Demographic-specific TyG and TyG-BMI thresholds improve insulin resistance and mortality risk

Introduction: Insulin resistance (IR) is a key contributor to metabolic disorders and mortality. The triglyceride–glucose (TyG) and TyG–body mass index (TyG-BMI) indices are increasingly used as surrogate markers of IR. However, the influence of demographic factors, such as age, biological sex, and ethnicity, on their performance remains unclear. Materials and methods: This observational, cross-sectional study utilized 11,061 adults (n-weighted = 39,603,280) from the 1999–2018 National Health and Nutrition Examination Survey (NHANES), linked to mortality follow-up through 2018. TyG and TyG-BMI indices were calculated, and their associations with IR defined by Homeostatic Model Assessment for IR (HOMA-IR) and all-cause mortality were examined using Receiver Operating Characteristic (ROC) curves, Cox regression, and Kaplan–Meier survival analyses. Thresholds were derived for the overall population and by age, biological sex, and ethnicity. Results: TyG and TyG-BMI were significantly influenced by age, biological sex, and ethnicity. Males and older individuals had higher index values, while ethnic differences were notable, particularly among Mexican Americans and non-Hispanic Black individuals. TyG-BMI showed a stronger correlation with HOMA-IR (r = 0.574, p < 0.001) than TyG (r = 0.409, p < 0.001). Stratified thresholds were associated with improved diagnostic performance, with TyG-BMI showing the highest sensitivity (93.8%) and specificity (82.4%) among older males from underrepresented ethnic groups. Elevated TyG and TyG-BMI indices were significantly associated with all-cause mortality, with TyG showing a stronger association (Hazard Ratio = 2.3, 95% confidence interval: 1.8–2.8, p < 0.001) than TyG-BMI or HOMA-IR. Conclusions: Demographic-specific TyG and TyG-BMI thresholds are associated with improved identification of IR and stratification of mortality risk, supporting their potential utility as accessible, cost-effective tools for metabolic risk assessment in diverse populations.