Cross-Cohort Variation in First Reported Myocardial Infarction and Stroke Risk Among Middle-Aged and Older Adults with Diabetes: A Harmonized Longitudinal Cohort Study

Background: Composite cardiovascular outcomes can obscure which non-fatal event is first reported among adults with diabetes. We compared standardized five-year risks of first reported myocardial infarction and stroke across longitudinal ageing cohorts and separated cross-cohort variation in absolute risk from women–men contrasts. Methods: We harmonized the English Longitudinal Study of Ageing, the US Health and Retirement Study, and the Mexican Health and Aging Study. Participants were aged at least 50 years, had diabetes, and had no previously reported myocardial infarction or stroke. Original interview intervals were analyzed with a joint competing-risk likelihood for myocardial infarction, stroke, same-wave dual reports, and death. Models incorporated multiple imputation, inverse-probability-of-censoring weights, and g-computation standardized to a common age distribution; uncertainty used 500 participant-cluster bootstrap replicates per cohort. Results: The analysis included 11,465 participants (ELSA, 1452; HRS, 6074; MHAS, 3939). Standardized myocardial-infarction risk ranged from 3.23% to 4.16% in men and 2.24% to 3.92% in women; stroke risk ranged from 2.50% to 3.67% in men and 2.36% to 4.20% in women. The pooled women-minus-men risk difference was −0.73 percentage points (95% CI, −1.74 to 0.27) for myocardial infarction and 0.30 percentage points (−0.75 to 1.36) for stroke. Conclusions: Absolute risks varied across cohorts, whereas pooled women–men contrasts were small and imprecise. Cohort, outcome-definition, and observation context should be explicit when cardiovascular risk estimates are interpreted or calibrated for adults with diabetes.