Abstract
BACKGROUND: The atherogenic index of plasma (AIP) and obesity indices have been introduced as cost-effective indicators for cardiovascular disease (CVD) risk. We aimed to investigate the association between cumulative changes in AIP and its obesity-related derivatives (AIP, AIP-BMI, AIP-WC, AIP-WHtR, AIP-BRI, and AIP-CVAI) and CVD using a nationally representative cohort. METHODS: A total of 4,519 CVD-free participants from the China Health and Retirement Longitudinal Study were included. Based on repeated measurement data from Waves 1 and 3, we classified the control levels of AIP and its obesity-related derivatives using k-means clustering and evaluated their cumulative exposure. A multi-model joint analysis strategy was adopted to systematically assess the associations of cumulative changes in AIP and its obesity-related derivatives with CVD, the component contributions, and the mediating role of glycometabolic factors. RESULTS: Over 8 years' median follow-up, the study found that both AIP and its obesity-related derivatives (including baseline levels and cumulative changes) exhibited linear positive associations with CVD events. Compared to baseline assessments, evaluating the cumulative changes in AIP and its obesity-related derivatives further enhanced the assessment of CVD risk. Compared to individuals with better control level or lowest cumulative exposure, those with poor control level or highest cumulative exposure of AIP, AIP-BMI, AIP-WC, AIP-WHtR, AIP-BRI, and AIP-CVAI exhibited odds ratios of 1.37/1.34, 1.48/1.43, 1.44/1.42, 1.36/1.42, 1.43/1.42, and 1.51/1.42, respectively. Notably, diabetic status further amplified the impact of cumulative changes in AIP and its obesity-related derivatives on CVD risk. Further weighted quantile sum analysis revealed that cumulative triglycerides and cumulative obesity indices exhibited the highest relative contribution weights to CVD events. Finally, mediation analysis demonstrated that cumulative haemoglobin A1c partially mediated the CVD risk associated with cumulative AIP and its obesity-related derivatives. CONCLUSIONS: AIP and its obesity-related derivatives (particularly AIP-CVAI) exhibited significant positive associations with incident CVD risk. The strength of these associations dynamically influenced with cumulative changes in these metrics and significantly modified by diabetic status. Comprehensive analysis using weighted quantile sum and mediation models revealed that cumulative glucose effect partially mediated these associations, while sustained exposure to triglycerides and obesity emerged as primary drivers.