Interoception in insula subregions as a possible state marker for depression-an exploratory fMRI study investigating healthy, depressed and remitted participants

Abstract

Background: Interoceptive awareness (iA), the awareness of stimuli originating inside the body, plays an important role in human emotions and psychopathology. The insula is particularly involved in neural processes underlying iA. However, iA-related neural activity in the insula during the acute state of major depressive disorder (MDD) and in remission from depression has not been explored.

Methods: A well-established fMRI paradigm for studying (iA; heartbeat counting) and exteroceptive awareness (eA; tone counting) was used. Study participants formed three independent groups: patients suffering from MDD, patients in remission from MDD or healthy controls. Task-induced neural activity in three functional subdivisions of the insula was compared between these groups.

Results: Depressed participants showed neural hypo-responses during iA in anterior insula regions, as compared to both healthy and remitted participants. The right dorsal anterior insula showed the strongest response to iA across all participant groups. In depressed participants there was no differentiation between different stimuli types in this region (i.e., between iA, eA and noTask). Healthy and remitted participants in contrast showed clear activity differences.

Conclusions: This is the first study comparing iA and eA-related activity in the insula in depressed participants to that in healthy and remitted individuals. The preliminary results suggest that these groups differ in there being hypo-responses across insula regions in the depressed participants, whilst non-psychiatric participants and patients in remission from MDD show the same neural activity during iA in insula subregions implying a possible state marker for MDD. The lack of activity differences between different stimulus types in the depressed group may account for their symptoms of altered external and internal focus.

Keywords: fMRI; hopelessness; insula; interoception; interoceptive awareness; major depressive disorder; neuroimaging; remission.

Figure 1

Neural activity during interoceptive awareness (iA) in the left (L) and right (R) dorsal anterior insula (dAI), ventral anterior insula (vAI) and the posterior insula (PI) in groups of healthy (H), depressed (D) and remitted (R) participants. (A) Illustration of regions of interest in the left and right insula (L-dAI: blue, L-vAI: cyan, L-PI: purple, R-dAI: orange, R-vAI: red, R-PI: yellow) (provided by Deen et al., 2011). (B) Mean iA-related BOLD responses in insula subregions. Data points within columns represent healthy (H), depressed (D) and remitted (R) groups. In the left hemisphere, the PI (purple, square symbols) shows significant differences between healthy and depressed participants. In the right hemisphere, the depressed group differs compared to the healthy group in all regions, whereas differences to the remitted group occur in the dAI (orange, triangular symbols) and vAI (red, diamond symbols). Healthy and remitted groups show no differences. [^***P < 0.005, ^**P < 0.01, ^*P < 0.05, (^*) P < 0.1, post-hoc Bonferroni] (C) Results shown in (B) are re-represented here across groups (healthy: black line, depressed: red line, remitted: green line). Columns represent insula subregions. Healthy and remitted participants show identical neural activation during iA, whilst depressed patients show the lowest activity in each region. Bilateral regions of the dAI show the highest degree of iA-related activity in each study group. In the right hemisphere, this activity linearly decreases with the PI showing the lowest degree of neural activity. (D) Voxel-wise results within the bilateral insula comparing the different groups during iA performance (P = 0.05 FWE-corrected, Z threshold = 3.1, gray and white matter volumes included as confound). Healthy (left image) and remitted (middle image) participants show increased iA-related BOLD responses compared to depressed participants. Healthy and remitted groups show less extensive differences (right image), with increased BOLD responses in the inferior insula in healthy individuals.

Figure 2

BOLD responses (mean ± SEM) during different task conditions in healthy (H), depressed (D) and remitted (R) participants in the left and right dorsal anterior insula (dAI). Interoceptive awareness (iA) is marked in yellow, exteroceptive awareness (eA) is marked in purple and no particular task (noTask) in orange. Healthy and remitted participants show a clear distinction between iA and specifically noTask conditions in both regions. Additionally, iA-related BOLD responses differ to eA-related BOLD responses in both groups and regions, whereas the R-dAI reveals also a differentiation between eA- and noTask-related BOLD responses in the non-psychiatric group. Depressed participants show no differentiation in both regions between any of the three conditions. [^***P < 0.0005, ^**P < 0.01, ^*P < 0.05, (^*) P ≤ 0.1, post-hoc Bonferroni].

Figure 3

Detailed overview of the distribution of neural activity in the left and right dorsal anterior insula (dAI) across groups for different conditions (awareness toward internal stimuli on the left side, awareness toward external stimuli in the middle and awareness toward no particular task on the right side). Black bars indicate the non-psychiatric group, red bars indicate the depressed group and green bars indicate the remitted group.

Figure 4

Visualization of the fMRI paradigm to study interoceptive awareness. Each condition contained both stimuli types: external tone and internal heartbeat were concurrently ongoing events throughout a scanning session. The different conditions were matched as closely as possible and participants had to direct their awareness either to internal, external or no stimuli.

Figure 5

Mean BOLD responses during interoceptive awareness (iA) across four scanning sessions in the left and right dorsal anterior insula (dAI). Graphs show responses for healthy (black line, diamond symbols), depressed (red line, squared symbols) and remitted participants (green line, triangular symbols). No group differences occurred over the time course between the four functional scanning sessions.

Figure 6

Neural activity during exteroceptive awareness (eA) in the right posterior insula in groups of healthy (H), depressed (D) and remitted (R) participants and its relation to scores of the Beck Hopelessness Scale (BHS). (A) Illustration of right posterior insula region of interest (indicated in pink). This region showed significant between-subjects effects during eA as revealed by MANOVA (see Table 2B). (B) BOLD responses during eA (mean ± SEM) show differences between healthy and depressed participants in the R-PI (^*P < 0.05). (C) BHS scores correlate negatively (Pearson, two-tailed) with signal changes for eA in healthy participants (^**P < 0.01).