The role of glymphatic-related structural and functional imaging biomarkers in the pathological progression of Alzheimer’s disease: a multimodal imaging-based study

Abstract

Background The comparative utility of choroid plexus volume (CPV) and the diffusion tensor imaging-based perivascular space (DTI-ALPS) index as glymphatic biomarkers across the Alzheimer’s disease (AD) continuum is unclear. This study aimed to perform a head-to-head comparison of their relationships with AD pathology, cognition, and diagnostic performance. Methods This study analyzed data from 848 AD Neuroimaging Initiative (ADNI) participants [426 cognitively normal (CN), 309 with mild cognitive impairment (MCI), and 113 with AD dementia]. Group differences in CPV and the ALPS index were assessed using generalized linear models (GLM), and their associations with AD biomarkers and cognition were examined via partial correlation. Diagnostic performance using logistic regression, and longitudinal predictive value was assessed with linear mixed models. Results Compared with the CN and MCI groups, patients with AD exhibited significantly increased CPV and reduced ALPS indices (all P<0.001). Amyloid-β-positive (Aβ+) participants also showed significantly higher CPV and lower ALPS indices than Aβ-negative (Aβ−) individuals (all P<0.001). Elevated CPV was associated with lower cerebrospinal fluid (CSF) Aβ42 (r=−0.15, P<0.001), poorer Mini-Mental State Examination (MMSE; r=−0.12, P<0.001), Montreal Cognitive Assessment (MoCA; r=−0.17, P<0.001), and reduced hippocampal volume (r=−0.21, P<0.001). Conversely, higher ALPS indices were associated with increased CSF Aβ42 (r=0.12, P<0.001), better MMSE (r=0.13, P<0.001), and preserved hippocampal volume (r=0.15, P<0.001). CPV achieved superior diagnostic performance for differentiating AD from MCI [area under the curve (AUC) =0.774] and CN (AUC =0.912). Moreover, integrating CPV, ALPS, and hippocampal volume further improved classification performance across diagnostic and Aβ stratification tasks. Longitudinal analyses demonstrated that higher baseline CPV was associated with better baseline cognitive performance and slower decline in executive, language, and memory functions, whereas lower baseline ALPS indices predicted accelerated memory decline over time. Conclusions Our findings identify CPV and the ALPS index as a pair of promising, complementary neuroimaging biomarkers for AD. Their synergistic integration into diagnostic models improves the precision of early detection and intervention strategies.