0001), cardiac death (adjusted HR, 1.02, 95%CI, 1.01–1.04, p =0.0006) and all-cause death (adjusted HR, 1.02, 95%CI, 1.01–1.03, p =0.0004) significantly increased (Table 2). The interaction between the TyG index and ischemic events remained statistically significant after accounting for competing risks (p = 0.0001), consistent with the findings from the conventional Cox proportional hazards model (Supplementary Table 2). The Interaction Effect Between the Diabetes Duration and Clinical Outcomes Ischemic event risk demonstrated a monotonic increase with ascending TyG index levels in each diabetes duration stratification (T1: adjusted HR, 1.43, 95%CI, 1.16–1.76, p =0.0007, T2, adjusted HR, 1.24, 95%CI, 1.03–1.50, p =0.0201, and T3, adjusted HR, 1.22, 95%CI, 1.05–1.43, p = 0.0111). However, the HRs in the T1 group were significantly higher than those in the T2 and T3 groups of diabetes duration (p for interaction = 0.0418). Moreover, the HRs of TyG index for the group with T1 of diabetes duration (adjusted HR, 1.92, 95%CI, 1.41–2.61, p <0.0001) were significantly higher than the HRs for the T2 and T3 groups (adjusted HR, 1.15, 95%CI, 0.87–1.51, p =0.3420, and adjusted HR, 1.15, 95%CI, 0.93–1.42, p =0.2103, respectively) for death from cardiac cause (p for interaction = 0.0055) (Table 3). The interaction between TyG index and diabetes duration remained statistically significant after the sensitivity analysis, confirming the robustness of our primary findings.Table 3Association Between Tertiles of Duration and OutcomesDuration of DiabetesTertiles 1Tertiles 2Tertiles 3p for Interaction†p for Interaction#Adjusted HR (95%CI)*p valueAdjusted HR (95%CI)*p valueAdjusted HR (95%CI)*p valueIschemic events1.43 (1.16–1.76)0.00071.24 (1.03–1.50)0.02011.22 (1.05–1.43)0.01110.04180.0287Cardiac death1.92 (1.41–2.61)<0.00011.15 (0.87–1.51)0.34201.15 (0.93–1.42)0.21030.00550.0040MI1.32 (0.86–2.03)0.20241.46 (1.03–2.08)0.03261.13 (0.82–1.56)0.44490.19900.4163Stroke0.96 (0.68–1.35)0.82641.24 (0.93–1.67)0.14901.33 (1.01–1.75)0.03970.54590.5904All-cause death1.68 (1.32–2.15)<0.00011.03 (0.83–1.28)0.77661.21 (1.02–1.43)0.02820.18460.0580Notes: *The TyG index per 1 unit increase. †Duration of diabetes as continuous variable. # Duration of diabetes as categorized variable. Model adjusted for age, gender, hypertension, diabetes, previous myocardial infarction, previous percutaneous coronary intervention, previous stroke, smoking, type of ACS, anemia, eGFR, arterial access, coronary arteries treated, number of stents, and medical treatment (aspirin, P2Y12 inhibitor, statin, ACEI/ARB, β blockers, PPI and antidiabetic therapy).Abbreviations: CI, confidence interval; HR, hazard ratio; MI, myocardial infarction. TyG: triglyceride-glucose. Table 3Association Between Tertiles of Duration and OutcomesDuration of DiabetesTertiles 1Tertiles 2Tertiles 3p for Interaction†p for Interaction#Adjusted HR (95%CI)*p valueAdjusted HR (95%CI)*p valueAdjusted HR (95%CI)*p valueIschemic events1.43 (1.16–1.76)0.00071.24 (1.03–1.50)0.02011.22 (1.05–1.43)0.01110.04180.0287Cardiac death1.92 (1.41–2.61)<0.00011.15 (0.87–1.51)0.34201.15 (0.93–1.42)0.21030.00550.0040MI1.32 (0.86–2.03)0.20241.46 (1.03–2.08)0.03261.13 (0.82–1.56)0.44490.19900.4163Stroke0.96 (0.68–1.35)0.82641.24 (0.93–1.67)0.14901.33 (1.01–1.75)0.03970.54590.5904All-cause death1.68 (1.32–2.15)<0.00011.03 (0.83–1.28)0.77661.21 (1.02–1.43)0.02820.18460.0580Notes: *The TyG index per 1 unit increase. †Duration of diabetes as continuous variable. # Duration of diabetes as categorized variable. Model adjusted for age, gender, hypertension, diabetes, previous myocardial infarction, previous percutaneous coronary intervention, previous stroke, smoking, type of ACS, anemia, eGFR, arterial access, coronary arteries treated, number of stents, and medical treatment (aspirin, P2Y12 inhibitor, statin, ACEI/ARB, β blockers, PPI and antidiabetic therapy).Abbreviations: CI, confidence