Xiaomeng Mi, Anqi Shi, Suting Xiong, Wenchao Xu, Fang Yao, Wenguo Xu
This study aimed to identify longitudinal triglyceride-glucose (TyG) index trajectories and evaluate their association with annual estimated glomerular filtration rate (eGFR) change in a high cardiovascular-risk Chinese cohort, with sex-stratified analyses. A prospective cohort analysis was conducted using data from the China PEACE Million Persons Project. The primary longitudinal analysis included 4814 participants (2233 men and 2581 women) contributing 9378 observations. TyG index trajectories were identified using finite mixture modeling. Linear mixed-effects models after multiple imputation with 20 imputed datasets were used to assess associations between trajectory groups and annual eGFR change, adjusting for demographics, blood pressure, body mass index, and baseline medication use. Three short-term stable TyG trajectories were identified: Stable-Low, Stable-Middle, and Stable-High. In the m = 20 multiple-imputation analysis, the Stable-Middle trajectory was associated with slower annual eGFR decline compared with the Stable-Low trajectory in the overall cohort (beta for group × time = 0.94; 95% CI 0.34–1.53; P = 0.002), in men (beta = 1.01; 95% CI 0.17–1.85; P = 0.019), and in women (beta = 0.87; 95% CI 0.03–1.71; P = 0.043). The Stable-High trajectory was not significantly different from the Stable-Low trajectory in either sex. In this high cardiovascular-risk cohort, an intermediate stable TyG trajectory was associated with slower short-term eGFR decline than a low stable trajectory, with broadly similar direction in both sexes. These findings support cautious trajectory-based interpretation of TyG rather than a simple lower-is-better assumption, but the short follow-up and modest trajectory separation require longer-term validation.