Association of cumulative exposure and change patterns of triglyceride–glucose index combined with novel obesity indices with glycemic transitions in prediabetes: evidence from the China Health and Retirement Longitudinal Study (CHARLS)

BACKGROUND: Prior research indicates that the triglyceride-glucose (TyG) index and its derivatives are cost-effective and easily accessible tools for assessing glucose metabolism risk. This study aimed to systematically evaluate the longitudinal associations between TyG-novel obesity indices [TyG-body roundness index (TyG-BRI), TyG-chinese visceral adiposity index (TyG-CVAI), TyG-weight-adjusted waist index (TyG-WWI), TyG-a body shape index (TyG-ABSI)] and glycemic transitions in individuals with prediabetes. METHODS: This analysis used data from the China Health and Retirement Longitudinal Study, encompassing 2004 participants with prediabetes. Employing longitudinal data from baseline and a 3‑year follow‑up, we systematically quantified the baseline values, cumulative exposure, and change patterns of TyG‑novel obesity indices. Multiple statistical models were applied to assess the associations of these indices with glycemic transitions in prediabetes and to examine the potential mediating effect of inflammatory burden. RESULTS: During follow‑up, 306 participants progressed to diabetes and 452 reverted to normoglycemia. Compared with static baseline measurements, cumulative exposure and change patterns of TyG‑novel obesity indices strengthened their associations with glycemic transitions. Notably, TyG-BRI and TyG-CVAI exhibited the strongest associations at baseline, whereas TyG-WWI showed the most prominent performance in the analyses of both cumulative exposure and change patterns. Compared with individuals exhibiting well-controlled (low baseline with a slight increase over follow-up) or low cumulative exposure, those with poor control (high baseline and minimal change over follow-up) or the highest cumulative exposure to TyG-WWI, TyG-ABSI, TyG-BRI, and TyG-CVAI faced a significantly higher risk of diabetes progression, with TyG-WWI showing the strongest association with diabetic outcomes. Conversely, elevated levels of all four TyG-novel obesity indices were associated with a reduced likelihood of reverting to normoglycemia, with the strongest inverse associations observed for TyG-BRI (both at baseline and cumulative exposure) and TyG‑WWI (cumulative exposure). For the predictive analysis of glycemic status transitions in prediabetes, the continuous net reclassification improvement demonstrated that the TyG-BRI model exhibited the greatest ability to reclassify individuals into more accurate risk categories, followed by TyG-CVAI, TyG-WWI, and TyG-ABSI. Finally, mediation analysis showed that the progression of prediabetes associated with the TyG-novel obesity indices was partially mediated (2%) by long-term inflammatory burden. CONCLUSION: Cumulative exposure and change patterns of TyG-novel obesity indices are significant factors influencing glycemic transitions in individuals with prediabetes.