interval; HR, hazard ratio; MI, myocardial infarction. TyG: triglyceride-glucose. Association Between Tertiles of Duration and Outcomes Notes: *The TyG index per 1 unit increase. †Duration of diabetes as continuous variable. # Duration of diabetes as categorized variable. Model adjusted for age, gender, hypertension, diabetes, previous myocardial infarction, previous percutaneous coronary intervention, previous stroke, smoking, type of ACS, anemia, eGFR, arterial access, coronary arteries treated, number of stents, and medical treatment (aspirin, P2Y12 inhibitor, statin, ACEI/ARB, β blockers, PPI and antidiabetic therapy). Abbreviations: CI, confidence interval; HR, hazard ratio; MI, myocardial infarction. TyG: triglyceride-glucose. Although a higher TyG index was associated with poor prognosis in most subgroups stratified by diabetes duration, no nonlinear relationship was observed (Figure 1). Figure 1Restricted cubic splines for the association between the TyG index and clinical outcomes across tertiles of diabetes durations: ischemic events (A) T1 diabetes duration; (D) T2 diabetes duration; (G) T3 diabetes duration, cardiac death (B) T1 diabetes duration; (E) T2 diabetes duration; (H) T3 diabetes duration, and all-cause death (C) T1 diabetes duration; (F) T2 diabetes duration; (I) T3 diabetes duration.The image A showing a spline line graph of ischemic events. The x-axis label is, Triglyceride Glucose Index, from 8.0 to 11.0 with ticks at 8.0, 9.0, 9.3, 10.0, 11.0. The y-axis label is, Hazard ratio of ischemic events, from 0 to 5 with ticks at 0, 1, 2, 3, 4, 5. A vertical dashed line is at 9.3 and a horizontal dashed line is at hazard ratio 1. The curve is near 1 around 9.3, below 1 near 8.0 and rises to about 2 by 11.0. Text: p value for overall equals 0.004; p value for nonlinearity equals 0.671. The image B showing a spline line graph of cardiac death. The x-axis label is, Triglyceride Glucose Index, 8.0 to 11.0 with ticks at 8.0, 9.0, 9.3, 10.0, 11.0. The y-axis label is, Hazard ratio of cardiac death, 0 to 5 with ticks 0 to 5. Dashed reference lines at 9.3 and 1. The curve is below 1 near 8.0, near 1 at 9.3 and increases steeply to about 4 by 11.0. Text: p value for overall less than 0.001; p value for nonlinearity equals 0.978. The image C showing a spline line graph of all cause death. The x-axis label is, Triglyceride Glucose Index, 8.0 to 11.0 with ticks at 8.0, 9.0, 9.3, 10.0, 11.0. The y-axis label is, Hazard ratio of all cause death, 0 to 5 with ticks 0 to 5. Dashed reference lines at 9.3 and 1. The curve is below 1 near 8.0, near 1 at 9.3 and rises to about 3 by 11.0. Text: p value for overall less than 0.001; p value for nonlinearity equals 0.861. The image D showing a spline line graph of ischemic events. The x-axis label is, Triglyceride Glucose Index, 8.0 to 11.0 with ticks at 8.0, 9.0, 9.3, 10.0, 11.0. The y-axis label is, Hazard ratio of ischemic events, 0 to 5 with ticks 0 to 5. Dashed reference lines at 9.3 and 1. The curve is near 1 across the range, slightly below 1 near 8.0 and slightly above 1 near 11.0, reaching about 1.3 at 11.0. Text: p value for overall equals 0.068; p value for nonlinearity equals 0.689. The image E showing a spline line graph of cardiac death. The x-axis label is, Triglyceride Glucose Index, 8.0 to 11.0 with ticks at 8.0, 9.0, 9.3, 10.0, 11.0. The y-axis label is, Hazard ratio of cardiac death, 0 to 5 with ticks 0 to 5. Dashed reference lines at 9.3 and 1. The curve stays close to 1, with a shallow dip below 1 near 9.0 and a mild rise above 1 toward 11.0, reaching about 1.2 at 11.0. Text: p value for overall equals 0.629; p value for nonlinearity equals 0.827. The image F showing a spline line graph of all cause death. The x-axis label is, Triglyceride Glucose Index, 8.0 to 11.0 with ticks at 8.0, 9.0, 9.3, 10.0, 11.0. The y-axis label is, Hazard ratio of all cause death, 0 to 5 with ticks 0 to 5. Dashed reference lines at 9.3 and 