Distinct patterns of default mode network functional connectivity between adolescents with bipolar disorder and major depressive disorder
Source: PubMed Central Open Access, NCBI / U.S. National Library of Medicine
Background Bipolar disorder (BD) and major depressive disorder (MDD) exhibit overlapping clinical presentations, posing significant challenges for differential diagnosis and often leading to misidentification; therefore, elucidating the neural mechanisms that distinguish these two disorders is of critical importance. Methods In this study, 122 adolescents (43 with BD, 39 with MDD, and 40 healthy controls) completed resting-state functional magnetic resonance imaging (rs-fMRI). Voxel-level seed-based functional connectivity (FC) analysis using default mode network (DMN) subregions and machine learning classification were applied. Results Group-level analysis revealed that compared with healthy controls (HCs), patients with bipolar disorder (BD) exhibited significantly reduced FC between the anterior medial prefrontal cortex (aMPFC) and regions including the bilateral superior temporal gyrus (STG) and right temporal pole (TPOsup), as well as between the posterior inferior parietal lobule (pIPL) and the left middle temporal gyrus (MTG) and left STG. Relative to BD patients, patients with major depressive disorder (MDD) showed stronger FC between the aMPFC and left STG, and between the pIPL and right inferior frontal gyrus (IFG). No significant MDD–HC differences were detected in these circuits. Furthermore, no significant associations were found between altered DMN FC and clinical symptoms. Machine-learning analyses showed modest and unstable classification performance under nes
Abstract
Background Bipolar disorder (BD) and major depressive disorder (MDD) exhibit overlapping clinical presentations, posing significant challenges for differential diagnosis and often leading to misidentification; therefore, elucidating the neural mechanisms that distinguish these two disorders is of critical importance. Methods In this study, 122 adolescents (43 with BD, 39 with MDD, and 40 healthy controls) completed resting-state functional magnetic resonance imaging (rs-fMRI). Voxel-level seed-based functional connectivity (FC) analysis using default mode network (DMN) subregions and machine learning classification were applied. Results Group-level analysis revealed that compared with healthy controls (HCs), patients with bipolar disorder (BD) exhibited significantly reduced FC between the anterior medial prefrontal cortex (aMPFC) and regions including the bilateral superior temporal gyrus (STG) and right temporal pole (TPOsup), as well as between the posterior inferior parietal lobule (pIPL) and the left middle temporal gyrus (MTG) and left STG. Relative to BD patients, patients with major depressive disorder (MDD) showed stronger FC between the aMPFC and left STG, and between the pIPL and right inferior frontal gyrus (IFG). No significant MDD–HC differences were detected in these circuits. Furthermore, no significant associations were found between altered DMN FC and clinical symptoms. Machine-learning analyses showed modest and unstable classification performance under nested 10-fold cross-validation, with a pooled out-of-fold AUC of 0.609, accuracy of 0.598, sensitivity of 0.564, and specificity of 0.628. Conclusion Our results indicate that patients with MDD and BD exhibit distinct patterns of DMN connectivity with regions subserving sensory and cognitive processing, which may provide a potential neurobiological marker worthy of further investigation for discriminating between these disorders.