Atherogenic index of plasma and subclinical vascular disease: predictive value for coronary and carotid atherosclerosis in a health screening population
Xiaowei Li, Ping Shuai, Xincheng Huang, Yan Mou, Peiyuan He
Abstract
OBJECTIVE: To evaluate the predictive value of the atherogenic index of plasma (AIP) for coronary artery calcification (CAC), carotid atherosclerosis (CA), and their coexistence, and to develop a multidimensional risk prediction model to optimize early cardiovascular disease (CVD) screening. METHODS: A total of 32,992 individuals who underwent carotid intima-media thickness (CIMT) measurement and coronary CT calcium scoring at Sichuan Provincial People's Hospital between January and December 2023 were enrolled. Participants were classified into healthy, CAC, CA, and combined groups. AIP quartiles were derived from the cohort distribution. Nonparametric tests, Spearman correlation, and multivariable logistic regression with sequential covariate adjustments (Models 1-4) were applied. Predictive performance was assessed by sROC curves, AUC, and validated using bootstrap resampling and ten-fold cross-validation. RESULTS: AIP levels differed significantly across groups, being lowest in healthy individuals (0.01) and highest in the CAC and combined groups (0.09 and 0.06, respectively). Spearman correlations between AIP and atherosclerotic burden were strongest in healthy participants (ρ = 0.89) and progressively attenuated in CAC (ρ = 0.64) and combined groups (ρ = 0.43). Higher AIP quartiles were associated with increased risk of CA, CAC, and combined lesions, with Model 4 showing the highest predictive performance (AUC = 0.91 for combined lesions). Bootstrap and cross-validation confirmed model stability. CONCLUSION: AIP may help identify early subclinical coronary and carotid atherosclerosis, and sequential adjustment with metabolic indicators improves predictive performance. External validation is warranted.