1. The curve is near 1 throughout, with a slight rise after 9.3 and about 1.1 to 1.2 near 11.0. Text: p value for overall equals 0.934; p value for nonlinearity equals 0.816. The image G showing a spline line graph of ischemic events. The x-axis label is, Triglyceride Glucose Index, 8.0 to 11.0 with ticks at 8.0, 9.0, 9.3, 10.0, 11.0. The y-axis label is, Hazard ratio of ischemic events, 0 to 5 with ticks 0 to 5. Dashed reference lines at 9.3 and 1. The curve is below 1 near 8.0, near 1 at 9.3 and rises gradually to about 1.5 at 11.0. Text: p value for overall equals 0.020; p value for nonlinearity equals 0.153. The image H showing a spline line graph of cardiac death. The x-axis label is, Triglyceride Glucose Index, 8.0 to 11.0 with ticks at 8.0, 9.0, 9.3, 10.0, 11.0. The y-axis label is, Hazard ratio of cardiac death, 0 to 5 with ticks 0 to 5. Dashed reference lines at 9.3 and 1. The curve remains close to 1, slightly below 1 near 8.0 to 9.0 and slightly above 1 near 11.0, around 1.2 at 11.0. Text: p value for overall equals 0.443; p value for nonlinearity equals 0.738. The image I showing a spline line graph of all cause death. The x-axis label is, Triglyceride Glucose Index, 8.0 to 11.0 with ticks at 8.0, 9.0, 9.3, 10.0, 11.0. The y-axis label is, Hazard ratio of all cause death, 0 to 5 with ticks 0 to 5. Dashed reference lines at 9.3 and 1. The curve is near 1 at 9.3, below 1 near 8.0 and rises to about 1.4 at 11.0. Text: p value for overall equals 0.056; p value for nonlinearity equals 0.236.Nine restricted cubic splines of TyG index over different diabetes duration for ischemic events, cardiac death and all cause death. Restricted cubic splines for the association between the TyG index and clinical outcomes across tertiles of diabetes durations: ischemic events (A) T1 diabetes duration; (D) T2 diabetes duration; (G) T3 diabetes duration, cardiac death (B) T1 diabetes duration; (E) T2 diabetes duration; (H) T3 diabetes duration, and all-cause death (C) T1 diabetes duration; (F) T2 diabetes duration; (I) T3 diabetes duration. An interaction effect was visualized between the TyG index and clinical outcomes modified by diabetes duration considered as a categorical variable (tertiles) and continuous variable in Figures 2 and 3, respectively. The association between TyG index and ischemic events remained statistically significant across the entire range of diabetes duration, but the HR gradually decreases with the duration of the disease (p for interaction = 0.0418, Figure 3A). The relationship between TyG index and cardiac death also showed a decrease trend (p for interaction = 0.0055) in Figure 3B. In Figure 3C, the association between the TyG index and all-cause mortality demonstrated an inverse J-shaped pattern with increasing duration of diabetes (p for interaction = 0.1846). Figure 2Interaction analysis of the association between the TyG index and clinical outcomes across tertiles of diabetes durations. The 3 lines indicate tertiles of diabetes duration (< 4.9 years, 4.9 years to 10.5 years, and ≤ 10.5 years, respectively). (A) Ischemic events. (B) Cardiac death. (C) All-cause death. Model adjusted for age, gender, hypertension, diabetes, previous myocardial infarction, previous percutaneous coronary intervention, previous stroke, smoking, type of ACS, anemia, eGFR, arterial access, coronary arteries treated, number of stents, and medical treatment (aspirin, P2Y12 inhibitor, statin, ACEI/ARB, β blockers, PPI and antidiabetic therapy).The dataset includes 12 line graphs categorized by 3 outcomes and 4 diabetes duration groups, analyzing the Triglyceride Glucose Index (7-12) and hazard ratios for ischemic events, cardiac death and all-cause death. Each graph shows tertiles of diabetes duration: Tertile 1 (<4.9 years), Tertile 2 (4.9-10.5 years) and Tertile 3 (less than or equal to 10.5 years). Graph 1 (Ischemic events) lines rise from 0.2-0.6 at index 7 to 1.6-2.0 at index 12 (P