Cross-national patterns of cardiometabolic disease progression from healthy state to stroke: a cause-specific Cox transition analysis of six international cohorts

Cardiometabolic disease progression to stroke involves intermediate transitions through hypertension (HTN) and diabetes (DM). However, cross-national comparisons of these transition patterns remain unexplored. We compared cardiometabolic progression pathways, including death as a competing state, across six countries using harmonized longitudinal data. We analyzed 247,863 participants aged 45–100 from CHARLS (China), ELSA (England), HRS (USA), KLoSA (South Korea), MHAS (Mexico), and SHARE (Europe). Six states were defined: healthy, HTN-only, DM-only, HTN + DM, stroke, and death (absorbing state). Cause-specific hazard models with complementary log–log links estimated the association of age, sex, education, and smoking with each of 17 observed transition types, adjusted for within-person correlation. Cross-national differences were formally tested using country-by-covariate interaction terms. Death was modeled as a sixth state, with persons remaining at risk until death or censoring. Among 599,161 observed transitions (including 40,911 transitions to death), baseline health-state distributions summed to 100% within each country, with the proportion of healthy participants ranging from 45.5% (Mexico) to 74.5% (China). Conditional on leaving the healthy state, the most common destination was HTN-only in all countries (42.2–66.9%), followed by death (9.0–33.7%). Age was associated with increased hazards of progression to stroke from healthy (HR range 1.032–1.084 per year), from HTN (1.022–1.050), and from HTN + DM (1.013–1.032), as well as with death (1.107–1.125). Formal interaction tests identified statistically significant cross-national differences in age effects for all major transitions (Wald p < 0.05). Smoking was associated with progression to HTN + DM, and education showed protective associations in several cohorts (full results in Supplementary Tables). Cardiometabolic progression patterns and their determinants differ statistically across countries, with consistent age gradients but heterogeneous effect magnitudes. These findings support context-specific prevention strategies, while design and measurement differences across cohorts should be considered as competing explanations.