Abstract
BACKGROUND AND AIMS: Recent studies have suggested an interplay between conicity index (C-index)-related diabetes risk and lipid burden. It is plausible that the atherogenic index of plasma (AIP), fatty liver, and HbA1c mediate the association between C-index and diabetes risk, though this has not been fully explored. This study explored whether AIP, fatty liver, and HbA1c mediate the relationship between C-index and diabetes risk, as well as their combined effect. METHODS: Data from 15,453 participants in the NAGALA Cohort were analyzed (median follow-up 5.39 years). Restricted Cubic Spline (RCS) and univariate Cox regression models adjusted for risk factors were used to assess the role of AIP in modifying the C-index-diabetes relationship. Mediation analysis assessed the contributing factors, and predictive models for diabetes were established. RESULTS: Among normoglycemic individuals, the AIP and C-index remained significantly and positively associated with diabetes risk. Higher AIP levels strengthened the C-index-diabetes association, particularly in the AIP range of 0.11-≤1.21. In the initial model, hazard ratios (HRs) for those in the fourth quartile of the C-index distribution in this group showed a significant HR of 2.22 (1.37-3.59). As fatty liver and HbA1c levels were progressively adjusted, the HRs gradually decreased, but a significant HR of 1.70 (1.05-2.76) was retained in the fully adjusted model. No significant association was observed in the other AIP strata. Furthermore, AIP, fatty liver, and HbA1c mediated the relationship between C-index and diabetes risk, with mediation effects of 9.8%, 25.0%, and 13.4%, respectively. Notably, the combined model incorporating AIP, fatty liver, HbA1c, and the C-index achieved the highest predictive performance (AUC = 0.86), outperforming the C-index alone (AUC = 0.68). CONCLUSIONS: C-index was significantly associated with diabetes risk, modified by AIP, fatty liver, and HbA1c. These findings emphas ize the importance of AIP along with the C-index, particularly in the context of fatty liver and HbA1c, for diabetes risk screening and management.