2026 American Academy of Neurology Abstract Website

Aksel D Laudon¹, Xinrui Shi², Hugo Javier Aparicio¹, Vasileios-Arsenios Lioutas³, Oluchi Ekenze¹, Virginia Howard⁴, Myriam Fornage⁵, Sudha Seshadri⁶, Alexa Beiser⁷, Qiong Yang², Jose Romero¹
¹Department of Neurology, Boston University Chobanian & Avedisian School of Medicine; Framingham Heart Study, ²Department of Biostatistics, Boston University School of Public Health; Framingham Heart Study, ³Department of Neurology, Beth Israel Deaconess Medical Center, ⁴Department of Epidemiology, School of Public Health, University of Alabama at Birmingham, ⁵Institute of Molecular Medicine, McGovern Medical School, University of Texas Health Science Center at Houston, ⁶Glenn Biggs Institute for Alzheimer's and Neurodegenerative Diseases; Framingham Heart Study, ⁷Department of Neurology, Boston University Chobanian & Avedisian School of Medicine; Department of Biostatistics, Boston University School of Public Health; Framingham Heart Study

Objective:

To evaluate the association of long-term Framingham Stroke Risk Profile (FSRP) trajectories with incident stroke or transient ischemic attack (TIA) and post-stroke incident dementia (PSD) in a community-dwelling population.

Background:

Cross-sectional FSRP assessments predict 10-year probability of incident stroke. Whether long-term trajectory of FSRP score better reflects the risk of stroke and PSD is unclear.

Design/Methods:

Stroke- and dementia-free Framingham Heart Study Original and Offspring Cohort participants with at least 9 and 4 FSRP observations, respectively, were included. FSRP trajectories over age were analyzed using functional principal component (FPC) analysis, followed by K-means clustering of FPC scores. Cox proportional hazards regression was performed with the FPC scores and clusters as main exposures in models adjusted for age, FSRP (at first FSRP observation for stroke/TIA and at stroke/TIA for PSD), sex, and cohort.

Results:

FPC analysis identified 2 major modes of variation in FSRP trajectories of 6,142 participants, where the first FPC explained 90% of variance, and the second an additional 9%. 945 (15.4%) had incident stroke/TIA over mean follow-up 31.9 years (SD 9.2). Increased FPC1 score was associated with increased stroke/TIA risk (HR 1.54, 95%CI 1.39-1.69) while increased FPC2 score was associated with decreased risk (HR 0.25, 95%CI 0.17-0.39). K-means clustering identified 2 groups: cluster 1 (n=5,554) had 14.0% stroke/TIA incidence versus 28.1% in cluster 2 (n=588). Cluster 2 was associated with higher stroke/TIA risk (HR 1.54, 95%CI 1.30-1.83). Among 756 dementia-free stroke/TIA survivors, 141 (18.7%) developed dementia over mean follow-up 6.0 years (SD 6.2), with a greater proportion in cluster 2 (27/129, 20.9%) versus cluster 1 (114/627, 18.2%). However, cluster 2 was not associated with PSD after adjusting for FSRP at stroke/TIA (HR 0.77, 95% CI 0.46-1.29).

Conclusions:

Long-term FSRP trajectory was associated with risk of incident stroke/TIA but not PSD. Our findings support consideration of longitudinal FSRP trajectory to assess stroke risk.