Prefrontal hypoactivation induced via social stress is more strongly associated with state rumination than depressive symptomatology

Previous studies have consistently shown a pattern of prefrontal hypoactivation in depressed patients (DP); however, it remains unclear whether this neural correlate is a consequence or concomitant feature of depression and/or whether ruminative thinking might be underlying. Using a sample comprising 65 healthy controls (HC) and 77 DP, we investigated the behavioral and neural correlates in response to stress and their association with depressive symptomatology, trait and state rumination. Fitting repeated-measurement MANOVAs including 21 fNIRS-channels covering the bilateral Inferior Frontal Gyrus (IFG), Dorsolateral Prefrontal Cortex (DLPFC) and Somatosensory Association Cortex (SAC), we investigated the predictive value of diagnostic group (HC vs. DP) and state rumination. In DP, we observed significantly lower increases in cortical oxygenation under stress in channels of the right IFG and bilateral DLPFC. Participants reporting lower state rumination and no increases in state rumination under stress showed higher increases in cortical oxygenation compared to the other groups and in more channels compared to the analysis on diagnostic group. Re-running our fNIRS-analysis while correcting for performance resulted in time-dependent changes dependent on group (DP vs. HC) no longer yielding significance, however for the differentiation of state rumination groups.


Supplementary material
Supplementary material S1.Items of the state rumination questionnaire including adapted items from the Ruminative Response Scale (Nolen-Hoeksema & Morrow, 1991), Amsterdam Resting-State Questionnaire (Diaz et al., 2013) and the Perseverative Thinking Questionnaire (Ehring et al., 2011).Subjects were instructed to rate if the items were in line with their mental state during the last resting-state measurement.

German translation English translation
Ich dachte immer wieder an meine Probleme.
I repeatedly thought about my problems.
I kept thinking about things that bother me.
I dwelled on my thoughts without coming to a solution.
Ich verlor mich in meinen negativen Gedanken.I got lost in my negative thoughts.
I had difficulties holding on to my thoughts.
I could not let go of my negative thoughts.
Ich war bei der Sache.I was present.
I thought about why I acted wrong in certain situations.
I thought why I have problems other people don´t have.
I thought about whereby I deserved my current life situation.
I thought why I can´t handle things better.
I thought about all my shortcomings, failings, faults, mistakes.
I could switch between my thoughts flexibly.
I thought about past situations that I regret.
Ich verlor mich in Gedanken an Vergangenes.I got lost in thoughts about the past.
I was consumed by my problems and worries.
I couldn´t let go of my negative thoughts.
Supplementary material S2.Results of Benjamini-Hochberg-corrected pairwise comparisons main effect of time of the NIRS-analysis.

NIRS dependent on group (DP vs. HC):
Benjamini-Hochberg-corrected post-hoc pairwise comparisons of the main effect of time (CTL1 vs. TSST and CTL2 vs. TSST for significant channels) indicated significant increases from CTL1 to the TSST in the case of all three channels of the left IFG, left DLPFC, right DLPFC, all nine channels of the SAC as well as one of three channels of the right IFG (channel 18) respectively.Significant increases from CTL2 to TSST were observed in the case of all three channels of the left DLPFC, two out three channels of the left IFG (channel 6, 9), two out of three channels of the right DLPFC (channel 20.24) and two out of nine channels of the SAC (channel 27, 31).

NIRS dependent on SR-cluster:
We investigated the Benjamini-Hochberg-corrected pairwise comparisons of the main effect of time and observed significant increases from CTL1 to the TSST in the case of two channels of the left IFG (channel 7, 9), one channel of the right DLPFC (channel 20), and two channels covering the SAC (channel 25, 32).We further observed significant increases from CTL2 to the TSST in the case of a subset of the aforementioned channels, namely channel 9 and 20.
Performance-corrected NIRS dependent on group (DP vs. HC): Pairwise comparisons of the time effect in the case of these channels yielded significant differences between CTL1 and TSST in the case of all channels of the left and right DLPFC, all channels of the SAC as well as two out of three channels of the left IFG (channel 6, 9).We further observed significant differences between CTL2 and TSST in the case of all channels of the left DLPFC, two out of three channels of the right DLPFC (channel 20.24), a subset of two channels of the SAC (channel 27, 31) as well as the same channels of the left IFG (channel 6, 9).
Performance-corrected NIRS dependent on SR-cluster: Pairwise comparisons of the time effect in these channels yielded significant increases between CTL1 and TSST in the case of all channels of the right DLPFC, one channel of the left DLPFC (channel 10), seven out of nine channels of the SAC (channel 25, 26, 27, 28, 30. 32. 35) and one channel of the left IFG (channel 6).We further observed significant increases between CTL2 and TSST in the case of three channels of the SAC (25,28,32).
Supplementary material S4.T-tests investigating the differences in NIRS cortical oxygenation dependent on group (DP = depressed patients vs. HC = healthy controls).Please note that this is for illustrative purposes only and therefore not corrected for multiple comparisons.Note: p-value of one-sided t-tests material S3.Illustration of the results of the fNIRS-MANOVAs.Please note that only channels with significant group*time interactions are depicted.Error bars indicate +/-1 SE.