The joint association of intrinsic capacity and physical activity with cardiovascular risk in older adults: evidence from four longitudinal cohorts

Abstract Background Intrinsic capacity (IC) serves as an integrative measure of biological reserve, and physical activity (PA) is a key modifiable determinant of cardiovascular health. The joint association of IC and PA with incident cardiovascular disease (CVD), and the mediating role of PA, remain unclear across populations. Methods We analyzed 14,823 adults aged ≥ 60 years without baseline CVD from four longitudinal cohorts: CHARLS, ELSA, HRS, and MHAS. IC was assessed across five domains and classified as intact or impaired. PA was categorized as regular or inactive. Cox models estimated hazard ratios (HR) for incident CVD across four groups defined by IC and PA status. Interaction and mediation analyses evaluated combined effects, and random-effects meta-analysis pooled cohort estimates. Results Both impaired IC and inactivity were independently associated with higher CVD risk. Participants with both impaired IC and inactivity had the highest risk (CHARLS HR: 2.10; ELSA HR: 1.80; HRS HR: 1.62; MHAS HR: 1.68; pooled HR: 1.74, 95% CI: 1.57–1.93). Risk increased with the number of impaired IC domains, and within both PA groups, greater IC impairment was consistently associated with higher CVD risk. Mediation analysis suggested that PA accounted for part of the association between IC and CVD, with larger estimated proportions in ELSA (17.97%) and HRS (15.68%) than in CHARLS (5.29%) and MHAS (2.43%). Conclusion Impaired IC and physical inactivity were jointly associated with higher CVD risk, with IC impairment consistently associated with elevated risk regardless of PA status. PA appeared to account for part of this association, although the magnitude varied across cohorts.