Graph theory analysis reveals the topological property changes of the gray matter structure covariant network in patients with primary dysmenorrhea and its association with pain-emotion regulation

TIAN Xin; WEI Wei; FAN Junfeng; ZHOU Feng; ZHENG Yunsong; FAN Lihua

doi:10.12122/j.issn.1674-4500.2025.12.03

Abstract：Objective To systematically evaluate alterations in the structural covariance networks(SCNs)of primary dysmenorrhea(PDM)patients using structural magnetic resonance imaging and graph theory analysis.Methods A total of 31 PDM patients and 30 healthy controls(HCs)were recruited from Shaanxi University of Chinese Medicine from September 2021 to September 2022.High-resolution T1-weighted imaging was conducted on days 1-3 of the menstrual cycle.Structural images were preprocessed using SPM8.Gray matter SCNs were reconstructed utilizing the graph analysis toolbox(GAT),and global and local network metrics were computed and compared between groups based on graph theory.Results Compared to the HC group,PDM patients exhibited trends of decreased clustering coefficient(Cp),local efficiency,and transitivity,alongside increased assortativity in global network metrics.However,these differences were not statistically significant when compared against the null distribution derived from permutation tests(P>0.05).The area under the curve(AUC)results for global metrics also indicated no significant intergroup differences(P>0.05),suggesting the overall architecture of the brain network remains relatively intact in PDM patients.At the local nodal level,under the minimum density threshold,PDM patients showed significantly altered nodal metrics prior to FDR correction(P<0.05):decreased nodal degree was observed in the left cuneus,left superior occipital gyrus,and right postcentral gyrus;increased nodal degree was found in the bilateral middle cingulate gyrus and right middle frontal gyrus;betweenness centrality was decreased in the left frontal inferior operculum and increased in the left cerebellum,cingulate gyrus,and olfactory cortex;Cp was decreased in the left cerebellum and increased in the left postcentral gyrus.These differences were no longer statistically significant after FDR correction(P>0.05).AUC analysis of standardized nodal metrics revealed that PDM patients had increased nodal degree in the bilateral cingulate gyrus,left olfactory cortex,and right parahippocampal gyrus;decreased nodal degree in the left cuneus,left middle occipital gyrus,left superior occipital gyrus,and right postcentral gyrus;increased betweenness centrality in the left middle cingulate gyrus,left olfactory cortex,right parahippocampal gyrus,and right precentral gyrus;decreased betweenness centrality in the left orbital inferior frontal gyrus,right orbital middle frontal gyrus,and right insula;increased Cp in the right orbital middle frontal gyrus;and decreased Cp in the left middle occipital gyrus and right superior occipital gyrus.Local efficiency was significantly higher in the right orbital middle frontal gyrus and lower in the right superior occipital gyrus in the PDM group(P<0.05).Analyses using both targeted and random network attacks demonstrated no significant differences in the size of the largest component of the residual network between the two groups(P>0.05).AUC results for network attack metrics also showed no significant differences(P>0.05).Conclusion Alterations in both global and nodal metrics of SCNs were observed in PDM patients,primarily involving brain regions associated with pain processing and emotion regulation,which may contribute to the manifestation of dysmenorrhea symptoms.Although the overall stability of the brain network appears largely preserved,these topological changes provide important neuroimaging evidence for the central mechanisms underlying PDM.The findings could aid in identifying potential central markers for PDM in clinical diagnosis(e.g.,local efficiency in the right orbital middle frontal gyrus)and inform the development of novel neuroregulatory treatment strategies targeting pain-related brain regions.