TyG指数对冠状动脉支架内再狭窄的预测价值及机器学习预测模型构建

The predictive value of the TyG index for coronary in-stent restenosis and the construction of a machine learning prediction model

摘要

目的 探讨甘油三酯-葡萄糖（TyG）指数对冠状动脉支架内再狭窄（ISR）的预测价值，并基于TyG指数及多维临床特征构建机器学习预测模型。方法 回顾性纳入2021年1月至2024年12月于福建医科大学附属龙岩第一医院心内科的270例首次接受冠状动脉支架置入术患者资料，根据随访造影结果将其分为ISR组（ n = 62）和非ISR组（ n = 208）。计算TyG指数，采用单因素及多因素Logistic回归分析ISR的危险因素，绘制受试者操作特征（ROC）曲线评估TyG指数的预测效能；采用Logistic回归（LR）、随机森林（RF）、极限梯度提升算法（XGBoost）、轻量级梯度提升算法(LightGBM)、支持向量机（SVM）5种机器学习模型，经5折交叉验证评估预测性能，并以SHAP方法解释最优模型。结果 ISR发生率为23.0%（62/270）。ISR组TyG指数高于非ISR组（9.17±0.42 vs. 8.82±0.45， P ＜ 0.001）。多因素Logistic回归分析显示TyG指数是ISR的独立危险因素（OR = 5.641，95%CI：2.428 ～ 13.108， P ＜ 0.001）。ROC曲线示TyG指数预测ISR的曲线下面积（AUC）为0.708，最佳截断值为9.111，灵敏度为62.9%，特异度为73.1%。5种机器学习模型中LR表现最优（AUC = 0.841±0.041），SHAP分析证实TyG指数是模型中重要的预测特征。结论 TyG指数是冠状动脉支架术后ISR的独立危险因素，具有良好的预测价值。基于多维临床特征构建的机器学习模型可进一步提升ISR预测效能。

Abstract

Objective This study aimed to evaluate the predictive value of the triglyceride-glucose (TyG) index for coronary in-stent restenosis (ISR) and develop a machine learning model integrating TyG with multidimensional clinical features. Methods Two hundred and seventy patients who underwent coronary stent implantation in Department of Cardiovascular Medicine, Longyan First Affiliated Hospital of Fujian Medical University from January 2021 to December 2024 were enrolled. They were categorized into the ISR ( n = 62) and non-ISR ( n = 208) groups based on follow-up angiography. The TyG index was calculated. Risk factors for ISR were identified using univariate and multivariate Logistic regression, and the predictive performance of TyG was assessed using ROC curves. Five machine learning algorithms [Logistic regression(LR), random Forest(RF), extreme gradient boosting(XGBoost), light gradient boosting machine(LightGBM), support vector machine(SVM)] were trained and validated using 5-fold cross-validation. Model interpretability was achieved via SHAP analysis. Results The ISR incidence was 23.0% (62/270). Patients with ISR exhibited a significantly higher TyG index compared to controls (9.17±0.42 vs. 8.82±0.45, P ＜ 0.001). Multivariable analysis identified the TyG index as an independent predictor of ISR ( OR = 5.641, 95% CI: 2.428 ～ 13.108, P ＜ 0.001). The TyG index yielded an AUC of 0.708 (cutoff: 9.111; sensitivity: 62.9%; specificity: 73.1%). Among the machine learning models, LR demonstrated superior performance (AUC = 0.841±0.041), with SHAP analysis highlighting the TyG index as a critical predictive variable. Conclusions The TyG index serves as an independent risk factor and predictor for ISR following coronary stenting. Integrating TyG into machine learning models based on multidimensional clinical data significantly improves ISR prediction accuracy.

关键词

TyG指数 / 支架内再狭窄 / 机器学习 / SHAP

Key words

TyG index / In-stent restenosis / Machine learning / SHAP

参考文献

[1] The Writing Committee of the Report on Cardiovascular Health and Diseases in China 2024. Report on Cardiovascular Health and Diseases in China 2024：prevalence status of cardiovascular diseases and related risk factors[J]. Chin J Cardiovasc Res， 2025， 23（9）： 771-793. DOI: 10.3969/j.issn.1672-5301.2025.09.001.

[2] Cassese S， Byrne R A， Tada T， et al. Incidence and predictors of restenosis after coronary stenting in 10 004 patients with surveillance angiography[J]. Heart， 2014， 100（2）： 153-159. DOI: 10.1136/heartjnl-2013-304933.

[3] Ge J B， Xu Y J， Wang C. Internal Medicine[M]. 9th ed. Beijing： People’s Medical Publishing House， 2018： 221-235.

[4] XU S， QIAO Y， TANG C. Research progress on risk factors for coronary in-stent restenosis[J]. J Southeast Univ（Med Sci Ed）， 2023， 42（3）： 475-480. DOI: 10.3969/j.issn.1671-6264.2023.03.022.

