Increased brain entropy of resting-state fMRI mediates the relationship between depression severity and mental health-related quality of life in late-life depressed elderly

Abstract

Background: Entropy analysis is a computational method used to quantify the complexity in a system, and loss of brain complexity is hypothesized to be related to mental disorders. Here, we applied entropy analysis to the resting-state functional magnetic resonance imaging (rs-fMRI) signal in subjects with late-life depression (LLD), an illness combined with emotion dysregulation and aging effect. Methods: A total of 35 unremitted depressed elderly and 22 control subjects were recruited. Multiscale entropy (MSE) analysis was performed in the entire brain, 90 automated anatomical labeling-parcellated ROIs, and five resting networks in each study participant. Limitations: Due to ethical concerns, all the participants were under medication during the study. Results: Regionally, subjects with LLD showed decreased entropy only in the right posterior cingulate gyrus but had universally increased entropy in affective processing (putamen and thalamus), sensory, motor, and temporal nodes across different time scales. We also found higher entropy in the left frontoparietal network (FPN), which partially mediated the negative correlation between depression severity and mental components of the quality of life, reflecting the possible neural compensation during depression treatment. Conclusion: MSE provides a novel and complementary approach in rs-fMRI analysis. The temporal-spatial complexity in the resting brain may provide the adaptive variability beneficial for the elderly with depression. © 2019