Attraction through similarity in autistic traits: A group communication study using social relations model and fNIRS hyperscanning.

Background: The double empathy problem (DEP) reconceptualizes autism's social challenges as bidirectional differences rather than unidirectional deficits. Following the DEP, the dialectical misattunement hypothesis (DMH) predicts that interaction between people with similar autistic traits will be smoother and reflected in neural synchronization. However, evidence remains inconsistent due to methodological limitations in dyadic designs and unstructured tasks, and it remains unclear whether neural mechanisms differ between passive and active social contexts across autistic trait levels.

Methods: Using the social relations model, we measured the relational attraction within four-person groups (20 female and 10 male groups), composed of two high-autistic-trait individuals and two low-autistic-trait individuals following a turn-taking discussion. Simultaneously, we recorded brain activity using functional near-infrared spectroscopy (fNIRS) during both passive story listening and active turn-taking discussion.

Results: Individuals with similar autistic traits reported higher interpersonal attraction when sharing consistent opinions. Neural analyses revealed context-dependent interbrain coupling patterns: During passive story listening, low-autistic-trait dyads exhibited higher inter-subject correlation (ISC) compared to high-autistic-trait dyads. In contrast, during active communication, low-autistic-trait dyads exhibited higher interbrain synchronization (IBS) in the right temporoparietal junction, while high-autistic-trait dyads showed higher IBS in the right dorsolateral prefrontal cortex, suggesting distinct neural mechanisms underlying social interaction across autistic trait levels.

Conclusions: Our findings support the DMH and reveal that neural synchronization mechanisms vary across both autistic trait levels and social contexts. These context-dependent patterns challenge deficit-based models of autism, suggesting that high-autistic-trait individuals may employ alternative but effective neural strategies during social interaction, particularly in active communication contexts.