Cerebrovascular reactivity in amnestic and non‐amnestic mild cognitive impairment

Background Cerebrovascular disease (CVD) is implicated in the pathogenesis and clinical presentation of Alzheimer’s Disease. Cerebrovascular reactivity (CVR) is considered a more direct estimate of cerebrovascular integrity. CVR is a potential promising biomarker for examining pathological cognitive...

Full description

Saved in:
Bibliographic Details
Published inAlzheimer's & dementia Vol. 19; no. S17
Main Authors Umfleet, Laura Glass, Cohen, Alexander D, Pommy, Jessica, Allen, Margaret, Obarski, Shawn, Mason, Lilly, Berres, Halle, Franczak, Malgorzata, Wang, Yang
Format Journal Article
LanguageEnglish
Published 01.12.2023
Online AccessGet full text

Cover

Loading…
More Information
Summary:Background Cerebrovascular disease (CVD) is implicated in the pathogenesis and clinical presentation of Alzheimer’s Disease. Cerebrovascular reactivity (CVR) is considered a more direct estimate of cerebrovascular integrity. CVR is a potential promising biomarker for examining pathological cognitive decline in older adults though, to date, findings have been inconclusive, possibly partly due to variability in methodological approaches. In this study, we add to the limited literature on CVR in mild cognitive impairment (MCI). We provide preliminary evidence for CVR differences in MCI phenotypes compared to older adult cognitively normal controls. We hypothesized that they amnestic phenotype (aMCI) will have lower CVR than non‐amnestic (naMCI) and both groups will show lower CVR compared to controls. We also performed exploratory correlational analyses between CVR and neuropsychological test scores. Method CVR was assessed in 22 controls, 11 aMCI, 9 naMCI cohorts using a breath‐holding task functional magnetic resonance imaging (fMRI). fMRI data were preprocessed and analyzed using AFNI. All participants underwent neuropsychological testing. T‐tests and ANOVA/ANCOVA analyses were conducted to compare controls to MCI groups on CVR metrics and cognitive tests of memory, language, processing speed, and executive functioning. Partial correlation analyses between CVR derived from regions‐of‐interest (Yeo 1‐7 ROIs) and different cognitive functions were conducted. Results CVR was found to be significantly lower in MCI compared to controls across all Yeo functional networks, ps < .001, which was mostly driven by the aMCI phenotype. naMCI showed intermediate patterns between aMCI and controls, though results were not significant, ps > .05. Regional CVR values along Yeo networks were positively correlated with cognitive measures of processing speed, executive functioning, and memory. Conclusion This is the first study to evaluate CVR differences in MCI phenotypes. Consistent with previous research, we found that CVR is lower in MCI compared to controls. While group differences between naMCI and controls (also between naMCI and aMCI) did not reach statistical significance, visual inspection of CVR mean values showed preliminary evidence that naMCI had lower mean values across all ROIs compared to controls and higher mean values compared to aMCI. Future research is needed to examine differences in CVR between MCI phenotypes.
ISSN:1552-5260
1552-5279
DOI:10.1002/alz.079025