RESUMO
BACKGROUND: Carotid endarterectomy (CEA) and carotid artery stenting (CAS) are effective interventions for treating extracranial carotid artery stenosis (ECAS), but long-term prognosis is limited by postoperative restenosis. Carotid restenosis is defined as carotid stenosis >50% by various examination methods in patients after carotid revascularization. This retrospective cohort study examined the value of the triglyceride-glucose (TyG) index for predicting vascular restenosis after carotid revascularization. METHODS: A total of 830 patients receiving CEA (408 cases, 49.2%) or CAS (422 cases, 50.8%) were included in this study. Patients were stratified into three subgroups according to TyG index tertile (high, intermediate, and low), and predictive value for restenosis was evaluated by constructing multivariate Cox proportional hazard regression models. RESULTS: Incidence of postoperative restenosis was significantly greater among patients with a high TyG index according to univariate analysis. Kaplan-Meier survival curve analysis revealed a progressive increase in restenosis prevalence with rising TyG index. Multivariate Cox regression models also identified TyG index as an independent predictor of restenosis, while receiver operating characteristic (ROC) curve analysis showed that TyG index predicted restenosis with moderate sensitivity (57.24%) and specificity (67.99%) (AUC: 0.619, 95% CI 0.585-0.652, z-statistic=4.745, p<0.001). Addition of the TyG index to an established risk factor model incrementally improved restenosis prediction (AUC: 0.684 (0.651-0.715) vs 0.661 (0.628-0.694), z-statistic =2.027, p = 0.043) with statistical differences. CONCLUSION: The TyG index is positively correlated with vascular restenosis risk after revascularization, which can be used for incremental prediction and has certain predictive value.