article · Discover Neuroscience
Alzheimer’s disease (AD) is characterized by progressive, regionally heterogeneous cortical atrophy. Whether whole-brain cortical thickness homogeneity serves as an informative macro-scale marker of this structural disruption remains unclear. This study evaluated cortical thickness homogeneity across a large, multi-cohort sample and assessed its independent diagnostic utility. Structural MRI data comprising 11,382 scans from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) and the Open Access Series of Imaging Studies 3 (OASIS-3) were analyzed. Cortical thickness was extracted across 68 Desikan-Killiany regions. A cortical thickness homogeneity index-defined as the mean pairwise similarity of absolute regional thickness differences-was computed from within-dataset Z-scored values. Linear mixed-effects models with random intercepts for participant ID, adjusted for age and sex, compared diagnostic groups. Logistic regression evaluated added diagnostic value beyond medial temporal lobe (MTL) thickness. Statistical path decomposition evaluated the contribution of MTL atrophy. Dementia patients exhibited significantly lower cortical homogeneity than controls (total effect β = −0.035; LME-adjusted β = −0.029, Cohen’s d = 0.95, p < 0.001), independent of age and sex. The effect was most pronounced in the temporal lobe. The baseline diagnostic model (MTL thickness, age, sex) achieved an area under the curve (AUC) of 0.738 on external validation. Adding cortical homogeneity yielded an AUC of 0.746, a non-significant improvement (DeLong’s test, p = 0.198). Statistical path decomposition showed that MTL atrophy statistically accounted for 31.5% of the association between diagnosis and whole-brain homogeneity. Cortical thickness homogeneity captures macro-scale structural degradation in AD, showing partial overlap with regional temporal atrophy, but does not appear to provide meaningful incremental diagnostic value beyond MTL thickness or improve diagnostic accuracy beyond standard regional measures.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.1186/s13064-026-00315-z
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.