[5] Pelliccia F， Zimarino M， Niccoli G， et al. In-stent restenosis after percutaneous coronary intervention：emerging knowledge on biological pathways[J]. Eur Heart J Open， 2023， 3（5）： oead083. DOI: 10.1093/ehjopen/oead083.

[6] Bai H， Zhang B， Sun Y， et al. Research advances in the etiology of in-stent restenosis of coronary arteries[J]. Front Cardiovasc Med， 2025， 12： 1585102. DOI: 10.3389/fcvm.2025.1585102.

[7] Liu J， Bursill C . When stents become plaques：in-stent neoatherosclerosis[J]. Nat Rev Cardiol， 2026， 23（2）： 85. DOI: 10.1038/s41569-025-01241-z.

[8] Fan Y， Yan Z， Li T， et al. Primordial drivers of diabetes heart disease：comprehensive insights into insulin resistance[J]. Diabetes Metab J， 2024， 48（1）： 19-36. DOI: 10.4093/dmj.2023.0110.

[9] Liao J， Wang L， Duan L， et al. Insulin resistance surrogates are associated with all-cause mortality and cardiovascular mortality in population with metabolic syndrome：a retrospective cohort study of NHANES[J]. Sci Rep， 2025， 15（1）： 4706. DOI: 10.1038/s41598-025-88296-7.

[10] Landowska M， Kałuża B， Watała C， et al. Is insulin resistance an independent predictor of atherosclerosis?[J]. J Clin Med， 2025， 14（3）： 969. DOI: 10.3390/jcm14030969.

[11] Gounden V， Devaraj S， Jialal I . The role of the triglyceride-glucose index as a biomarker of cardio-metabolic syndromes[J]. Lipds Heal Dis， 2024， 23（1）： 416. DOI: 10.1186/s12944-024-02412-6.

[12] Li C， Zhao S， Li Y， et al. Triglyceride-glucose index predicts adverse cardiovascular events in patients with H-type hypertension combined with coronary heart disease：a retrospective cohort study[J]. Cardiovasc Diabetol， 2025， 24（1）： 45. DOI: 10.1186/s12933-024-02555-x.

[13] Miao Y， Wang Y， Wan Q . Association between TyG index with obesity indicators and coronary heart disease：a cohort study[J]. Sci Rep， 2025， 15： 8920. DOI: 10.1038/s41598-025-93922-5.

[14] Zhao Q N， Wang Y T， Zhang H M， et al. Comparative analysis of six surrogate insulin resistance indices for comparison of the predictive value of insulin resistance substitution index for cardiovascular disease risk in adult patients with cardiovascular-kidney-metabolic syndrome 0-3 stage[J]. J Sun Yat Sen Univ（Med Sci）， 2026， 47（3）： 539-550. DOI: 10.11714/jsysu.med.YX20260005.

[15] Jia H， Zhang W， Jia S， et al. Association between dynamic changes in the triglyceride-glucose index and prognosis in patients with acute ST-segment elevation myocardial infarction[J]. Cardiovasc Diabetol， 2025， 24（1）： 357. DOI: 10.1186/s12933-025-02858-7.

[16] Xi Y， Wang H， Sun N . Machine learning outperforms traditional logistic regression and offers new possibilities for cardiovascular risk prediction：a study involving 143,043 Chinese patients with hypertension[J]. Front Cardiovasc Med， 2022， 9： 1025705. DOI: 10.3389/fcvm.2022.1025705.

[17] Alahdab F， El Shawi R， Ahmed A I， et al. Improving prognostic risk assessment of cardiovascular events with machine learning：an evaluation using positron emission tomography myocardial perfusion imaging[J]. J Nucl Cardiol， 2026， 57： 102539. DOI: 10.1016/j.nuclcard.2025.102539.

[18] Kim J， Park S， Lee H， et al. Explainable machine learning models for coronary artery disease diagnosis using Shapley value analysis to identify key risk factors[J]. BMC Med Inform Decis Mak， 2024， 24： 68. DOI: 10.1186/s12911-024-02417-3.

[19] Patel R， Desai A， Mehta N， et al. Leveraging interpretable machine learning with SHAP to uncover modifiable risk factors for cardiovascular complications in patients with metabolic syndrome[J]. J Biomed Inform， 2025， 149： 104587. DOI: 10.1016/j.jbi.2024.104587.

[20] Zhang L， Chen X， Wu D， et al. Comparison of machine learning algorithms for predicting in-stent restenosis after percutaneous coronary intervention：a real-world cohort study[J]. BMC Med Inform Decis Mak， 2021， 21（1）： 301. DOI: 10.1186/s12911-021-01667-8.

