Abstract
BACKGROUND: The obesity and Atherogenic Index of Plasma (AIP) have been identified as independent predictors of abnormal glucose metabolism. The objective of this study is to systematically evaluate the relationships between control patterns for AIP and its obesity-associated indices (AIP/AIP-WHtR/AIP-WC/AIP-BMI) and transitions in prediabetic glycemic status. METHODS: The analytical sample comprised 2,535 prediabetic individuals derived from the national cohort. Utilizing repeated measures data from the first and third waves of the national survey, a k-means clustering algorithm was employed to classify control patterns for AIP and its obesity-associated indices. Using Cox regression, weighted quantile sum regression, restricted cubic spline regression, and the four-way decomposition method, the researcher systematically evaluated the associations of AIP and its obesity-associated indices—including baseline static levels and dynamic control patterns—with transitions in prediabetic glycemic status, while also quantifying component contributions and assessing inflammation's mediating role. RESULTS: Longitudinal analysis revealed that both baseline static values and dynamic control patterns of AIP and its obesity-associated indices were positively associated with the progression of prediabetes. For prediabetes reversal, the baseline static values of AIP and its obesity-associated indices showed no significant association with outcomes; in contrast, their dynamic control patterns demonstrated significant or borderline significant associations. Weighted quantile sum regression analysis indicated that cumulative exposure to obesity and triglycerides during follow-up had the highest weight in prediabetes progression. Further analysis using the four-way decomposition method revealed that cumulative inflammatory exposure demonstrated a significant mediated interaction in the association of control patterns for AIP and its obesity-associated indices with prediabetes progression. CONCLUSION: The dynamic control patterns of AIP-obesity-associated indices (particularly AIP-BMI) significantly influence transitions in prediabetic glycemic status. These associations are primarily driven by cumulative exposure to obesity and triglycerides, and are further mediated by persistent inflammatory exposure. Dynamic long-term monitoring of AIP and its obesity-associated indices may facilitate the management of glycemic status transitions in prediabetes. GRAPHICAL ABSTRACT: [Image: see text] SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12944-026-02900-x.