interaction = 0.0400). Graph 2 (Cardiac death) lines rise from 0.1-0.6 to 2.0-4.0 (P interaction = 0.0049). Graph 3 (All-cause death) lines rise from 0.2-0.7 to 1.8-2.1 (P interaction = 0.1705). Graphs 4-6 show T1 duration effects, lines rising from 0.2-0.3 to 2.0-4.0. Graphs 7-9 show T2 duration effects, lines rising from 0.2-0.3 to 1.8-3.0. Graphs 10-12 show T3 duration effects, lines rising from 0.4-0.6 to 1.6-2.5. Shaded bands indicate variability.12 line graphs: Interaction analysis of the association between the TyG index and clinical outcomes across tertiles of diabetes durations. Figure 3Effect of TyG index on clinical outcomes across the range of diabetes duration. (A) Ischemic events. (B) Cardiac death. (C) All-cause death. Model adjusted for age, gender, hypertension, diabetes, previous myocardial infarction, previous percutaneous coronary intervention, previous stroke, smoking, type of ACS, anemia, eGFR, arterial access, coronary arteries treated, number of stents, and medical treatment (aspirin, P2Y12 inhibitor, statin, ACEI/ARB, β blockers, PPI and antidiabetic therapy).Image A: Ischemic events show a decreasing Hazard Ratio (HR) from 1.45 at 0 years to 1.02 at 30 years, with a p interaction of 0.0418. The 95% CI is widest at 0 years (1.1 to 1.8) and narrows around 10-20 years (1.05 to 1.35), widening again near 30 years (0.7 to 1.3). Image B: Cardiac death HR decreases from 1.8 at 0 years to 0.95 at 30 years, with a p interaction of 0.0055. The 95% CI spans 1.2 to 2.5 at 0 years, narrows to 0.9 to 1.3 around 20 years and is 0.5 to 1.3 at 30 years. Image C: All-cause death HR decreases from 1.5 at 0 years to 1.15 by 15 years, remaining stable through 30 years, with a p interaction of 0.1846. The 95% CI is 1.05 to 2.0 at 0 years, narrows to 1.0 to 1.3 around 15-20 years and is 0.75 to 1.5 at 30 years.A set of three line graphs showing hazard ratio of TyG index across the range of diabetes duration for ischemic events and deaths. Interaction analysis of the association between the TyG index and clinical outcomes across tertiles of diabetes durations. The 3 lines indicate tertiles of diabetes duration (< 4.9 years, 4.9 years to 10.5 years, and ≤ 10.5 years, respectively). (A) Ischemic events. (B) Cardiac death. (C) All-cause death. Model adjusted for age, gender, hypertension, diabetes, previous myocardial infarction, previous percutaneous coronary intervention, previous stroke, smoking, type of ACS, anemia, eGFR, arterial access, coronary arteries treated, number of stents, and medical treatment (aspirin, P2Y12 inhibitor, statin, ACEI/ARB, β blockers, PPI and antidiabetic therapy). Effect of TyG index on clinical outcomes across the range of diabetes duration. (A) Ischemic events. (B) Cardiac death. (C) All-cause death. Model adjusted for age, gender, hypertension, diabetes, previous myocardial infarction, previous percutaneous coronary intervention, previous stroke, smoking, type of ACS, anemia, eGFR, arterial access, coronary arteries treated, number of stents, and medical treatment (aspirin, P2Y12 inhibitor, statin, ACEI/ARB, β blockers, PPI and antidiabetic therapy). Incremental Predictive Value of the TyG Index Beyond the GRACE Risk Score As shown in Supplementary Table 3, the addition of the TyG index significantly improved discrimination for ischemic events, increasing the AUC from 0.63 (95% CI: 0.61–0.66) to 0.65 (95% CI: 0.63–0.67; p = 0.0026). Significant improvements were also observed in risk reclassification (NRI = 0.14, 95% CI: 0.06–0.22, p = 0.0003) and discrimination (IDI = 0.33%, 95% CI: 0.17%–0.50%, p < 0.0001). For all-cause mortality, although the increase in AUC was modest (0.72 vs 0.72, p = 0.0373), the TyG index still significantly improved risk reclassification (NRI = 0.14, 95% CI: 0.06–0.23, p = 0.0005) and discrimination (IDI = 0.26%, 95% CI: 0.09%–0.44%, p = 0.0029). Discussion This study utilized data from a large-scale real-world registry to