Disrupted Brain Hierarchical Organization in Alzheimer's Disease Progression

Area of Science:

• Neuroscience
• Systems Neuroscience
• Computational Neuroscience

Background:

• Alzheimer's disease (AD) involves complex interactions between amyloid-β, tau, and neurodegeneration.
• The impact of these pathological processes on the brain's large-scale dynamic hierarchical organization remains poorly understood.

Purpose of the Study:

• To investigate how Alzheimer's disease alters the hierarchical organization of large-scale brain dynamics.
• To identify neuroimaging biomarkers for AD progression and cognitive decline.

Main Methods:

• Quantified directed causal interactions using resting-state fMRI in 134 participants across the AD continuum (healthy controls, mild cognitive impairment, AD dementia).
• Modeled brain activity with generative effective connectivity and assessed hierarchy using trophic levels and directedness.
• Employed machine learning to classify different stages of the AD continuum.

Main Results:

• Healthy individuals display a canonical brain hierarchy with distinct source, mediating (Default Mode Network), and sink regions.
• Alzheimer's disease patients show elevated visual network activity and reduced activity in somatomotor, salience, control, and DMN systems, indicating a flattened hierarchy.
• Mild cognitive impairment participants exhibited disruptions in somatomotor and dorsal attention networks.
• Machine learning successfully differentiated all disease stages, including early differences between healthy groups.
• Hierarchical alterations correlated with amyloid, tau, and neurodegeneration (ATN) biomarkers and cognitive decline.

Conclusions:

• Reduced hierarchical brain structure is a core systems-level signature of Alzheimer's disease.
• Trophic metrics derived from brain network analysis serve as sensitive biomarkers for AD progression.
• These findings suggest potential for early AD detection and targeted therapeutic interventions.