Cumulative C-reactive protein-triglyceride-glucose index and longitudinal trajectories as predictors of new-onset diabetes in middle-aged and older adults: A prospective cohort analysis from CHARLS

BACKGROUND: The C-Reactive Protein-Triglyceride-Glucose Index (CTI) has recently emerged as a novel biomarker reflecting both insulin resistance (IR) and systemic inflammation. While its association with cardiovascular outcomes has been explored, evidence regarding the impact of cumulative CTI (cumCTI) exposure and its dynamic trajectories on the risk of new-onset diabetes remains limited. This study aims to investigate the association between dynamic changes in CTI and the incidence of diabetes in a Chinese population. METHODS: This prospective cohort analysis utilized data from the China Health and Retirement Longitudinal Study (CHARLS). A total of 6044 middle-aged and older adults without diabetes at baseline were included. cumCTI was calculated as the time-weighted average of CTI between 2012 and 2015. K-means clustering was employed to identify distinct CTI trajectory patterns. Cox proportional hazards regression models, restricted cubic splines (RCS), and receiver operating characteristic (ROC) curves were utilized to evaluate the associations and discrimination performance. RESULTS: During the follow-up period, 1209 participants (20.00%) developed new-onset diabetes. Three distinct CTI trajectories were identified: Low-Stable, Moderate-Stable, and High-Stable. In the fully adjusted Cox regression model, each 1-unit increase in cumCTI was associated with a 22% higher risk of diabetes (Hazard Ratio HR = 1.22, 95% CI: 1.18-1.26, P < 0.001). Compared to the Low-Stable group, participants in the High-Stable trajectory faced a 162% increased risk (HR = 2.62, 95% CI: 2.20-3.13, P < 0.001). RCS analysis demonstrated a continuous, linear dose-response relationship (P for non-linearity = 0.205). Subgroup analyses revealed that these associations remained highly consistent across all clinical strata with no significant interactions. Furthermore, cumCTI outperformed baseline TyG and hs-CRP alone in distinguishing both 7-year (AUC = 0.653) and 9-year (AUC = 0.643) diabetes risk. CONCLUSION: Cumulative exposure to CTI and its longitudinal High-Stable trajectories are robust and independent predictors of new-onset diabetes in middle-aged and older adults. Monitoring long-term immuno-metabolic dynamics provides superior risk stratification and significant prognostic implications for personalized diabetes prevention strategies.