Multi-dimensional Functional Brain Connectome Signatures In Occupational Burnout

Occupational burnout has become a serious public health issue globally,particularly for healthcare workers who have been under immense work pressure during the COVID-19 pandemic Functional magnetic resonance imaging(f MRI)reveals that similarities between spontaneous neural activation patterns during rest characterize the functional connectivity of the brain,forming an efficient and complex brain network.While previous research has suggested a relationship between occupational burnout and anatomical as well as functional abnormalities in specific brain regions,these studies have been limited to local changes in the brain(i.e.,based on isolated brain regions),ignoring the brain’s full-scale patterns of connectivity.In addition,genetic factors play an important role in shaping the organization structure of the brain in occupational burnout.Therefore,in what form do the abnormal fluctuations of neural activity in occupational burnout induces the instability of the large-scale functional network and the underlying molecular mechanisms need to be solved urgently.Advances in neuroimaging-transcriptomic analysis provide unprecedented opportunities to understand the microscopic molecular basis underlying macroscopic brain functional abnormalities.Therefore,this study uses brain network technology,based on rs-f MRI data,gene expression data,and clinical data,to characterize the abnormal functional brain network mechanisms in occupational burnout from a multi-dimensional perspective.The main contributions of this research are as follows:First,the functional network based on Pearson correlation coefficient and the effective network based on Granger causality analysis(GCA)were respectively constructed.We employed graph theoretical and network-based statistic(NBS)methods to analyze the functional network,and the NBS approach was used to explore connectivity difference in effective network.Finally,partial correlation analyses were conducted between clinical scores and the metrics with significant differences.Results showed that although the “small world” architecture was retained in the burnout group,participants with burnout showed significantly increased shortest path length and decreased global efficiency,indicating that its brain functional network shifted to a relatively regular network,revealing the global integration disruption of the functional network of participants with burnout.Second,nodal metrics increased mainly in visual and auditory network regions,suggesting their strengthened effects in coordinating the work of large networks as a compensatory response to the deficits in integrating visual and audio information,revealing the essential function of primary sensory cortex in the pathogenesis of burnout.Results in the directed networks showed decreased connectivity mainly between the visual network and the right hippocampus,supporting the concept of breakdown in the process of extracting visual information from memory.Finally,correlation analysis showed that these abnormalities may be intrinsic features of occupational burnout and are not related to the severity of symptoms.Second,a similarity network was first constructed based on the Pearson correlation coefficient and cosine similarity function.Then,we introduced a non-linear dimensionality reduction technique based on manifold learning(diffusion embedding technology)to derive the full-cortex functional gradient network.The between-group differences were determined by cortex-wide multivariate analyses spanned by the first three gradients which together explained approximately 34% of the total variance.We observed macroscale distortions in occupational burnout in multiple networks,with primary effects in sensorimotor and transmodal cortices,especially in the postcentral gyrus involving somatosensory centers and the precentral gyrus of somatomotor centers,and with hemispheric asymmetry.After controlling for the effect of morphological changes(cortical thickness),results were broadly similar.When simplifying multivariate gradient space into a scalar metric(eccentricity)for group comparisons,we found evident expansions in somatomotor cortices in individuals with burnout relative to controls.We applied multiple thresholds to functional connectomes to perform validation analyzes,and group-level gradient patterns were found to be highly spatially correlated across all thresholds.Finally,macroscale distortions in the posterior cingulate cortex,medial temporal lobe,and visual cortex were significantly associated with burnout severity.Our findings provide new evidence for imbalances in network hierarchy in occupational burnout,which offers a parsimonious reference frame to consolidate the deficits of sensory and cognitive processing function.By establishing a fusion model of brain neuroimaging and transcriptomics,we discovered the relationship between the macroscopic functional hierarchy perturbations in occupational burnout and brain gene expression.Specifically,the gradient difference t map was linearly fitted with the gene expression data obtained from the AHBA dataset using the partial least squares(PLS)method.These significant gene subsets were subjected to gene enrichment analysis and cell-type-specific gene expression analysis.Gene enrichment analysis revealed gradient changes associated with gene expression profiles involved in synaptic transmission,cellular response to metal ion,mineral absorption,and circadian rhythm.Furthermore,gene expression in excitatory neurons,inhibitory neurons,and astrocyte cells was spatially associated with abnormal patterns of burnout.These results advance the understanding of the underlying neurobiological information of macroscale dysconnectivity and may provide new clues for exploring the biological etiology of burnout.