NIDER - Neurobiology of Individual Differences in Emotion Regulation
A collaborative image-based meta-analysis
Cognitive reappraisal of stressors is a fundamental strategy people use to regulate their emotions, supporting adaptive goal-oriented behavior and psychological well-being. Meta-Analyses have integrated a large number of fMRI studies to find robust brain networks which underlie such reappraisal processes. Still, there is limited evidence whether differential activity in these networks is related to individual differences in the ability or tendency to employ cognitive reappraisal. This move from within- to between-person associations is an important intermediate step to build robust and interpretable neurobiological models for emotion regulation as a transdiagnostic dimension of psychopathology.
The purpose of this meta-analysis is to quantify the association between regulatory brain networks and individual differences in the capability (and tendency) to employ cognitive reappraisal. In many previous studies, questionnaire measures of cognitive reappraisal have been included, but only a limited number of studies explicitly reported analyses on associations between questionnaires and brain activity during reappraisal. Therefore, this meta-analysis will be both collaborative and image-based, meaning researchers with applicable data are invited to become co-authors on the resulting manuscripts for providing group-level t-maps for the contrasts of interest.
A preregistration of the study protocol can be found here.
If you are interested in contributing or have any questions, send a mail to email@example.com or contact any of the organizers:
About this project
Central Institute of Mental Health Mannheim, Germany
Central Institute of Mental Health Mannheim, Germany
Institute of Psychology, University of Innsbruck, Austria
Who can join?
To collaborate on this project, authors will need to provide group-level t-maps of studies which fulfill the following criteria:
Task-based fMRI study
Cognitive reappraisal task during fMRI, comparing the response to negative stimuli (e.g., IAPS images) in a down-regulation condition with a control condition (i.e., maintaining/permitting emotional responses)
Included a questionnaire that measures reappraisal capabilities or tendencies (e.g. ERQ, CERQ)
Healthy adult sample
We will give you access to a template where you can fill in the following information:
Your name/mail address
Name and year of main data publication
Mean age and standard deviation
Gender proportion of the sample
Questionnaire mean and standard deviation
If applicable: Self-rating type (online, post-hoc)
If applicable: Rating means and standard deviations for regulate and view conditions
If applicable: Correlations between questionnaires, self-ratings (view-regulate), and amygdala activity (view-regulate)
How to prepare and submit study information?
How to prepare and submit brain data?
The meta-analysis will summarize associations between brain activity in the contrast [regulate – view] and three different markers of emotion regulation success:
and (if applicable) self-rating-based
Hence, 2-3 unthresholded whole-brain group-level contrast maps need to be submitted as described below.
For all analyses, first prepare the contrast [regulate – view] for each participant. In general, univariate regressions of this contrast on a covariate (which reflects reappraisal tendencies/success) need to be conducted.
Make sure you test for positive correlations between contrast and covariates reflecting reappraisal tendencies/success (e.g., in SPM assigning a contrast weight is necessary. This contrast weight must positive).
Make sure to employ no thresholding on the t-maps. One option to check this is plotting the t-map as a histogram to see
(1) the values on the x-axis are plausible for t-values and
(2) low t-values are not removed.
Any preprocessing pipeline, data screening strategy, software, or toolbox is permitted to ensure the generalizability of meta-analytic results.
We provide a brief walkthrough how these analyses could be conducted in matlab here, but users of other software might find this helpful as well.
Contact us if you have questions or comments regarding the procedures described here.
Analysis 1: Questionnaire based
Set up a 2nd(group)-level model on the contrast [regulate – view] with the participant-wise scores on your reappraisal scale as a 2nd-level covariate. Please note: Questionnaires scores should be coded so that higher scores reflect increased tendency or capability to employ reappraisal strategies.
Name the resulting t-map according to the pattern “FirstauthorYear_quest_abbreviatedName.nii”, e.g. “Mueller2019_quest_ERQ.nii”. In this example, "ERQ" is the abbreviated name of the "emotion regulation questionnaire". Authorship and year refer to the main publication of the dataset. For unpublished data, use the form: “CoauthorNameUnpublished_quest_abbreviatedName.nii"
Analysis 2: Self-rating-based
Set up a 2nd(group)-level model on the contrast [regulate – view] with the participant-wise difference in affective self-ratings between conditions [view – regulate] as a covariate. Importantly, note that affective self-ratings should be coded so that higher values mean a more negative response. Then, higher values in the difference score for affective self-ratings [view – regulate] correspond to higher reappraisal success. Name the resulting t-map according to the pattern “FirstauthorYear_rating.nii”.
Analysis 3: Amygdala-based
Download a binary map for the amygdala region of interest (ROI) here. Calculate the difference in average activity for this ROI between the view and the regulate condition (i.e. [view – regulate]), so that higher values correspond to a “effective down-regulation” of the amygdala in the regulate condition. We use ROI-wise averaged regression estimates as an indicator of activity, but other strategies might be valid as well.
Then, set up a 2nd(group)-level model on the contrast [regulate – view] with the participant-wise difference in average amygdala activity between conditions [view – regulate] as a covariate. Name the resulting t-map according to the pattern “FirstauthorYear_amy.nii”.
Open Science and Reproducibility
We intend to make the unthresholded group-level t-maps openly available to increase the reproducibility of meta-analytic results. With group-level maps, the privacy of participants is ensured. If you have any questions concerning this, please feel free to contact us.