Triglyceride–glucose index‑integrated nomogram for six‑month outcomes after acute ST-elevation myocardial infarction

Risk stratification after ST‑elevation myocardial infarction (STEMI) treated with percutaneous coronary intervention (PCI) remains challenging, particularly for identifying residual risk beyond established clinical scores. The triglyceride–glucose (TyG) index, a surrogate marker of insulin resistance, may provide additional prognostic information. To develop and internally validate a TyG‑integrated nomogram for 6‑month major adverse cardiovascular events (MACE) after index hospital admission in STEMI patients undergoing PCI. In this single‑centre ambispective cohort, 169 consecutive STEMI patients treated with primary PCI were enrolled. The primary endpoint was time to first MACE within 6 months after index hospital admission, defined as all‑cause death, non‑fatal myocardial infarction, target‑vessel revascularisation, hospitalisation for unstable angina or heart failure, or stroke/transient ischaemic attack. Candidate predictors were analysed using Cox regression with elastic‑net penalisation. The final model included TyG index, GRACE score, hemoglobin, and left ventricular ejection fraction (LVEF). Apparent and bootstrap optimism-corrected model performance were assessed. Thirty-nine of 169 patients (23.1%) experienced a first MACE within 6 months after index hospital admission. Compared with the TyG-excluded clinical model containing GRACE score, hemoglobin, and LVEF, addition of TyG improved global model fit (likelihood-ratio χ2 = 8.36, df = 1, p = 0.0038; Akaike information criterion, 314.6 to 308.3) and reclassification (integrated discrimination improvement, 0.038; continuous net reclassification improvement, 0.574), but did not improve apparent or optimism-corrected C-index discrimination. The full model had an apparent C-index of 0.859 (95% confidence interval, 0.799–0.920), an optimism-corrected C-index of 0.848, and an optimism-corrected calibration slope of 0.923. A nomogram combining the TyG index with GRACE score, hemoglobin, and LVEF was associated with improved global model fit and reclassification relative to a TyG-excluded clinical model, but did not improve C-index discrimination. External validation in larger, contemporary cohorts is needed before clinical implementation.