Background Non-neuropsychiatric systemic lupus erythematosus (non-NPSLE) has been confirmed to have subtle changes in brain structure before the appearance of obvious neuropsychiatric symptoms. Previous literature mainly focuses on brain structure loss in non-NPSLE; however, the results are heterogeneous, and the impact of structural changes on the topological structure of patients’ brain networks remains to be determined. In this study, we combined neuroimaging and network analysis methods to evaluate the changes in cortical thickness and its structural covariance networks (SCNs) in patients with non-NPSLE. Methods We compare the cortical thickness of non-NPSLE patients (N=108) and healthy controls (HCs, N=88) using both surface-based morphometry (SBM) and regions of interest (ROI) methods, respectively. After that, we analyzed the correlation between the abnormal cortical thickness results found in the ROI method and a series of clinical features. Finally, we constructed the SCNs of two groups using the regional cortical thickness and analyzed the abnormal SCNs of non-NPSLE. Results By SBM method, we found that cortical thickness of 34 clusters in the non-NPSLE group was thinner than that in the HC group. ROI method based on Destrieux atlas showed that cortical thickness of 57 regions in the non-NPSLE group was thinner than that in the HC group and related to the course of disease, autoantibodies, the cumulative amount of immunosuppressive agents, and cognitive psychological scale. In the SCN analysis, the cortical thickness SCNs of the non-NPSLE group did not follow the small-world attribute at a few densities, and the global clustering coefficient appeared to increase. The area under the curve analysis showed that there were significant differences between the two groups in clustering coefficient, degree, betweenness, and local efficiency. There are a total of seven hubs for non-NPSLE, and five hubs in HCs, the two groups do not share a common hub distribution. Conclusion Extensive and obvious reduction in cortical thickness and abnormal topological organization of SCNs are observed in non-NPSLE patients. The observed abnormalities may not only be the realization of brain damage caused by the disease, but also the contribution of the compensatory changes within the nervous system.

BackgroundCognitive dysfunction (CI) is frequently reported in patients with systemic lupus erythematosus (SLE), but the identification and assessment of SLE-related CI remain challenging. Previous studies have focused on changes in static brain activity, and no studies have investigated the characteristics of dynamic brain activity in SLE patients with CI.ObjectsWe calculated the dynamic amplitude of low-frequency fluctuation (dALFF) by combining the ALFF with a sliding window method to assess the temporal variability of brain functional activity in SLE patients with and without CI.MethodsThirty-eight SLE with CI, thirty-eight SLE without CI, and thirty-eight healthy controls (HCs) were recruited. By comparing static ALFF (sALFF) and dALFF among the three groups, changes in brain activity intensity and its temporal variability were assessed in patients with SLE with or without CI. Spearman correlation coefficients were calculated between the brain function indicator and Mini-mental State Examination (MMSE) scores of SLE with CI.ResultsSubjects among the three groups exhibited significant sALFF differences in the right parahippocampal gyrus, left caudate nucleus, right putamen, and left cuneus. Compared to the SLE without CI, the right parahippocampal gyrus exhibited higher sALFF in the SLE with CI group. Compared to the HCs, the left caudate nucleus exhibited increased sALFF in the SLE with CI group. Participants in the three groups exhibited significant dALFF variability in the right parahippocampal gyrus, right lingual gyrus, and bilateral inferior occipital gyrus. Compared to the HCs, the right lingual gyrus exhibited reduced dALFF in the SLE without CI group. Compared to the HCs, the right parahippocampal gyrus exhibited increased dALFF, left calcarine fissure, and the surrounding cortex exhibited reduced dALFF in the SLE with CI group. There was no significant correlation between the MMSE score, sALFF, and dALFF in the SLE with CI group.ConclusionSLE patients with CI have abnormal brain activity intensity and stability. By analyzing the dynamics of intrinsic brain activity, it provides a new idea for evaluating SLE-related CI. However, more research and validation with multiple metrics are needed to determine the link between the severity of cognitive impairment (CI) and brain activity in patients with SLE.

Background Non-neuropsychiatric systemic lupus erythematosus (non-NPSLE) has been confirmed to have subtle changes in brain structure before the appearance of obvious neuropsychiatric symptoms. Previous literature mainly focuses on brain structure loss in non-NPSLE; however, the results are heterogeneous, and the impact of structural changes on the topological structure of patients' brain network remains to be determined. In this study, we combined neuroimaging and network analysis methods to evaluate the changes in cortical thickness and its structural covariance networks (SCNs) in patients with non-NPSLE. Methods We compare the cortical thickness of non-NPSLE patients (N = 108) and healthy controls (HCs, N = 88) using both surface-based morphometry (SBM) and regions of interest (ROIs) method, respectively. After that, we analyzed the correlation between the abnormal cortical thickness results found in the ROIs method and a series of clinical features. Finally, we constructed the SCNs of two groups using the regional cortical thickness, and analyzed the abnormal SCNs of non-NPSLE. Results By SBM method, we found that cortical thickness of 34 clusters in the non-NPSLE group was thinner than that in the HCs group. ROIs method based on Destrieux altas showed that cortical thickness of 57 regions in the non-NPSLE group was thinner than that in the HCs group and related to the course of disease, autoantibodies, the cumulative amount of immunosuppressive agents, and cognitive psychological scale. In the SCNs analysis, the cortical thickness SCNs of the non-NPSLE group did not follow the small-world attribute at a few densities, and the global clustering coefficient appeared to increase. The area under the curve analysis showed that there were significant differences between the two groups in clustering coefficient, degree, betweenness, and local efficiency. There are a total of seven hubs for non-NPSLE, and five hubs in HCs, the two groups do not share a common hub distribution. Conclusion Extensive and obvious reduction in cortical thickness and abnormal topological organization of SCNs are observed in non-NPSLE patients. The observed abnormalities may not only be the realization of brain damage caused by the disease, but also the contribution of the compensatory changes within the nervous system.