BMC Cardiovascular Disorders
Differences in the triglyceride–glucose index across echocardiographic phenotypes and its association with obstructive coronary artery disease in patients with diabetes aged < 60 years
By Peiyan Peng, Zexin Zhang, Zhuangxiong Wu and 1 others.
Abstract
The triglyceride–glucose (TyG) index is a surrogate marker of insulin resistance, whereas global longitudinal strain (GLS) and left atrial volume index (LAVI) reflect subclinical myocardial dysfunction and chronic left ventricular filling-pressure burden, respectively. This study compared the TyG index across echocardiographic phenotypes and examined its association with obstructive coronary artery disease (CAD) in patients with type 2 diabetes mellitus aged < 60 years. This single-center retrospective study included 171 patients with type 2 diabetes mellitus who underwent elective coronary angiography for suspected stable CAD between May 2024 and June 2026. Patients were classified into four echocardiographic phenotypes according to GLS and LAVI: phenotype A (normal GLS and normal LAVI), phenotype B (normal GLS and increased LAVI), phenotype C (abnormal GLS and normal LAVI), and phenotype D (abnormal GLS and increased LAVI). Obstructive CAD was defined as ≥ 50% diameter stenosis in at least one major epicardial coronary artery. Multivariable logistic regression and receiver operating characteristic analyses were performed. The TyG index differed significantly among the four phenotypes ( P < 0.001) and was highest in phenotype D. Of the 171 patients, 53 had obstructive CAD and 118 had non-obstructive CAD. Compared with patients with non-obstructive CAD, those with obstructive CAD had a higher TyG index, a larger LAVI, less negative GLS, and a higher prevalence of phenotype D (all P < 0.05). In multivariable analysis, the TyG index (odds ratio [OR], 3.183; 95% confidence interval [CI], 1.039–9.747; P = 0.043) and phenotype D (OR, 2.188; 95% CI, 1.036–4.630; P = 0.040) were independently associated with obstructive CAD. The TyG index yielded an area under the curve (AUC) of 0.714, whereas the model combining the TyG index and phenotype D yielded an AUC of 0.730. The increase in AUC was not statistically significant (ΔAUC = 0.017; 95% CI, − 0.036 to 0.069; P = 0.534). In patients with type 2 diabetes mellitus aged < 60 years, a higher TyG index and phenotype D were independently associated with obstructive CAD. Although the combination of the TyG index and phenotype D showed numerically greater discrimination than the TyG index alone, a significant incremental improvement was not demonstrated.






