Objective To investigate the association and predictive value of the triglyceride-glucose-waist circumference index (TyG-WC) with the risk of diabetes mellitus (DM). Methods Based on the preliminary data from the "Early Screening and Comprehensive Intervention Project for High-Risk Populations of Cardiovascular and Cerebrovascular Diseases" in Hainan Province, a multi-stage stratified cluster sampling method was used to select research subjects aged 35-75 years from eight project sites during February 2023 to January 2024. Data were collected through questionnaires, physical examinations, and laboratory tests. A multivariate logistic regression model was employed to analyze the association strength between TyG-WC and diabetes (odds ratio OR and 95% confidence interval CI). A restricted cubic spline (RCS) model was used to assess the dose-response relationship. Stratified analyses by sex and age were conducted to evaluate the heterogeneity in the association between TyG-WC and diabetes risk. Receiver operating characteristic (ROC) curves were used to compare the predictive performance of the triglyceride-glucose-waist circumference index (TyG-WC), triglyceride-glucose index (TyG), and waist circumference (WC). The DeLong test was applied to compare the differences in the ROC curves between TyG, WC, and TyG-WC to determine whether the differences in predictive efficacy were statistically significant. Results A total of 6,148 participants were ultimately screened, with 5,843 included in the analysis. The mean age of male participants was 59.68 ± 9.99 years, and that of female participants was 57.79 ± 10.57 years. The overall prevalence of diabetes was 15.15%, and the age-standardized prevalence was 12.90%.. After adjusting for confounding factors, the risk of DM in the highest quartile was 7.108 times that of the lowest quartile (OR = 7.108, 95% CI: 5.264–9.599). The restricted cubic spline analysis showed a linear association between TyG-WC and DM risk (Pnon-linearity= 0.782). No significant interaction was found in the subgroup analysis (Pinteraction > 0.05). The DeLong test results indicated that the differences in predictive performance among the ROC curves were statistically significant (P < 0.05). ROC analysis demonstrated that the predictive performance of TyG-WC was optimal (AUC = 0.783), which was significantly higher than that of TyG (AUC = 0.752) and waist circumference (AUC = 0.698). Conclusion In the adult population aged 35-75 years in Hainan Province, TyG-WC is independently and positively correlated with the risk of diabetes and can serve as an effective indicator for diabetes screening, especially among women.

Diabetes Mellitus, Triglyceride-Glucose-Waist Circumference Index, Prevalence