Mind-body interactions in the regulation of airway inflammation in asthma: A PET study of acute and chronic stress

Abstract

Background: Psychological stress has long been recognized as a contributing factor to asthma symptom expression and disease progression. Yet, the neural mechanisms that underlie this relationship have been largely unexplored in research addressing the pathophysiology and management of asthma. Studies that have examined the mechanisms of this relationship in the periphery suggest that it is the superimposition of acute stress on top of chronic stress that is of greatest concern for airway inflammation.

Methods: We compared asthmatic individuals with high and low levels of chronic life stress in their neural and peripheral physiological responses to the Trier Social Stress Test and a matched control task. We used FDG-PET to measure neural activity during performance of the two tasks. We used both circulating and airway-specific markers of asthma-related inflammation to assess the impact of acute stress in these two groups.

Results: Asthmatics under chronic stress had a larger HPA-axis response to an acute stressor, which failed to show the suppressive effects on inflammatory markers observed in those with low chronic stress. Moreover, our PET data suggest that greater activity in the anterior insula during acute stress may reflect regulation of the effect of stress on inflammation. In contrast, greater activity in the mid-insula and perigenual anterior cingulate seems to reflect greater reactivity and was associated with greater airway inflammation, a more robust alpha amylase response, and a greater stress-induced increase in proinflammatory cytokine mRNA expression in airway cells.

Conclusions: Acute stress is associated with increases in markers of airway inflammation in asthmatics under chronic stress. This relationship may be mediated by interactions between the insula and anterior cingulate cortex, that determine the salience of environmental cues, as well as descending regulatory influence of inflammatory pathways in the periphery.

Keywords: ACC; Asthma; Cortisol; IL-1; IL-17; Inflammation; Insula; PET; Stress; TSST.

Figure 1. Measures collected and schedule of a typical challenge day

Note: The FDG brain scan was only 30 min in duration. FDG lung scans were also acquired, but not reported on here.

Figure 2. Stress hormone responses to the TSST and control task

Means displayed by group and challenge for (A) cortisol data and (B) alpha amylase data. Error bars represent standard error of the mean.

Figure 3. Change in lung function and asthma-related inflammatory markers in response to stress and control tasks

Means displayed by group and challenge for (A) FEV₁, (B) FeNO, (C) blood EOS, and (D) sputum EOS. Error bars represent standard error of the mean.

Figure 4. Individuals with asthma who have low levels of chronic life stress show increased glucose metabolism in the anterior insula during a social stressor, relative to a control task, compared to asthmatic individuals with high levels of chronic life stress

(A) Cluster showing a group x challenge interaction in the insula ROI analysis. MNI coordinates of the peak voxel in mm (34, 16, −16), 286 voxels. (B) Mean glucose metabolism extracted from the cluster shown in (A) for each condition and group. This plot is provided only to show the data over the significant voxels, and inferences are not provided since the voxel-wise analysis and ROI-averaged analyses utilize the same contrast, which is a case of circular analysis (Kriegeskorte et al., 2009). (C) Overlap of the cluster showing a group x challenge interaction in the right insula in the current data (red) with the cluster showing a group x challenge x valence interaction (blue), and correlation with sputum EOS (green) published in Rosenkranz et al., 2012..(D) Intersection of the three clusters. Image A is thresholded at p < .05, uncorrected.

Figure 5. Individuals with asthma who have low levels of chronic life stress show decreased glucose metabolism in the mid-cingulate cortex (MCC) during a social stressor, relative to a control task, compared to asthmatic individuals with high levels of chronic life stress

(A) Cluster showing a group x challenge interaction in the ACC ROI analysis, thresholded at p < .05, uncorrected. MNI coordinates of the peak voxel in mm (10, −4, 44), 102 voxels. (B) Mean glucose metabolism extracted from the cluster shown in (A) for each condition and group. This plot is provided only to show the data over the significant voxels, and inferences are not provided since the voxel-wise analysis and ROI-averaged analyses utilize the same contrast, which is a case of circular analysis (Kriegeskorte et al., 2009).

Figure 6. Activity in the mid-insula is positively associated with mean FeNO

Glucose metabolism in the mid-insula during performance of the TSST, relative to the control challenge (stress – control), and FeNO averaged across measurements. MNI coordinates of the peak voxel in mm (34, 0, 14), 128 voxels. Image thresholded at p < .05, corrected.

Figure 7. Activity in the cingulate cortex is positively associated stress-related change in mRNA expression of genes in the IL1β/IL-17 pathway

(A) Glucose metabolism in the MCC during performance of the TSST, relative to the control task (stress – control), and the increase in IL23A from baseline to post-stress, relative to post-control. MNI coordinates of peak voxel in mm (−2, −14, 36), 39 voxels (B) Glucose metabolism in the perigenual ACC during performance of the TSST, relative to the control task (stress – control), and the increase in IL1R1 from baseline to post-stress, relative to post-control. MNI coordinates of peak voxel in mm (8, 36, 6), 65 voxels. Images thresholded at (A) p < .05 and (B) p < .06, corrected.

Figure 8. Relationship between cortisol response to stress and percent EOS post-stress differs by chronic stress group

Interaction between Life Stress Interview (LSI) score and log cortisol AUC following stress challenge on (A) % blood EOS at 4h post-stress (B) % sputum EOS at 24h post-stress