Utility of TyG-Based Indices for Predicting Insulin Resistance in Turkish Adults: Insights from the TEKHAP Study

Background/Objectives: Insulin resistance (IR) is a key feature of metabolic disorders and a major precursor of type 2 diabetes and cardiovascular disease. The triglyceride-glucose (TyG) index and TyG-based indices-including TyG-body mass index (TyG-BMI), TyG-waist circumference (TyG-WC), and TyG-waist-to-height ratio (TyG-WHtR), have been proposed as simple, cost-effective surrogate markers of IR. However, population-based data from Türkiye are limited. To evaluate the association between TyG-based indices, triglycerides/high-density lipoprotein cholesterol (TG/HDL-C) ratio, and IR defined by Homeostasis Model Assessment-IR (HOMA-IR) in adults from the TEKHAP study-a province-wide, population-based survey in Tokat, Türkiye. Methods: A total of 1854 adults (≥20 years) were included in the analysis. Physiologically implausible HOMA-IR outliers were identified using statistical criteria and excluded. IR was defined as HOMA-IR ≥ 2.46, previously validated in this population. TyG and its derivatives were calculated from fasting triglyceride, glucose, and anthropometric measurements. Group comparisons between IR and non-IR individuals, correlation analyses, receiver operating characteristic (ROC) curves, and multivariate logistic regression models were conducted to evaluate the diagnostic and independent associations of these surrogate markers with IR. Results: IR prevalence was 27.2%. Participants with IR had significantly higher triglycerides, fasting glucose, insulin, C-peptide, BMI, and waist circumference and lower HDL-C levels (all p < 0.001). All TyG-based indices were higher in the IR group and showed weak-to-moderate positive correlations with HOMA-IR. TyG-BMI showed the highest diagnostic accuracy in ROC curve analyses (AUC = 0.765); however, this association could not be interpreted as an independent predictive effect in adjusted models because of collinearity with BMI. In multivariable logistic regression, the TyG index demonstrated the strongest independent association with IR (OR = 4.14; 95% CI: 3.32-5.18; p < 0.001), while TyG-WC and the TG/HDL-C ratio also retained significant independent predictive value. Conclusions: The TyG index showed the strongest independent association with IR, while the TG/HDL-C ratio and TyG-WC also demonstrated significant independent predictive value. TyG-based indices may represent practical, low-cost surrogate markers for early metabolic risk stratification in community settings; however, their role in formal screening strategies requires external validation and calibration in independent populations.