Temporal dynamics of spontaneous default-mode network activity mediate the association between reappraisal and depression

Abstract

Cognitive reappraisal is associated with major depressive disorder (MDD), while spontaneous activity patterns of the default mode network (DMN) is implicated in reappraisal and MDD. However, neural mechanisms subserving the close association of spontaneous reappraisal and depression are unclear. Spontaneous reappraisal, depression and resting-state functional magnetic resonance imaging (rsfMRI) were measured from 105 healthy subjects. We assessed the temporal complexity (Hurst exponent), Regional Homogeneity (ReHo) and fractional Amplitude of Low Frequency Fluctuation (fALFF) profiles of DMN, a network involved in both reappraisal and depression. Mediation effects of these standard measures on the relationship between reappraisal and depression, and the contributions of each DMN subregion, were assessed. Results indicated that Hurst exponent (H) of DMN, whether extracted by independent component analysis (ICA) or region of interest (ROI), was significantly associated with reappraisal scores. An individual with a higher reappraisal score has a lower Hurst value of DMN. Mediation analyses suggest that H of DMN partially mediates the association between reappraisal and the degree of depression, and this mediation effect arises from the contribution of medial prefrontal cortex. Neither ReHo nor fALFF showed a similar correlation or mediation effect. These findings suggest that temporal dynamics of DMN play an important role in emotion regulation and its association with depression. H of DMN may serve as a neural marker mediating the association between reappraisal and depression.

Fig. 1

The power spectra of all subjects without band-pass filtering from the cumulative sum of the time courses of ICA (red) and region of interests (ROIs; green). The scale parameters were chosen to correspond to a frequency range of (0.016, 0.063) Hz.

Fig. 2

The spatial patterns of DMN extracted by ICA and ROI. DMN extracted by group ICA (Top) and DMN extracted by ROI seeded in PCC region (Bottom). Axial map was shown to be superimposed on the MNI152 standard space template image. The statistical thresholds of both spatial patterns were P < 0.05 (FDR-corrected, 20 adjacent voxels).

Fig. 3

Spearman rank-correlations and scatter-plots displaying the relationship between the scores of reappraisal and depression scores. Significant negative correlation between reappraisal and depression scores, r (105) = −0.288; P < 0.01.

Fig. 4

Spearman rank correlations and scatter plots displaying the relationship between the scores of reappraisal and the Hurst exponents in DMN_ICA(A) and in DMN_ROI(B). Significant negative correlation between reappraisal scores and H of DMN_ICA, r (105) = −0.311; P < 0.01; H of DMN_ROI, r (105) = −0.294; P < 0.01.

Fig. 5

The relationship between reappraisal and depression scores was mediated by H of DMN. The lines are labeled with path coefficients and standard errors are shown in parentheses. The direct path between spontaneous reappraisal and depression is calculated controlling for H. Indirect path a = −0.31136, indirect path b = 0.2289, total relationship c = −0.3078 and direct path c = −0.2360. The values in parentheses indicate the strength of the path prior to the inclusion of the mediating variable, ^*P < 0 .05.