A meta-analysis of fMRI studies on in-group and out-group categorization

796 

Abstract Submission 

Kelly Sng^1,2, Xavier Lim², SH Annabel Chen^2,3,4,5, Gianluca Esposito⁶

¹Neuroscience, Interdisciplinary Graduate Programme, Nanyang Technological University, Singapore, Singapore, ²Psychology Program, School of Social Sciences, Nanyang Technological University, Singapore, Singapore, ³Centre for Research and Development in Learning (CRADLE), Nanyang Technological University, Singapore, Singapore, ⁴Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore, Singapore, ⁵Office of Educational Research, National Institute of Education, Singapore, Singapore, ⁶Department of Psychology and Cognitive Science, University of Trento, Rovereto, Italy

Kelly Sng  
Neuroscience, Interdisciplinary Graduate Programme, Nanyang Technological University|Psychology Program, School of Social Sciences, Nanyang Technological University
Singapore, Singapore|Singapore, Singapore

Xavier Lim  
Psychology Program, School of Social Sciences, Nanyang Technological University
Singapore, Singapore

SH Annabel Chen, PhD  
Psychology Program, School of Social Sciences, Nanyang Technological University|Centre for Research and Development in Learning (CRADLE), Nanyang Technological University|Lee Kong Chian School of Medicine, Nanyang Technological University|Office of Educational Research, National Institute of Education
Singapore, Singapore|Singapore, Singapore|Singapore, Singapore|Singapore, Singapore

Gianluca Esposito  
Department of Psychology and Cognitive Science, University of Trento
Rovereto, Italy

Different social groups exhibit varying degrees of closeness to the self, with some memberships readily discernible based on appearance (e.g., ethnicity) while others (e.g., nationality) necessitate higher-level cognitive processing. Although different group memberships could elicit distinct neural networks, most neuroimaging meta-analyses examining functional networks on in-out-group biases do not make such distinctions Similarly, differential functional activation may exist for bias toward the in- over the out-group (IG>OG) and out- over the in-group (OG>IG) across different tasks, depending on the modality involved. The activation likelihood estimation (ALE) has 2 aims: (1) summarise the cortical networks for overall IG>OG and OG>IG processing, and (2) provide additional information on ethnicity as group membership, and empathy processing.

A systematic review (PRISMA guidelines) was performed on PubMed and Scopus on fMRI studies (Figure 1). Inclusion criteria: task-fMRI studies, whole-brain analyses in a standard stereotaxic space, involved healthy adults, and published in English. The overall IG>OG processing included 64 studies (106 contrasts, 493 foci), and overall OG>IG processing included 34 studies (59 contrasts, 347 foci). Group membership classification found 37 Ethnicity studies on in-group bias (IG>OG; 58 contrasts, 268 foci). fMRI task types found 29 studies in Empathy processing on in-group bias (IG>OG; 49 contrasts, 249 foci). All other group memberships and fMRI task-types did not have enough studies for adequate power to conduct the ALE analyses. Meta-analyses were performed using GingerALE. An uncorrected threshold of p<.005, 200mm³ was used to minimise false negative findings (Saarinen et al., 2021).

ALE analysis for IG>OG processing revealed 24 significant clusters (Figure 2a), with the largest (3480mm³) located in the left insula, inferior frontal gyrus, and uncus. OG>IG processing revealed 30 significant clusters (Figure 2b), with the largest (2000mm³) located in the bilateral presupplementary motor area. Analysis for Ethnicity (IG>OG) revealed 31 significant clusters (Figure 2c), with the largest (1680mm³) located in the right middle occipital gyrus, middle temporal gyrus, inferior occipital gyrus, and fusiform. Analysis of the task type, Empathy Processing (IG>OG) revealed 23 significant clusters (Figure 2d), with the largest (2648mm³) located in the right postcentral gyrus and posterior cingulate gyrus.

Overall IG>OG processing engaged left insula and inferior frontal gyrus, both of which are activated when the self is involved (Northoff & Bermpohl, 2004), in line with the idea that the in-group is closely related to the self (Smith & Henry, 1996). Our meta-analysis showed more recruitment of self-related areas in in-group bias as expected. The bilateral presupplementary motor area, implicated in cognitive control (Hertrich, 2016) was observed in overall OG>IG processing, suggesting that in-group processing may be more automatic whereas more cognitive resources are required for out-group processing. Occipital involvement for Ethnicity group processing is consistent with a review (Bagnis et al., 2020) reporting recruitment of these regions for racial traits perception. As occipital regions are linked with facial representation and encoding of invariant visual traits like ethnicity, heightened activation seen in IG>OG contrast could imply greater focus on individual distinctions within the in-group. Our observation of the engagement of postcentral gyrus and posterior cingulate in Empathy Processing aligns with prior literature associating these areas with empathy's sensory and affective facets, respectively (Singer et al., 2004; Völlm et al., 2006). This stronger empathic response appears evident for the in-group than the out-group as expected. The findings add to our understanding of the neuro-networks involved in in-group/out-group processes.

Social Cognition ¹

Social Interaction ²

Social Neuroscience Other

Activation (eg. BOLD task-fMRI)

Cognition

FUNCTIONAL MRI

Meta- Analysis

Social Interactions

Other - In-Group & Out-Group; Ethnicity & Race; Empathy