investigate the association between the TyG index and the risks of ischemic events and mortality among ACS patients with diabetes undergoing PCI. Our findings were as follows: 1) Elevated TyG index is associated with an increased risk of ischemic events, cardiac death, and all-cause death in patients with ACS and T2DM; 2) This association was significantly attenuated with increasing duration of for ischemic events and cardiac death, but not in all-cause mortality in ACS patients with DM. The TyG index, serves as a surrogate marker for insulin resistance and has shown superior performance compared to homeostasis model assessment of insulin resistance in various populations.13 The TyG index shows good prognostic predictive value in different diabetes populations, such as young people, middle-aged and elderly people, and sudden cardiac arrest patients, and our study provides new evidence for the value of TyG index in ACS patients with different diabetes durations.8,14,15 The TyG index has demonstrated associations with cardiovascular risk factors and cerebrovascular disease, such as stroke, hypertension, carotid plaque, kidney injury, and moreover, as a strong predictor of coronary artery stenosis, coronary artery calcification, and the prevalence of multivessel disease.16,17 Moreover, emerging evidence suggests a close relationship between elevated TyG levels and metabolic dysfunction-associated steatotic liver disease (MASLD), a condition affecting more than half of patients with T2DM. MASLD is increasingly recognized not merely as a hepatic manifestation of metabolic dysfunction but as a multisystem disease characterized by insulin resistance, chronic low-grade inflammation, oxidative stress, mitochondrial dysfunction, and atherogenic dyslipidemia.18,19 In the present study, the association between TyG index and ischemic event and cardiac death varies in patients with different durations of T2DM. During initial-stage diabetes, insulin resistance remains clinically significant, while end-organ damage and irreversible vascular remodeling are not yet fully established, such as end-stage kidney disease, heart failure hospitalization, diabetic eye disease, myocardial infarction, and lower-extremity amputation.20 This may render the TyG index a more sensitive indicator of ongoing atherogenic activity at this stage. With prolonged disease duration, patients with diabetes often exhibit improved medication adherence, better compliance with dietary recommendations, and benefit from the introduction of novel first-line agents such as DDP-4 inhibitors, GLP-1 agonists, etc20–22 These factors contribute to significantly improved clinical outcomes, which may partially attenuate the association between the TyG index and adverse events. In our study, TyG index provides incremental prognostic information beyond the GRACE score, particularly by improving risk reclassification despite relatively modest gains in overall discrimination. Although current treatments may mitigate the existing metabolic burden, persistent metabolic dysregulation can exert lasting effects on cardiovascular health.23 Following the prolong of diabetes durations, β-cell function progressively declines, while insulin resistance worsens over time.24 Metabolic disorders such as T2DM and obesity are often accompanied by prolonged hypertriglyceridemia.25 Progressive diabetes duration correlates with worsening metabolic dysregulation, poor glycemic control, oxidative stress, and prothrombotic states become more prevalent.26 Aging-related insulin resistance further worsens these metabolic abnormalities, contributing to endothelial dysfunction, inflammation, and platelet activation, which collectively elevate cardiovascular risk.27 In early-stage diabetes, insulin resistance plays a dominant role in promoting endothelial dysfunction, inflammation, oxidative stress, and atherosclerotic progression. Consequently, TyG serves as a stronger