Context:

Alzheimer’s disease (AD) often co-occurs with multiple brain aging
pathologies affecting cognition. However, the relative contribution
and interaction of neurodegeneration, small vessel disease, and
glymphatic dysfunction remain poorly understood.

Objective:

We aimed to investigate the independent and relative associations
of MRI biomarkers from these systems with cognitive performance
along the AD spectrum.

Methods:

MRI analysis:
 - Gray Matter (GM) volume segmented on T1w-MPRAGE image
using MRI synthseg1 from FreeSurfer
 - White Matter Hyperintensities (WMH) segmented on T2-FLAIR
using FLAMeS2
 - Choroid Plexus (ChP) segmented on T1w-MPRAGE using Deep
Learning UNet3

Cognitive scores:
Mini-Mental State Examination (MMSE), and composite z-scores for
memory, language, visuospatial, executive functions and global
cognition

We included: 
85 amyloid-positive individuals spanning 
 - Cognitively normal (CN — 8), 
 - Mild cognitive impairment (MCI — 55), 
 - Alzheimer’s disease dementia (ADD  — 22) stage

Results:

Univariable analysis+:
 - GM is the most consistent predictor of
cognitive performance
 - WMH shows marginal association with
executive performance 
 - ChP demonstrate broad association pattern

Multi-variable analysis+:
 - GM remained the dominant predictor
 - WMH showed domain-specific effects
 - ChP demonstrated weaker independent
associations

Conclusion:

 - Among the MRI measures examined, Cortical volume loss showed the strongest association with
cognitive performance across domains, except for executive function 
 - Vascular pathology contributes selectively, with a stronger impact on executive dysfunction
 - Glymphatic dysfunction has broad univariable effects but weaker independent contribution,
highlighting its likely indirect or shared role with broader brain aging processes

Overall, results supports a multi-system framework, with neurodegeneration as the dominant
clinical determinant