Exploratory cluster analysis of noninvasive vascular measures reveals heterogeneous vascular profiles after stroke
- 1Department of Neurology, University of Kansas Medical Center, Kansas City, Kansas, United States of America
- 2Department of Physical Therapy, Rehabilitation Science, and Athletic Training, University of Kansas Medical Center, Kansas City, Kansas, United States of America
- 3Department of Neurology, University of Kansas Alzheimer’s Disease Research Center, Fairway, Kansas, United States of America
Abstract
Vascular dysfunction is prevalent after stroke and contributes to elevated cardiovascular risk; however, the extent to which vascular heterogeneity exists in this population remains unclear. This study examined whether distinct multi-domain cluster-derived vascular profiles are present in individuals with chronic stroke and whether these profiles are associated with functional performance. Vascular structure and function were assessed using flow-mediated dilation, carotid intima-media thickness, pulse wave velocity, and augmentation index, with age included in vascular clustering to account for vascular aging. Unsupervised k-means clustering was applied to identify vascular profiles. Functional performance was assessed using gait speed and predicted peak oxygen consumption (VO2peak) derived from a submaximal exercise test. Between-cluster differences were characterized using analysis of variance. Forty-five individuals with chronic stroke (62 ± 12 years; 29 ± 30 months post-stroke) were included. Three vascular profiles were identified (cluster 1: n = 13; cluster 2: n = 11; cluster 3: n = 21), with significant differences across clusters for all vascular measures (p < 0.01). Although age increased across clusters, vascular structure and function did not change uniformly. One profile exhibited elevated arterial wave reflection despite relatively preserved endothelial function and arterial structure, whereas another demonstrated impairment in all vascular variables. Vascular profile was not associated with predicted VO2peak (p = 0.31) or gait speed (p = 0.65). Three exploratory cluster-derived vascular profiles emerged among participants with chronic stroke, and age did not fully account for the observed vascular heterogeneity. Multi-domain cluster-derived vascular profiling revealed heterogeneous vascular profiles in individuals with chronic stroke, providing a basis for future studies to evaluate their relevance to cardiovascular risk stratification and long-term health management. NCT05936008: https://clinicaltrials.gov/study/NCT05936008
Related
- Vascular geometry and network topology reveal vascular remodelling beyond density measures across Chew grades in macular telangiectasia type 2
- Absolute and lung volume-normalized deep learning-derived pulmonary vascular measures in idiopathic pulmonary fibrosis: distinct clinical associations and the role of lung volume in prognosis
- Frontal Cerebral Blood Flow Deficits in Behavioral Variant Frontotemporal Dementia: Associations With Clinical Symptoms, Peripheral Vascular Biomarkers, and Neuroimaging Measures.
- Alkaline Phosphatase and Cumulative Vascular Access Burden in Maintenance Hemodialysis: An Exploratory Clinical Study
- Short-term changes in a chicken comb vascular model mimicking port-wine stain after synchronous ultrasound parallel beam treatment: an exploratory study
BibTeX
@article{bartsch2026exploratory,
title = {{Exploratory cluster analysis of noninvasive vascular measures reveals heterogeneous vascular profiles after stroke}},
author = {Bria L. Bartsch and Amanda N. Engler and Sandra A. Billinger},
journal = {PLOS One},
year = {2026},
month = nov,
doi = {10.1371/journal.pone.0362305},
url = {https://doi.org/10.1371/journal.pone.0362305},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Exploratory cluster analysis of noninvasive vascular measures reveals heterogeneous vascular profiles after stroke
AU - Bria L. Bartsch
AU - Amanda N. Engler
AU - Sandra A. Billinger
PY - 2026
DA - 2026/11/06
JO - PLOS One
DO - 10.1371/journal.pone.0362305
UR - https://doi.org/10.1371/journal.pone.0362305
AB - Vascular dysfunction is prevalent after stroke and contributes to elevated cardiovascular risk; however, the extent to which vascular heterogeneity exists in this population remains unclear. This study examined whether distinct multi-domain cluster-derived vascular profiles are present in individuals with chronic stroke and whether these profiles are associated with functional performance. Vascular structure and function were assessed using flow-mediated dilation, carotid intima-media thickness, pulse wave velocity, and augmentation index, with age included in vascular clustering to account for vascular aging. Unsupervised k-means clustering was applied to identify vascular profiles. Functional performance was assessed using gait speed and predicted peak oxygen consumption (VO2peak) derived from a submaximal exercise test. Between-cluster differences were characterized using analysis of variance. Forty-five individuals with chronic stroke (62 ± 12 years; 29 ± 30 months post-stroke) were included. Three vascular profiles were identified (cluster 1: n = 13; cluster 2: n = 11; cluster 3: n = 21), with significant differences across clusters for all vascular measures (p < 0.01). Although age increased across clusters, vascular structure and function did not change uniformly. One profile exhibited elevated arterial wave reflection despite relatively preserved endothelial function and arterial structure, whereas another demonstrated impairment in all vascular variables. Vascular profile was not associated with predicted VO2peak (p = 0.31) or gait speed (p = 0.65). Three exploratory cluster-derived vascular profiles emerged among participants with chronic stroke, and age did not fully account for the observed vascular heterogeneity. Multi-domain cluster-derived vascular profiling revealed heterogeneous vascular profiles in individuals with chronic stroke, providing a basis for future studies to evaluate their relevance to cardiovascular risk stratification and long-term health management. NCT05936008: https://clinicaltrials.gov/study/NCT05936008
SN - 1932-6203
ER -