Application of the C-reactive protein-triglyceride glucose index in predicting the risk of new-onset diabetes in the general population aged 45 years and older: a national prospective cohort study.

OBJECTIVE

Triglyceride-to-glucose index (TyG index) and inflammation are both independent risk factors for diabetes. However, only a few studies have combined TyG index with inflammation indices to predict diabetes risk. C-reactive protein-triglyceride-to-glucose index (CTI index), as a new type of lipid and inflammation marker, can comprehensively assess the severity of insulin resistance and inflammation. This study explores the association between CTI index and diabetes risk.

METHODS

We recruited a total of 6,728 participants from the China Health and Retirement Longitudinal Study (CHARLS) who had no history of diabetes at baseline. After determining the key predictors using the least absolute shrinkage and selection operator (LASSO) technique, the relationship between the CTI index and the risk of new-onset diabetes was assessed using multivariate COX regression, the mediating effect between insulin resistance and inflammatory indicators was explored, and restricted cubic splines (RCS) were applied to explore the association between the CTI index and the risk of new-onset diabetes. In addition, we used decision tree analysis to identify people at high risk of diabetes, calculated time-dependent Harrell's C index (95% CI) to assess the predictive ability of TyG, CRP, CTI and CRP + TyG for new-onset diabetes, and further calculated IDI and NRI to assess the predictive ability of CTI and TyG. Finally, we performed subgroup analyses for different subgroups using stratified COX proportional hazard regression models; and a series of sensitivity analyses were performed to verify the robustness of our results.

RESULTS

The incidence of diabetes was 15.9% during the 9-year follow-up. COX regression analysis showed that the risk ratio for diabetes increased gradually with an increase in the CTI index. The RCS curve confirmed the existence of a linear relationship between the CTI index and the risk of diabetes. Decision tree analysis showed that the CTI index was a key indicator of diabetes risk. In addition, the CTI index is a much better predictor of the onset of diabetes risk than the TyG index, as demonstrated by the integrated discrimination improvement (IDI) and net reclassification improvement (NRI).

CONCLUSION

An increase in CTI levels is closely related to diabetes risk, and the CTI index may become a unique predictor of diabetes risk.

CLINICAL TRIAL NUMBER

Not applicable.