The neural architecture of social threat: A meta-analytic comparison of indirect and direct social threat

Indirect and Direct Social Threat (IST and DST) are related but partially distinct forms of interpersonal adversity defined along two dimensions: threat perception, the explicitness and salience of harm, and behavioural reciprocity, the extent of bidirectional contingency and response opportunity. IST involves ambiguous or socially devaluing contexts with limited reciprocity, whereas DST entails overt hostility, clear antagonism, and reciprocal engagement. This distinction shows conceptual parallels with findings from rodent models of social interaction. We tested this framework using a coordinate-based Activation Likelihood Estimation (ALE) meta-analysis of human neuroimaging studies comprising 28 IST contrasts (N = 1081) and 28 DST contrasts (N = 855), with cluster-level family-wise error correction (p<.05). Findings were interpreted alongside passive and agonistic animal paradigms. Conjunction analysis identified shared spatial convergence in the left anterior insula (AI) extending into the inferior frontal gyrus (IFG), consistent with salience-detection and regulatory processes. ALE contrast analysis showed stronger convergence for DST than IST in the right AI, bilateral claustrum, and anterior and mid-cingulate cortices, regions implicated in motivational salience and action preparation; no clusters emerged for IST relative to DST. Predicted IST-specific convergence in the temporoparietal junction and posterior cingulate cortex, and DST-related convergence in the amygdala and ventral striatum, were not observed. This pattern is conceptually consistent with animal evidence associating reciprocal threat with defensive readiness and action selection. Overall, findings partially support the framework: threat perception and behavioural reciprocity may help organise neural responses to social threat, while correspondence with passive-agonistic animal models remains provisional.