[21] Topol E J， Rajcomar S， Harrell F E， et al. Machine Learning for Cardiovascular Risk Prediction：Methodological Advances and Clinical Applications[J]. Circulation， 2023， 147（22）： 1689-1705. DOI: 10.1161/CIRCULATIONAHA.122.062368.

[22] Jiang H， Liu Y， Guo H， et al. The association between the triglyceride-glucose index and in-stent restenosis in patients undergoing percutaneous coronary intervention：a systematic review and meta-analysis[J]. BMC Cardiovasc Disord， 2024， 24（1）： 234. DOI: 10.1186/s12872-024-03903-1.

[23] Mao J， Fang Z， Jiang S， et al. A positive linear correlation between the triglyceride-glucose index and in-stent restenosis after percutaneous coronary intervention in patients with coronary heart disease[J]. Front Cardiovasc Med， 2025， 12： 1544125. DOI: 10.3389/fcvm.2025.1544125.

[24] Taskaev D， Taskaeva I， Mazunina I， et al. Oxidized triglycerides and oxidized ldl：each playing their own role in atherosclerosis[J]. Lipids Health Dis， 2025， 24： 211. DOI: 10.1186/s12944-025-02587-8.

[25] Peng X， Wu H . Inflammatory links between hypertriglyceridemia and atherogenesis[J]. Curr Atheroscler Rep， 2022， 24（5）： 297-306. DOI: 10.1007/s11883-022-01006-w.

[26] Li S， Guo B， Chen H， et al. The role of the triglyceride（triacylglycerol）glucose index in the development of cardiovascular events：a retrospective cohort analysis[J]. Sci Rep， 2019， 9（1）： 7320. DOI: 10.1038/s41598-019-43776-5.

[27] Ormazabal V， Nair S， Elfeky O， et al. Association between insulin resistance and the development of cardiovascular disease[J]. Cardiovasc Diabetol， 2018， 17（1）： 122. DOI: 10.1186/s12933-018-0762-4.

[28] Liu J， Pan S， Wang X， et al. Role of advanced glycation end products in diabetic vascular injury：molecular mechanisms and therapeutic perspectives[J]. Eur J Med Res， 2023， 28（1）： 553. DOI: 10.1186/s40001-023-01431-w.

[29] Poteryaeva O N， Usynin I F . Molecular mechanisms of the regulatory action of high-density lipoproteins on the endothelial function[J]. Biomed Khim， 2024， 70（4）： 206-217. DOI: 10.18097/PBMC20247004206.

[30] Madaudo C， Bono G， Ortello A， et al. Dysfunctional high-density lipoprotein cholesterol and coronary artery disease：a narrative review[J]. JPM， 2024， 14（9）： 996. DOI: 10.3390/jpm14090996.

[31] Yiu J Y T， Hally K E， Larsen P D， et al. Neutrophil-enriched biomarkers and long-term prognosis in acute coronary syndrome：a systematic review and meta-analysis[J]. J Cardiovasc Transl Res， 2024， 17（2）： 426-447. DOI: 10.1007/s12265-023-10425-2.

[32] Gao H D， Kuang Y Y， Liu Y B， et al. Changes of plasma Rap1A levels in patients with in-stent restenosis after percutaneous coronary intervention and the underlying mechanisms[J]. J Cent South Univ(Med Sci)， 2023， 48（11）： 1650-1658. DOI: 10.11817/j.issn.1672-7347.2023.230285.

[33] Liu J H， Liu J. Interpretation of the highlights and key points of the 2024 revised edition of the chinese guidelines for the management of hypertension[J]. Chin J Front Med Sci（Electronic Ed）， 2025， 17（ 1）： 1-6. DOI: 10.12037/YXQY.2025.01-01.

[34] Islam K， Islam R， Nguyen I， et al. Diabetes mellitus and associated vascular disease：pathogenesis，complications，and evolving treatments[J]. Adv Ther， 2025， 42（6）： 2659-2678. DOI: 10.1007/s12325-025-03185-9.

[35] Wang W D . Research on Cardiovascular Disease Prediction Based on Machine Learning[D]. Wuhan： Central China Normal University， 2025.

[36] Qi X， Wang S， Fang C， et al. Machine learning and SHAP value interpretation for predicting comorbidity of cardiovascular disease and cancer with dietary antioxidants[J]. Redox Biol， 2025， 79： 103470. DOI: 10.1016/j.redox.2024.103470.

[37] Zhang X， Lin S， Zeng Q， et al. Machine learning and SHAP value interpretation for predicting cardiovascular disease risk in patients with diabetes using dietary antioxidants[J]. Front Nutr， 2025， 12： 1612369. DOI: 10.3389/fnut.2025.1612369.

基金资助

福建省卫健委科技计划项目(2023QNA101)