Correlation between triglyceride-glucose index and atrial fibrillation in acute coronary syndrome patients: a retrospective cohort study and the establishment of a LASSO-Logistic regression model.

Background: Insulin resistance (IR) is an independent predictor of atrial fibrillation (AF), but the specific utility of the triglyceride-glucose (TyG) index as a predictive marker for the incidence of AF in the acute coronary syndrome (ACS) population has not yet been explored.

Objective: To explore the correlation between TyG index and the risk of AF in ACS patients and to establish a predictive model.

Methods: A retrospective study was conducted on 613 ACS patients admitted to the Department of Cardiovascular Medicine at the First Teaching Hospital of Tianjin University of Traditional Chinese Medicine from January 2022 to September 2024. Patients were divided into four groups based on quartiles of TyG index. Patients were further divided into two groups based on the occurrence of AF: the AF group and the non-AF group. Patient information was collected through the hospital's HIS system. Variable selection was completed using LASSO regression algorithms. Multivariate logistic bidirectional stepwise regression analysis was used to explore the correlation between the TyG index and the risk of AF in ACS patients and to construct a regression model. Three different models were constructed by adjusting for confounding factors and restricted cubic spline plots were drawn to validate the significance of the TyG index combined with AF further. The predictive value of the LASSO-multivariate logistic bidirectional stepwise regression model and the TyG index alone for predicting AF in ACS patients was analyzed using the receiver operating characteristic curve.

Results: The LASSO-multivariate logistic bidirectional stepwise regression algorithm showed that coronary heart disease (CHD), valvular heart disease (VHD), TyG, age (AGE), and diastolic blood pressure (DBP) were risk factors for AF in ACS. The restricted cubic spline model demonstrated a significant linear relationship between a higher TyG index and an increased risk of AF in the ACS patient population. The area under the curve (AUC) for predicting AF in ACS patients using the TyG index and the LASSO-multivariate logistic bidirectional stepwise regression model was 0.65(95%CI = 0.58 ~ 0.73) and 0.71(95%CI = 0.65 ~ 0.77) respectively. Additionally, the correlation between the TyG index and AF was consistent across different subgroups.

Conclusion: In ACS patients, the TyG index is a stable and independent predictor of AF, with specific clinical value in identifying the occurrence of AF in this population.