Abstract
BACKGROUND: Understanding the evolving patterns of cardiovascular disease (CVD) burden attributable to ambient particulate matter pollution (APMP) is essential. Furthermore, research on the underlying mechanisms has mostly been limited to laboratory and animal models, with few large-scale population-based studies. METHODS: Using data from the Global Burden of Disease Study (GBD) 2021, we analyzed disability-adjusted life years and mortality for CVD attributable to APMP (measured as particulate matter [PM](2.5)) from 1990 to 2021. We examined shifts in burden between APMP and household air pollution (HAP), regional disparities by socio-demographic index (SDI), and predicted trends using a Bayesian age-period-cohort model. Additionally, we used UK Biobank (UKB) data (metabolomics: 230,000 + participants; proteomics: 50,000 +) to identify biomarkers mediating the association between PM(2.5) exposure and CVD outcomes, and further analyzed their biological roles. Metabolic and proteomic signatures were constructed using regression and elastic net models, with predictive performance assessed via time-dependent receiver operating characteristic analysis. Life expectancy was evaluated using flexible parametric survival models. Subgroup analysis was conducted by age, sex, lifestyle, socioeconomic status, and genetic susceptibility. RESULTS: In 2021, the global CVD absolute burden attributable to APMP was more than double that of 1990, with significant regional disparities. The burden shifted from HAP to APMP, with 15% of CVD cases globally attributed to APMP. The CVD burden attributable to APMP increased with age and is projected to rise through 2030. In the UKB, approximately 30 metabolites, including albumin, mediated the association between PM(2.5) exposure and CVD outcomes, primarily involving lipid and fatty acids metabolism. Over 60 proteins, including growth differentiation factor-15 and trefoil factor 2, mediated the association with CVD outcomes, enriched in cytokine-receptor interaction and leukocyte migration pathways. Metabolic and proteomic signatures outperformed PM(2.5) alone in predicting 1-, 5-, and 10-year CVD outcomes. Participants in the lowest decile of PM(2.5) exposure, metabolic, and proteomic signatures had longer life expectancy than those in the highest decile. CONCLUSION: The CVD burden attributable to APMP remains a critical public health concern. This study presents a novel approach for identifying and managing susceptible populations through metabolomic and proteomic perspectives.