Reduced structural connectivity of a major frontolimbic pathway in generalized anxiety disorder

Abstract

Context: Emotion regulation deficits figure prominently in generalized anxiety disorder (GAD) and in other anxiety and mood disorders. Research examining emotion regulation and top-down modulation has implicated reduced coupling of the amygdala with prefrontal cortex and anterior cingulate cortex, suggesting altered frontolimbic white matter connectivity in GAD.

Objectives: To investigate structural connectivity between ventral prefrontal cortex or anterior cingulate cortex areas and the amygdala in GAD and to assess associations with functional connectivity between those areas.

Design: Participants underwent diffusion-tensor imaging and functional magnetic resonance imaging.

Setting: University magnetic resonance imaging facility.

Participants: Forty-nine patients with GAD and 39 healthy volunteer control subjects, including a matched subset of 21 patients having GAD without comorbid Axis I diagnoses and 21 healthy volunteers matched for age, sex, and education.

Main outcome measures: The mean fractional anisotropy values in the left and right uncinate fasciculus, as measured by tract-based analysis for diffusion-tensor imaging data.

Results: Lower mean fractional anisotropy values in the bilateral uncinate fasciculus indicated reduced frontolimbic structural connectivity in patients with GAD. This reduction in uncinate fasciculus integrity was most pronounced for patients without comorbidity and was not observed in other white matter tracts. Across all participants, higher fractional anisotropy values were associated with more negative functional coupling between the pregenual anterior cingulate cortex and the amygdala during the anticipation of aversion.

Conclusions: Reduced structural connectivity of a major frontolimbic pathway suggests a neural basis for emotion regulation deficits in GAD. The functional significance of these structural differences is underscored by decreased functional connectivity between the anterior cingulate cortex and the amygdala in individuals with reduced structural integrity of the uncinate fasciculus.

Figure 1

Region of interest (ROI) placement for delineation of bilateral uncinate fasciculus. Bilateral frontal and temporal lobe seed ROIs (a) were manually drawn in the most posterior coronal slice that showed clear separation of the frontal and temporal lobes bilaterally, as depicted in (b). The Boolean ‘AND’ term was used to select only fibers that crossed through both the temporal and frontal seed ROIs for tract-based analysis. (c) Uncinate fasciculus tracts overlaid on anatomical T1-weighted image for a single subject. For 3D rendering, see eVideo 1 online.

Figure 2

Patients with generalized anxiety disorder (GAD) showed reduced mean fractional anisotropy (FA) values for the uncinate fasciculus (UF) relative to healthy control subjects, as indicated by a Group main effect for a Group × Hemisphere ANCOVA with covariates of age, sex, education, and whole-brain FA.

Figure 3

(a) Participants with higher uncinate fasciculus fractional anisotropy (FA) values showed greater negative coupling between the pregenual anterior cingulate cortex (pgACC) and amygdala during the anticipation of aversive vs. neutral pictures. (b) This relationship was seen in both patients with generalized anxiety disorders (red squares) and healthy controls (blue diamonds), as evidenced by a significant pgACC cluster for the FA predictor but not the Group or Group × FA predictors. Note that r values are presented for illustrative purposes only and are not intended as depicting additional statistical tests.^,