China has the largest number of diabetic patients in the world. Insulin resistance is a major pathogenetic mechanism and cause of type 2 diabetes. Recent studies have suggested potential relationship between triglyceride-glucose index (TyG) and its derivatives index and the development of type 2 diabetes. However, most available evidence is from cross-sectional study or longitudinal study, and there is a lack of studies in the northeast China.

To explore the association between TyG and its derivatives index called triglyceride glucose-body mass index (TyG-BMI) and the incidence risk of type 2 diabetes, analyze the predictive capacity of TyG and TyG-BMI for type 2 diabetes, so as to provide a scientific basis for the early screening of high risk population with type 2 diabetes.

A total of 6843 cases of physical examination in the health management center of the Second Affiliated Hospital of Dalian Medical University from January 2015 to December 2017 were included as research subjects. The baseline data of their physical examinations were collected. By the use of nested case-control study method, 209 patients with new-onset type 2 diabetes during follow-up from January 2018 to April 2021 were selected as the case group; 418 cases among those without new-onset type 2 diabetes, endocrine system diseases and malignancies were selected as the control group after 1∶2 matching by propensity score according to the same gender and age ±2 years. TyG and TyG-BMI were calculated based on baseline data. The Cox regression fitted conditional Logistic regression model was used to analyze the relationship between TyG, TyG-BMI and the development of type 2 diabetes. The restricted cubic spline regression model was used to analyze the dose-response relationship between TyG, TyG-BMI and type 2 diabetes. The receiver operating characteristic (ROC) curve was used to analyze the predictive value of TyG and TyG-BMI for type 2 diabetes.

BMI, AC, SBP, DBP, FPG, TG, TC, LDL-C, TyG, and TyG-BMI in the case group were higher than the control group, and HDL-C in the case group was lower than the control group (P<0.05) . Quartile Q1 group of baseline TyG: TyG<8.46, 157 cases; Quartile Q2 group of baseline TyG: 8.46≤TyG<8.83, 157 cases; quartile Q3 group of baseline TyG: 8.83≤TyG<9.19, 156 cases; Quartile Q4 group of baseline TyG: TyG≥9.19, 157 cases. The risk of type 2 diabetes in quartile Q2 group of baseline TyG, quartile Q3 group of baseline TyG and quartile Q4 group of baseline TyG was 1.57 〔95%CI (0.92, 2.70) 〕, 2.07 〔95%CI (1.21, 3.53) 〕and 3.18 〔95%CI (1.76, 5.75) 〕 times higher than quartile Q1 group of baseline TyG (P_trend<0.001) , respectively. The risk of type 2 diabetes in quartile Q2 group of baseline TyG, quartile Q3 group of baseline TyG and quartile Q4 group of baseline TyG was 2.21 times 〔95%CI (1.25, 3.94) 〕, 2.92 times 〔95%CI (1.58, 5.37) 〕, and 5.34 times 〔95%CI (2.39, 11.95) 〕 higher than quartile Q1 group of baseline TyG (P_trend< 0.001) , respectively. The association between continuous changes in TyG, TyG-BMI and type 2 diabetes showed a linear dose-response relationship (non-linear test, P>0.05) , with an increasing shape in the dose-response relationship. The incidence risk of type 2 diabetes gradually increased when TyG and TyG-BMI were higher than 8.838 and 229.364, respectively. The area under the ROC curve (AUC) for TyG and TyG-BMI to predict type 2 diabetes was 0.696 〔95%CI (0.658, 0.732) 〕and 0.725〔95%CI (0.688, 0.760) 〕with the optimal cut-offs of 8.650 and 224.859, respectively.

Increased TyG and TyG-BMI levels are independent risk factors for type 2 diabetes, showing a linear dose-response relationship with type 2 diabetes, both of which have predictive value for type 2 diabetes. TyG-BMI may be a better prediction index considering the strength of risk association, AUC, and clinical impact of screening results.