The Alterations in Brain Network Functional Gradients and Dynamic Functional Connectivity in Alzheimer's Disease:A Resting-State fMRI Study

Benqin Liu¹Chunbing Chen¹Pan Cai²Jiaren Zhang¹Yang Li¹Xu Chen¹Xuejin Ma¹Xuetao Jiang¹anjie zhang¹Linfeng Song^1*Lin Jiang^1*

• ¹Department of Radiology, the Third Affiliated Hospital of Zunyi Medical University (the First People’s Hospital of Zunyi), zunyi, China
• ²Department of Traditional, The Third Affiliated Hospital of Zunyi Medical University (The First People’s Hospital of Zunyi), zunyi, China

Background and Purpose: Alzheimer’s disease (AD) is characterized by progressive cognitive decline and large-scale brain network dysfunction. Dynamic functional connectivity (dFC) studies have consistently reported unstable functional states, reduced flexibility, and impaired transitions across the AD continuum. However, how these temporal abnormalities are embedded within the hierarchical spatial organization of brain networks, as captured by functional gradients (FG), remains unclear. This study integrated FG and dFC analyses to characterize hierarchical and dynamic network alterations in AD. Methods:Forty-six AD patients diagnosed using the Amyloid/Tau/Neurodegeneration (ATN) framework and 37 age- and sex-matched healthy controls underwent resting-state fMRI. Functional gradients were derived using diffusion embedding, and dFC was estimated with a sliding-window approach and clustered into four recurrent states. Group differences were assessed using Gaussian Random Field–corrected t tests. Correlations with ATN biomarkers and age were examined. A linear support vector machine (SVM) with leave-one-out cross-validation evaluated classification performance. Results:Compared with controls, AD patients showed widespread FG alterations predominantly along the principal gradient. Within the default mode network, gradient scores were reduced in the left precuneus but increased in the right medial superior frontal gyrus and bilateral angular gyri. Within the sensorimotor network, the right supplementary motor area showed increased scores, whereas bilateral superior temporal gyri were decreased. Second-gradient abnormalities were confined to the left postcentral gyrus and left middle occipital gyrus. Medial superior frontal gradient scores correlated negatively with T-Tau and age. dFC revealed four recurrent states; AD patients showed increased dwell time and fractional occupancy in states 1–2 and reduced values in state 3, accompanied by state-dependent bidirectional connectivity changes. The SVM achieved an AUC of 0.776. Conclusions:AD is characterized by hierarchical disorganization centered on the principal functional gradient and reduced dynamic network flexibility. Integrated FG–dFC analysis provides complementary spatiotemporal insights and identifies network features with potential translational value for disease monitoring and intervention studies.