Abstract Body: Introduction Cerebral small vessel disease (CSVD) is linked to stroke and dementia risk, often predating clinical events for years to decades. While preventive strategies focus on treatment of modifiable vascular risk factors (VRF), the complex evolution in these VRF over long periods and their association with CSVD burden are not fully understood.
Hypothesis Individuals exhibit different VRF trajectories over a lifetime, and these patterns determine their CSVD burden.
Methods Framingham Heart Study participants with ≥6 VRF assessments over their lifetime and CSVD markers on brain MRI were included. A multi-marker CSVD score quantified CSVD burden, assigning one point each to cerebral microbleeds, covert infarcts, extensive white matter hyperintensities, cortical superficial siderosis and high burden of visible perivascular spaces (range 0-5), and was categorized for analysis as 0, 1, and ≥2. Using functional data clustering to the trajectories of each VRF, we identified and characterized participant clusters of lifetime trends. Multivariable ordinal logistic regression was employed to associate cluster assignment with multi-marker CSVD scores.
Results In 7961 participants with longitudinal VRF data (baseline mean age 39.8 ± 10 years, 45% men), three clusters were identified for each VRF (Figure). These clusters broadly depict groups with long-term exposure to 1) abnormal or undesirable VRF levels, 2) optimal ranges, and 3) levels between optimal and abnormal, varying by VRF. Among 1625 participants with CSVD measurements (mean age at MRI 72.8 ± 9.4 years, 45% men), CSVD score 0 was seen in 47%, 1 in 31% and ≥2 in 22%. In age, sex, and cohort adjusted analyses, VRF cluster assignment for systolic blood pressure (p<0.01), pulse pressure (p<0.01), cigarettes per day (p=0.01), and body mass index (p<0.05) were associated with multi-marker CSVD scores, with the highest risk cluster generally showing the strongest effects. For instance, compared to participants with long-term optimal systolic blood pressure, those with abnormally high levels (OR 1.70; 95% CI 0.97, 3.00) or levels between optimal and abnormal (OR 1.39; 95% CI 1.14, 1.69) had increased odds of having higher CSVD scores.
Conclusions: Individuals exhibit different lifetime trajectories of modifiable VRF, which are associated with burden of CSVD. Preventive efforts should consider the trajectory of VRF to identify and target high risk individuals to prevent high CSVD burden and its adverse consequences.