surrogate marker of cardiometabolic risk. In contrast, among patients with long- duration diabetes, accumulated vascular damage, advanced glycation end-products (AGEs), chronic inflammation, and irreversible structural vascular changes may become predominant determinants of cardiovascular risk,28 thereby attenuating the relative contribution of insulin resistance reflected by TyG. The differential prognostic value of TyG across diabetes duration may reflect the evolving pathophysiology of cardiovascular risk in diabetes. With longer diabetes duration, cumulative hyperglycemic exposure promotes the formation of AGEs, which contribute to endothelial dysfunction, vascular stiffening, oxidative stress, and chronic inflammation through AGE-RAGE signaling.29 Consequently, cardiovascular risk in long-standing diabetes may be increasingly driven by glycation-related vascular damage rather than insulin resistance alone, potentially attenuating the predictive value of TyG. Interestingly, although the TyG index levels were similar across different DM durations in our cohort, patients with shorter disease duration tended to have higher triglyceride levels, while those with longer disease duration showed elevated blood pressure. This finding suggests the need for tailored management strategies across the diabetes continuum. For instance, in ACS patients with newly diagnosed DM, β-cell function may still be preserved, and thus, therapeutic strategies targeting TyG index should not only focus on glycemic control but also emphasize intensive lipid management.30 Conversely, in patients with long-standing diabetes, although low-density lipoprotein cholesterol levels were comparable among groups, triglyceride levels remained significantly different, underscoring the importance of comprehensive lipid regulation beyond low-density lipoprotein cholesterol, particularly in patients with prolonged diabetes duration. Previous studies have consistently shown that a higher TyG index is strongly associated with an increased risk of diabetes and cardiovascular disease.31 Our study reveals that T3 of TyG index was associated with the greatest risk of cardiovascular outcomes, including ischemic events and all-cause death. This is consistent with prior research showing that elevated TyG index significantly increased the risk of cardiovascular events in individuals with diabetes (HR = 1.5).32 A large cohort study found that TyG was independently associated with MACE and all-cause mortality in patients with T2DM after PCI, even after comprehensive adjustment for confounding variables. Importantly, these associations persisted across sex, age, and glycemic control strata, highlighting the robust nature of this index. Our study shows that patients with ACS with different stages of diabetes have different risks of time to adverse outcomes with the increase of TyG index. The risk of ischemic events was significantly increased in the T3 group except MI, the T2 group was mainly associated with the increase in MI, and the T1 group was reflected in cardiac death and all-cause death events. In previous studies from the Improving Care for Cardiovascular Disease in China cohort and a Korean registry, the proportion of statin use among ACS patients with diabetes was approximately 18% and 40%, respectively, while the use of antidiabetic therapies was 50% and around 70%.33 In contrast, over 90% of patients in our cohort received lipid-lowering therapy, and the rate of glucose-lowering treatment was slightly higher. Such interventional heterogeneity may account for the attenuation of expected nonlinear TyG index-outcome associations. From a clinical perspective, these findings have meaningful implications. First, the TyG index, given its simplicity and cost-effectiveness, can be readily implemented in routine clinical practice for early risk identification.34 Second, our study supports a stratified approach to cardiovascular