Cardiovascular Risk Trajectories and Risk of Developing Stroke and Dementia

Stroke and dementia share common cardiovascular risk factors, but few studies have evaluated long-term changes in cardiovascular disease (CVD) risk scores, which may better capture cumulative vascular burden. We aimed to investigate whether the longitudinal trajectory of CVD risk was associated with the risk of developing stroke and dementia. This prospective cohort study included residents aged 35 years and older from a community in northern Taiwan. We included participants without a history of stroke or dementia and assessed CVD risk at baseline (visit 1) and 3 follow-up visits (visits 2-4) from 1990 to 2000 using the Framingham general CVD risk function. CVD risk trajectories were modeled as linear changes over time using a pattern-mixture approach to account for attrition. Incident stroke and dementia were ascertained through linkage to National Health Insurance claims data starting in 2000. In the primary analysis, Cox proportional hazard models were used to estimate adjusted hazard ratios (HRs) for associations of CVD risk trajectory groups with risk of stroke and dementia. In the secondary analysis, we assessed the associations between baseline CVD risk and these outcomes. Among 2,335 participants (mean age 52.3 ± 11.2 years; 56.6% women), CVD risk trajectories over 10 years were classified as accelerated increase (29.2%), moderate increase (30.9%), or stable (39.9%). Over a median 21-year follow-up, an accelerated CVD risk trajectory, compared with the stable group, was associated with an increased risk of developing all stroke (HR 1.81, 95% CI 1.40-2.34), ischemic stroke (HR 1.84, 95% CI 1.37-2.48), hemorrhagic stroke (HR 2.38, 95% CI 1.49-3.80), and vascular dementia (HR 2.07, 95% CI 1.03-4.16). The secondary analysis revealed a positive association between baseline CVD risk and risk of developing stroke, but association with vascular dementia was weaker (baseline CVD risk ≥20% vs <10%: for stroke, HR 2.32, 95% CI 1.61-3.33; for vascular dementia, HR 1.84, 95% CI 0.74-4.56). No association was observed with all-cause or nonvascular dementia. Participants with an accelerated CVD risk had an elevated risk of developing stroke and vascular dementia. Our findings suggested that longitudinal trajectories of CVD risk may affect the risk of vascular dementia beyond individual baseline risks.