The Association Between New Insulin Resistance Indices and All-Cause Mortality in Elderly Patients with Diabetes: A Prospective Cohort Study

Background The association between newly developed insulin resistance (IR) indices and all-cause mortality in elderly patients with diabetes has not been investigated. Methods Baseline data and all-cause mortality for 1,248 elderly diabetes patients from the National Health and Nutrition Examination Survey (NHANES) conducted between 2001 and 2018 were collected. The traditional IR index homeostasis model assessment of insulin resistance (HOMA-IR) and several newly developed indices, including metabolic score for insulin resistance (METS-IR), triglyceride to high-density lipoprotein cholesterol ratio (TG/HDL-C), triglyceride glucose index (TyG), triglyceride glucose combined with body mass index (TyG-BMI), estimated glucose disposal rate (eGDR), and visceral adiposity index (VAI), were calculated for the patients. Cox proportional hazards regression and restricted cubic spline regression models assessed the relationship between IR indices and all-cause mortality. Results In a median follow-up period of 73.3 months, there were 381 recorded deaths. In the total cohort, METS-IR (p < 0.001), TyG-BMI (p < 0.001), and eGDR (p = 0.011) demonstrated a significant association with all-cause mortality as continuous variables. HOMA-IR, METS-IR, TyG-BMI, and eGDR exhibited significant correlations with all-cause mortality in the Cox regression models (p < 0.05) when analyzed as categorical variables. A U-shaped relationship exists between METS-IR, TyG-BMI, eGDR, and all-cause mortality (p-overall < 0.0001, p-nonlinear < 0.05). No significant associations were found between TyG, TG/HDL-C, VAI, and all-cause mortality. Among male patients, TyG-BMI and HOMA-IR exhibited superior prognostic value, whereas in female patients, METS-IR, TyG-BMI, and eGDR showed better performance. Conclusion HOMA-IR, TyG-BMI, METS-IR, and eGDR were associated with mortality in elderly diabetic patients, with gender differences in their prognostic values.