Neurodegenerative diseases do not erode the brain uniformly. Instead, they follow highly structured spatial and temporal trajectories, moving through neural circuits like a slow-moving wave. By mapping these patterns, researchers have moved beyond treating atrophy as a global metric, instead viewing it as a predictable sequence of regional failures that track clinical decline.
The Directional Flow of Atrophy
One of the most robust findings across different pathologies is the existence of spatial gradients—directional axes along which tissue loss progresses. In conditions like Parkinson’s disease and Progressive Supranuclear Palsy (PSP), this progression often follows an anterior-to-posterior spread.
Claassen et al. demonstrated that in Parkinson’s, cortical thinning initiates in the frontal lobes, which are responsible for motor and executive functions, before accelerating toward the occipital cortex in later stages. A similar pattern was identified by Scotton et al. in PSP, where atrophy begins in the insula and frontal regions and systematically extends to the temporal, parietal, and finally the occipital lobes.
Other diseases exhibit different geometric signatures. Ciarochi identified a superior-to-inferior gradient in Huntington’s disease, where gray matter loss starts in the upper regions of the frontal and parietal lobes and moves downward toward the base of the brain as the disease approaches clinical onset. These gradients suggest that the physical "flow" of neurodegeneration is a fundamental property of the disease mechanism.
Hierarchies of Regional Vulnerability
Within these broad gradients, specific structures act as early epicenters of decay. In Alzheimer’s disease (AD), the temporal hierarchy is particularly well-defined. Zhou et al. noted that atrophy in the entorhinal cortex typically precedes hippocampal changes by nearly a decade, serving as a preclinical marker. Even within the hippocampus itself, the involvement is sequential. Ho et al. found that in schizophrenia, volume loss is initially localized to the CA1 subfield before spreading to the entire structure as the illness becomes chronic.
This hierarchical spread is not limited to structural loss; it is mirrored by protein accumulation. Macedo et al. confirmed that in vivo tau PET imaging aligns with the Braak staging framework, showing that pathology climbs through a predictable ladder of brain regions. Kang et al. further refined this "atrophy atlas," showing a progression from memory centers (hippocampus and amygdala) outward to the temporal and frontal lobes.
From Localized Failure to Network Collapse
Why does pathology move in these specific patterns? The evidence suggests that neurodegeneration spreads along established anatomical communication pathways. Oxtoby et al. used data-driven modeling to show that Alzheimer’s begins with local connectivity failures in the temporal lobe, which then cascade into a systematic collapse of the brain's "wiring" toward the frontal and occipital regions.
As this spread occurs, the relationship between different biological markers changes. Park observed that the correlation between tau protein buildup and actual tissue loss strengthens and expands geographically as the disease advances. In early stages, these markers may appear disconnected or localized to limbic regions, but in late-stage disease, Simpson et al. found that structural atrophy and functional hypometabolism converge, creating a widespread spatial overlap of pathology.
Spatial Expansion as a Clinical Metric
Understanding these spatial hierarchies has significant diagnostic implications. Venkataraman et al. argued that the spatial expansion of pathological clusters—how much new territory the disease has claimed—is often a more sensitive indicator of clinical decline than the absolute intensity of the damage in a single area.
By mapping the "where" and "when" of brain decay, researchers can better predict the "what"—the specific cognitive and behavioral deficits a patient will face. As Henry et al. and Oh et al. demonstrated, the expansion of pathology into new anatomical stages triggers corresponding functional declines, allowing clinicians to use spatial imaging as a roadmap for disease progression.
Key Foundational Work
Baseline MRI Predictors of Conversion from MCI to Probable AD in the ADNI Cohort (2009) — contributed 1 concept elements to this direction.
A tensor based morphometry study of longitudinal gray matter contraction in FTD (2007) — contributed 1 concept elements to this direction.