BACKGROUND: This study aimed to investigate the complex associations of frailty and high-sensitivity C-reactive protein (hsCRP) with cardiovascular disease (CVD) through a nationwide prospective cohort, while also assessing the mediating associations. METHODS: According to critical criteria, a total of 5239 participants from the China Health and Retirement Longitudinal Study (CHARLS) in 2011 were ultimately enrolled in this study. Frailty was evaluated by the frailty index with 40 items, and CVD was defined as the presence of physician-diagnosed heart disease and/or stroke. A restricted cubic spline model, receiver operating characteristic curves, adjusted Cox proportional hazards regression, interaction analyses and mediation analyses were performed for association exploration. RESULTS: During a maximum follow-up of 7.0 years, 1204 (23.67%) people developed CVD. Both elevated hsCRP and frailty were significantly associated with CVD incidence. Compared with participants with a healthy status and low hsCRP (<
1.015 mg/L), those with a frailty status and elevated hsCRP had the highest risk of CVD (adjusted HR, 2.97
95% CI 2.29-3.84), heart disease (adjusted HR, 2.93
95% CI 2.16-3.96), and stroke (adjusted HR, 4.26
95% CI 2.81-6.44), which were still robust in the subgroup analysis. Moreover, frailty significantly mediated 19.60% of the associations between hsCRP and CVD. CONCLUSIONS: Combined assessment of frailty and hsCRP levels helps to better stratify the individual risk of CVD. Frailty could partly mediate the associations between hsCRP and CVD incidence.