Decreased brain interstitial fluid dynamics is associated with risk of Alzheimer's disease-related cognitive decline.
Academic Article
Overview
abstract
BACKGROUND: Diffusion-tensor image analysis along the perivascular space (ALPS) index that has the potential to reflect brain interstitial fluid (ISF) dynamics may predict the development of Alzheimer's Disease (AD). We aimed to study whether brain ISF dynamics indicated by the ALPS index relate to AD dementia diagnosis and AD-related changes. METHODS: This study included a discovery cohort (n = 180) and a validation cohort (n = 127), which were composed of cognitively normal, subjective memory concern, mild cognitive impairment, and AD dementia subjects. All participants underwent brain magnetic resonance imaging examination and neuropsychological evaluation. The diffusivities and diffusion-tensor image analysis along the perivascular space (ALPS) were calculated. The support vector machine (SVM) model for AD dementia diagnosis was built in the discovery cohort and validated in the validation cohort. Linear mixed-effects models were used to evaluate the association between the ALPS and cognitive decline. Cox regression models were used to evaluate the association between the ALPS and the risk of AD dementia. RESULTS: There was a lower median ALPS index in the AD dementia group compared to other groups (all P < 0.05) for both cohorts. The SVM model for AD dementia diagnosis produced an AUC of 0.802 in the discovery cohort (P < 0.001) and 0.783 in the external validation cohort (P < 0.001). Higher ALPS levels were associated with less cognitive decline (P < 0.001). Moreover, lower baseline ALPS had a greater risk of converting to AD dementia (P = 0.014). CONCLUSIONS: The SVM model based on diffusivities and ALPS was effective for AD dementia diagnosis, and higher ALPS levels are associated with a lower risk of AD-related changes. These findings suggest that ALPS may provide a useful AD progression or treatment biomarker.
