Abstract
PURPOSE: The triglyceride glucose (TyG) index and TyG-related indicators have been proposed as a marker of insulin resistance. It is unclear which is the best indicator to predict diabetes mellitus (DM) in Chinese. This study aimed to investigate the predictive value of different biomarkers for the incidence of DM. PATIENTS AND METHODS: Between January 2017 and December 2020, 5575 subjects who underwent health examinations in Hebei General Hospital were retrospectively included. The primary endpoint was new onset DM. RESULTS: During a median follow-up of 3.03 years, 133(2.39%) individuals developed DM. Multivariable cox proportional hazards models revealed that TyG index and TyG-related parameters were positively associated with DM risk. As the interaction analyses showed, there were significant interactions with sex and age levels in relation to DM risk (both P for interaction <0.05). Risk prediction for DM was significantly improved by adding TyG index to the baseline model using conventional diabetic risk factors in predicting DM at follow-up. CONCLUSION: This population-based cohort study suggested a causal relationship between TyG index and DM after adjusting for other confounding factors. This independent and significant association was more apparent in females and subjects younger than 65 years. Compared with the TyG-BMI, TyG-WC, TyG-WHtR, the TyG index was a more effective predictor of DM.