Abstract Background To investigate the triglyceride‐glucose index (TyG) as a predictor of glycaemic variability phenotypes in patients with type 2 diabetes mellitus (T2DM) using continuous glucose monitoring (CGM)‐derived metrics, guiding personalized management. Materials and Methods This cross‐sectional study analysed 279 T2DM patients who underwent 14‐day CGM monitering from community hospitals in Ningbo, China. Participants were stratified by TyG tertiles. Spectral clustering with complexity‐invariant dynamic time warping identified distinct glycaemic variability glucotypes. Comprehensive CGM metrics spanning time‐domain, frequency‐domain, event‐based, and circadian rhythm domains were evaluated. Multivariable logistic regression assessed associations between TyG and glycaemic phenotypes. Results Across TyG tertiles, we observed progressive increases in HbA1c (6.8 ± 1.5%, 7.2 ± 1.2%, 7.3 ± 1.3%, p = 0.011), fasting glucose (6.1 ± 1.1, 7.2 ± 1.6, 8.5 ± 2.8 mmol/L, p < 0.001), and triglycerides (0.9 ± 0.2, 1.4 ± 0.4, 2.7 ± 1.0 mmol/L, p < 0.001). Most notably, patients in the highest TyG tertile exhibited sustained hyperglycaemia with lower TIR (81.8% to 72.8 to 70.3%, p = 0.001), elevated TAR (15.9% vs. 29.1%, p < 0.001), and paradoxically reduced TBR (2.3% to 1.4% to 0.6%, p = 0.014) and CV (26.2% to 25.8% to 23.8%, p = 0.021). Frequency domain analysis demonstrated that elevated TyG was associated with high‐level, low‐frequency glucose oscillations. Higher TyG tertiles demonstrated increased prevalence of severe variability glucotype (32.3% to 53.8% to 63.4%), with TyG emerging as the strongest independent predictor (OR = 1.87, 95% CI: 1.42–2.50, p < 0.001). Conclusions Elevated TyG identifies a distinct glycaemic phenotype characterized by high‐level, low‐frequency oscillations, sustained hyperglycaemia, and minimal hypoglycaemic risk. T2DM patients with high TyG levels may be appropriate candidates for more intensified glucose‐lowering strategies.
Wu et al. (Wed,) studied this question.