摘要OBJECTIVES To assess the correlation between triglyceride glucose (TyG) index and in-hospital mortality in patients with ST-segment elevation myocardial infarction (STEMI). METHODS A total of 2190 patients with STEMI who underwent primary angiography within 12 h from symptom onset were selected from the prospective, nationwide, multicenter CAMI registry. TyG index was calculated with the formula: Ln [fasting triglycerides (mmol/L) × fasting glucose (mmol/L)/2]. Patients were divided into three groups according to the tertiles of TyG index. The primary endpoint was in-hospital mortality. RESULTS Overall, 46 patients died during hospitalization, in-hospital mortality was 1.5％, 2.2％, 2.6％ for tertile 1, tertile 2, and tertile 3, respectively. However, TyG index was not significantly correlated with in-hospital mortality in single-variable logistic regression analysis. Nonetheless, after adjusting for age and sex, TyG index was significantly associated with higher mortality when regarded as a continuous variable (adjusted OR = 1.75, 95％ CI: 1.16-2.63) or categorical variable (tertile 3 vs. tertile 1: adjus-ted OR = 2.50, 95％ CI: 1.14-5.49). Furthermore, TyG index, either as a continuous variable (adjusted OR = 2.54, 95％ CI: 1.42-4.54) or categorical variable (tertile 3 vs. tertile 1: adjusted OR = 3.57, 95％ CI: 1.24-10.29), was an independent predictor of in-hospital mortality after adjusting for multiple confounders in multivariable logistic regression analysis. In subgroup analysis, the pro-gnostic effect of high TyG index was more significant in patients with body mass index < 18.5 kg/m2 (Pinteraction = 0.006). CONCLUSIONS This study showed that TyG index was positively correlated with in-hospital mortality in STEMI patients who underwent primary angiography, especially in underweight patients.