Early identification of individuals at high risk for type 2 diabetes is a priority in primary care, especially among young and middle-aged adults, where subtle metabolic changes often go unrecognized despite long-term consequences. The homeostatic model assessment of insulin resistance (HOMA-IR) is widely used but has limited practicality in routine care. The triglyceride–glucose (TyG) index has emerged as a simpler surrogate marker, yet it remains unclear whether changes in TyG outperform changes in HOMA-IR in predicting diabetes risk.