Abstract
BACKGROUND: Abnormal lipid metabolism is a well-established risk factor for insulin resistance. However, the effect of the atherogenic index of plasma (AIP) on diabetes or impaired fasting glucose (IFG) has been rarely explored, particularly in large-scale, multicenter studies. Therefore, this study aims to assess the effect of AIP on diabetes or IFG in a large cohort of the Chinese population. METHODS: This multicenter cohort study encompassed 100,876 normoglycemic subjects in China. The effect of AIP on diabetes or IFG was analyzed using a Cox proportional hazards regression analysis. Nonlinear associations were examined using restricted cubic spline models. The use of time-dependent ROC analysis serves to evaluate the predictive performance of glycemic status over a time frame of 3, 4, and 5 years. RESULTS: During an average follow-up of 3.12 years, 13,090 participants (12.98%) developed diabetes or IFG. Multivariate analysis revealed that heightened AIP independently predicted an elevated risk of diabetes or IFG development (HR: 1.25, 95% CI: 1.17–1.34, P < 0.0001). Further analysis uncovered a nonlinear correlation between AIP and diabetes or IFG. Specifically, a significant positive association was identified when AIP was below 0.03 (HR: 1.74; 95%CI: 1.53–1.98, P < 0.0001). Conversely, for AIP values above 0.03, the association was no longer statistically significant (HR: 0.91, 95% CI: 0.81–1.04, P = 0.1582). Furthermore, the AIP demonstrates a non-linear relationship with both diabetes and IFG. The time-dependent ROC analysis reveals that the AUC for diabetes risk increases from 0.5931 at 3 years to 0.6795 at 5 years, while the AUC for IFG risk shows a slight increase from 0.5931 at 3 years to 0.6036 at 5 years, and the AUC for diabetes or IFG risk rises from 0.5966 at 3 years to 0.6102 at 5 years. CONCLUSION: AIP levels demonstrated a positive, nonlinear correlation with diabetes or IFG risk. These findings substantiate the utility of AIP as a predictive marker for future glycemic status. The clinical application of AIP might provide a novel approach for long-term monitoring and intervention in hyperglycemia prevention. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s12020-025-04490-7.