The study included 5755 participants from the Chinese Longitudinal Healthy Longevity Survey (CHARLS) and 2598 from the English Longitudinal Study of Ageing (ELSA). Cox proportional hazards models were used to examine the association between the SPISE index and CVD incidence, reporting hazard ratios (HR) and 95 % confidence intervals (CI). Within the Cox proportional hazards model framework, restricted cubic splines were applied to evaluate non-linear associations, and cumulative hazard curves were generated to illustrate the dynamics of risk over time. Subgroup analyses were conducted to assess effect heterogeneity, and a mediation analysis was performed to explore potential underlying mechanisms.

During the follow-up period (2004–2018 for ELSA and 2011–2018 for CHARLS), 935 (16.25 %) participants in CHARLS and 1026 (39.49 %) participants in ELSA developed CVD. In the fully adjusted Cox proportional hazards model, individuals in the highest tertile of the SPISE index had a significantly lower risk of CVD incidence compared with those in the lowest tertile (CHARLS: adjusted HR: 0.61, 95 % CI: 0.51–0.72; ELSA: adjusted HR: 0.83, 95 % CI: 0.71–0.98). RCS regression analysis showed a negative linear relationship between the SPISE index and the CVD incidence. A higher SPISE index had a lower cumulative incidence of CVD.

Subgroup analyses, stratified by gender, marital status, education level, smoking status, and alcohol drinking status, revealed no significant modification of the SPISE index–CVD incidence association. Mediation analysis examined how the SPISE index is linked to CVD events, with CHO, LDL-C, CRP, and HbA1c as potential mediators. In the CHARLS cohort, CHO (3.36 %) and LDL-C (3.81 %) partially mediated the association; in the ELSA cohort, CRP showed a significant mediating effect (11.98 %).