Objective White matter hyperintensities (WMHs), a hallmark of cerebral small vessel disease, frequently coexist with Alzheimer's disease (AD) and facilitate cognitive deterioration. Juxtaventricular WMH (JVWMH) is hypothesized to reflect pathological processes at the cerebrospinal fluid (CSF)-parenchyma interface and hold particular clinical significance. Therefore, this study specifically investigated associations between JVWMH burden and cognitive performance, CSF volume (CSFV), and amyloid-β (Aβ) pathology. Materials and methods Automated WMH segmentation was applied in two cohorts: 295 from the Australian Imaging, Biomarkers and Lifestyle (AIBL) study and 82 from a memory clinic. All participants underwent 3-T magnetic resonance imaging, amyloid-positron emission tomography, and cognitive assessment. Analyses evaluated JVWMH associations with Aβ, its discriminative value for cognitive impairment (CI) versus cognitively normal (CN), prediction of longitudinal decline, and mediation of CSFV-cognition relationships. Results JVWMH volume significantly discriminated CI from CN participants in both cohorts and predicted cognitive decline in longitudinal AIBL data. JVWMH volume showed strong correlations with CSFV in both cohorts. Notably, JVWMH partially mediated the associations between CSFV and cognition in the AIBL cohort. Non-juxtaventricular WMHs and Fazekas scores demonstrated no significant diagnostic or predictive power. Conclusion JVWMH volume provides diagnostic and prognostic information beyond conventional WMH metrics in AD. Its correlation with CSFV implicates CSF-parenchyma interface processes, though causality remains unproven. These findings highlight the importance of incorporating spatially stratified WMH metrics in AD research to better capture disease-relevant white matter pathology. Key Points <UnorderedList Mark="Bullet"> <ItemContent> Question: JVWMH volume provides diagnostic and prognostic value beyond conventional WMH metrics in AD. </ItemContent> <ItemContent> Findings: JVWMH volume discriminated CI from normal cognition, predicted longitudinal decline, correlated with CSF volume, and partially mediated CSF volume–cognition associations. </ItemContent> <ItemContent> Relevance statement: JVWMH volume provides diagnostic and prognostic information beyond conventional WMH metrics in AD, highlighting the value of spatially stratified WMH analysis in AD research to better capture disease-relevant white matter pathology. </ItemContent> </UnorderedList> Graphical Abstract

Topography of Juxtaventricular white matter hyperintensities and cognitive associations in Alzheimer's disease: a dual-cohort study

期刊论文
出版模式：

开放获取
发表日期：

2026年7月29日

查看全文

Yi-Wen Bao,
Ya-Qing Ji,
Li-Na Wang,
Yu Zhou,
Fen Wang,
Da-Ming Shen,
Dmytro Pylypenko,
Qiang Tong,
Ka-Fung Mak,
Li-Li Guo

摘要

Objective

White matter hyperintensities (WMHs), a hallmark of cerebral small vessel disease, frequently coexist with Alzheimer's disease (AD) and facilitate cognitive deterioration. Juxtaventricular WMH (JVWMH) is hypothesized to reflect pathological processes at the cerebrospinal fluid (CSF)-parenchyma interface and hold particular clinical significance. Therefore, this study specifically investigated associations between JVWMH burden and cognitive performance, CSF volume (CSFV), and amyloid-β (Aβ) pathology.

Materials and methods

Automated WMH segmentation was applied in two cohorts: 295 from the Australian Imaging, Biomarkers and Lifestyle (AIBL) study and 82 from a memory clinic. All participants underwent 3-T magnetic resonance imaging, amyloid-positron emission tomography, and cognitive assessment. Analyses evaluated JVWMH associations with Aβ, its discriminative value for cognitive impairment (CI) versus cognitively normal (CN), prediction of longitudinal decline, and mediation of CSFV-cognition relationships.

Results

JVWMH volume significantly discriminated CI from CN participants in both cohorts and predicted cognitive decline in longitudinal AIBL data. JVWMH volume showed strong correlations with CSFV in both cohorts. Notably, JVWMH partially mediated the associations between CSFV and cognition in the AIBL cohort. Non-juxtaventricular WMHs and Fazekas scores demonstrated no significant diagnostic or predictive power.

Conclusion

JVWMH volume provides diagnostic and prognostic information beyond conventional WMH metrics in AD. Its correlation with CSFV implicates CSF-parenchyma interface processes, though causality remains unproven. These findings highlight the importance of incorporating spatially stratified WMH metrics in AD research to better capture disease-relevant white matter pathology.

Key Points

Question: JVWMH volume provides diagnostic and prognostic value beyond conventional WMH metrics in AD.

Findings: JVWMH volume discriminated CI from normal cognition, predicted longitudinal decline, correlated with CSF volume, and partially mediated CSF volume–cognition associations.

Relevance statement: JVWMH volume provides diagnostic and prognostic information beyond conventional WMH metrics in AD, highlighting the value of spatially stratified WMH analysis in AD research to better capture disease-relevant white matter pathology.

Graphical Abstract