RESUMEN
Chronic pain has been linked with learning and memory processes and functional changes in brain plasticity in its development and maintenance via neuroimaging studies. However, the principle of reorganization of the migraine brain network as the brain progresses into chronic pain remain poorly understood. Here, using resting-state functional magnetic resonance imaging (rs-fMRI) and graph theory approaches, we aimed to investigate the dynamic dysfunctional connectivity in 108 patients with migraine without aura (MWoA) and 30 gender-matched healthy controls (HC). All patients were divided into 40 groups using a sliding boxcar grouping of subjects in disease duration order. As compared with HC, nonparametric permutation tests were applied for between-group comparisons of functional connectivity strength in each patient group. We focused only on the between-group differences of functional connections in MWoA, and the situation how these different connections were organized along with the changing trend. As the disease duration increased, the presence of chronic headache altered the functional connectivity from the local central nervous system (CNS) to a disruption in the whole-brain networks. These dysfunctional connections integrated into a connected component in relatively longer migraine duration groups, suggesting an abnormal integrated network configuration with ongoing central changes for long-term migraine. Within these between-group differences of the connected component, there were contained a small number of brain regions that had disproportionately numerous connections. Moreover, these brain regions exhibited a tendency to link to each other were organized into a strongly interconnected community. These interconnected brain regions were mainly located in the sensory-discriminative brain areas. Our results exhibited a working model of the central mechanisms of migraine where the brain functional connectivity was altered from the local central nervous system to a densely interconnected center, which may extend our understanding of the role of learning mechanisms which are likely involved in maintenance of chronic pain.