Abstract TP102: Lifetime Lipid Trajectories in Relation to Post-Stroke Dementia Prevalence in the Framingham Heart Study

Introduction: Abnormal levels of low-density lipoprotein (LDL), high-density lipoprotein (HDL), total cholesterol (TC) and triglycerides (TRIG) have been linked to cardiovascular diseases. However, the impact of lipid levels on neurological outcomes has been understudied, especially post-stroke dementia (PSD). Research across adverse outcomes has previously focused on short term measurements, rather than long term trajectories in dyslipidemia, which capture better lifelong effects. Evaluating long-term lipid trajectories in relation to PSD could help identify opportunities for earlier and targeted risk stratification. Aim: To assess how lifetime patterns of lipid levels are associated with PSD prevalence in the healthy, community-dwelling individuals within the Framingham Heart Study. Methods: Participants from the Original (1948–2005, 32 exams) and Offspring (1971–2016, 9 exams) cohorts of the Framingham Heart Study were included. Functional principal component analysis was applied to model trajectories for each lipid measure, and k-means clustering was used to group major trajectory patterns. Associations between lipid trajectory clusters and post-stroke dementia occurrence were assessed using logistic regression analysis. Model 1 adjusted for age, sex, and cohort; Model 2 additionally adjusted for baseline hypertension, smoking, and diabetes; and Model 3 further adjusted for lipid treatment duration. Results: In 1433 participants (44% male, mean age at baseline 43 years) there were 431 PSD cases. Three to four risk clusters of lifelong trajectories were identified (HDL n=559, LDL n=422, and TC n=1433, TRIG n=462). In all models using the lowest lifelong level (or highest for HDL) as reference, no significant associations were observed between any lipid trajectories and PSD. Conclusion: In healthy community dwelling participants, lifelong trajectories of lipid levels were not associated with PSD after multivariable adjusted analyses. These findings require replication in larger samples but suggest that other factors play a stronger role in post stroke dementia.