Triglyceride-glucose index and visceral fat area in the prediction of diabetic microvascular disease: A county-level real-world study in China.

Abstract

This study aimed to assess whether visceral fat area (VFA) enhances prediction of diabetic microvascular disease (DMV) beyond the triglyceride-glucose (TyG) index in a county-level population of individuals with type 2 diabetes mellitus. A cross-sectional study was conducted among 1042 adults with type 2 diabetes between 2022 and 2025. DMV was defined as the presence of diabetic retinopathy, chronic kidney disease, or diabetic peripheral neuropathy. Multivariable logistic regression analysis, receiver operating characteristic curve analysis, decision curve analysis, and calibration plots were applied to evaluate model performance. The predictive value of the TyG index was first assessed, after which the incremental contribution of VFA was quantified using net reclassification improvement (NRI) and integrated discrimination improvement (IDI). After adjustment for established risk factors, each 1-unit increase in TyG was associated with 31% higher odds of DMV (OR 1.31, 95% CI 1.04-1.65). VFA was not independently associated with DMV (OR 1.00, 95% CI 0.99-1.00). The area under the receiver operating characteristic curve (AUC) was 0.740 for the TyG-plus-covariates model and 0.741 for the TyG-plus-VFA-plus-covariates model. The categorical NRI was 0.8% (p = 0.310), and the IDI was 0.04% (p = 0.258). Subgroup analyses for diabetic retinopathy, chronic kidney disease, and diabetic peripheral neuropathy Subgroup analyses for diabetic retinopathy, chronic kidney disease, and diabetic peripheral neuropathy yielded consistent results: categorical NRI, IDI, and ΔAUC 95% CIs all crossed zero. Sensitivity analyses supported the primary findings. The TyG index was independently associated with diabetic microvascular disease, whereas the addition of visceral fat area provided negligible incremental predictive value beyond established risk factors.