risk assessment in T2DM patients with ACS, emphasizing the importance of considering diabetes duration. For instance, in patients with mid-stage diabetes, who appear most vulnerable, intensive secondary prevention strategies such as earlier use of GLP-1 receptor agonists may be particularly beneficial. Third, incorporating TyG into existing risk scores (eg, GRACE or OPT-CAD score) may improve their predictive accuracy for long-term events post-PCI.35 Our results affirm the prognostic association between TyG index and clinical outcomes including ischemic events and all-cause death, and critically, demonstrate that its prognostic strength varies significantly across different durations of T2DM. This heterogeneity underscores the need for a more nuanced risk stratification framework incorporating metabolic markers alongside disease duration. A major strength of our study is that the use of a large, real-world cohort with long-term follow-up enhances generalizability and highly adherent participants. Moreover, the stratification by diabetes duration and use of RCS modeling allow for a nuanced understanding of non-linear associations. Comprehensive multivariable adjustments were performed to minimize confounding. Our findings highlight the prognostic utility of the TyG index in ACS patients with T2DM undergoing PCI, and more importantly, illustrate that its prognostic association value is significantly influenced by diabetes duration. These results advocate for an individualized risk assessment paradigm that incorporates both metabolic and disease duration factors to optimize secondary prevention strategies in this high-risk population. The current study design presents certain inherent limitations that need be addressed. First, the retrospective nature of this single-institution study, though utilizing a sizeable PCI cohort of ACS/T2DM patients, may constrain generalizability of findings to broader populations. Therefore, these results should be interpreted cautiously and require validation in further well-designed studies. Second, as a dynamic metabolic marker, TyG may change over time owing to glycemic control, lipid-lowering therapy, lifestyle modifications, disease progression, and other clinical interventions. Consequently, a single measurement may not fully reflect cumulative metabolic exposure during the 5-year follow-up period. The lack of serial assessments prevented evaluation of temporal changes in TyG and may have resulted in exposure misclassification, potentially leading to an underestimation of the true association between long-term insulin resistance burden and cardiovascular risk. Third, information on lipid-lowering therapy was collected at discharge and analyzed as a baseline covariate. Longitudinal data regarding changes in treatment intensity, adherence, or the initiation of additional lipid-lowering agents during follow-up were unavailable. Given that lipid-lowering therapies, particularly statins, may influence triglyceride levels and cardiovascular risk over time, residual confounding related to treatment modifications cannot be excluded. Future studies incorporating time-varying treatment data are warranted to further clarify the relationship between the TyG index and cardiovascular outcomes. Lastly, unmeasured factors such as dietary habits, adherence to medication, and glycemic fluctuations may have influenced the outcomes but were not included in the analysis. Conclusions Among patients with ACS and T2DM, a higher TyG index was independently associated with adverse clinical outcomes, including ischemic events and mortality. These associations appeared to be attenuated with increasing diabetes duration, suggesting that the clinical relevance of the TyG index may vary across different stages of diabetes. Further studies are warranted to explore potential interactions between the TyG index and other patient characteristics in relation to cardiovascular outcomes.