Among the predisposing factors to developing bipolar disorder (BD), the presence of hypomanic personality traits in young adults has been seen as a reliable predictor of mania (Kwapil et al., 2000; Walsh et al., 2015) and is assessed through the Hypomanic Personality Scale (HPS; Eckblad and Chapman 1986), validated among college students (Klein et al., 1996), as well as family samples including individuals with BD (Johnson et al., 2015; Meyer, 2002; Savitz et al., 2008). The hypomanic personality style defines people who are in a mild manic state most of the time, but do not fulfill the DSM criteria for episodic hypomania (American Psychiatric Association, 2013; Eckblad and Chapman, 1986). Some of these traits are adaptive, as people are energetic, upbeat, gregarious, and hard-working, although they can also be irritable and risk-taking. Moreover, as soon as it was defined, some authors wondered whether this personality style might reflect an actual affective condition (Eckblad and Chapman, 1986). Hypomanic traits can indeed be understood as subsyndromal symptoms, and included in the bipolar spectrum from a dimensional perspective (Parker et al., 2014).

One interesting facet of this personality style, besides its direct link to mood symptoms, is its social component, sometimes called social vitality or hypersociability (Rawlings et al., 2000; Schalet et al., 2011). This dimension reflects both positive and maladaptive aspects, and has been related to cognitive aspects such as semantic ambiguity processing (Raucher-Chene et al., 2017), as well as psychopathological traits such as impulsivity (Schalet et al., 2011). This social potency and vivaciousness can be set against the theory of mind (ToM) impairments observed in people with hypomanic traits (Terrien et al., 2014). More specifically, ToM skills, namely the ability to represent one's own and other people's mental states (emotions, intentions, beliefs), can be predicted by the presence of hypomanic traits, especially in men (Terrien et al., 2014). Emotional context integration also seems to be modified by the presence of these traits. In Terrien et al. (2015a)’s study, participants with hypomanic traits had difficulty integrating negative emotional contexts, but exhibited enhanced integration of positive emotional contexts. The above studies highlighted the value of assessing this population with more ecological tasks, to capture the social processing anomalies encountered in everyday life. More naturalistic social stimuli could be used to detect subtle, more finely-grained impairments, especially when combined with neuroimaging (Redcay and Moraczewski, 2020; Zhang et al., 2021).

A number of tasks have been used to explore the neural correlates of ToM, defining a core network associating the medial prefrontal cortex, bilateral temporoparietal junction, and precunei, but each type of ToM task also elicits activity in specific brain regions (Molenberghs et al., 2016; Schurz et al., 2014). As these tasks tend to be static, they cannot capture the richness of social interactions, thereby limiting the application of their results in real life (Redcay and Moraczewski, 2020). For this reason, new tasks have been developed, featuring videos to make them more ecological (Henry et al., 2021; Serra-Mayoral et al., 2021). Ecological validity therefore refers to the potential for generalizability (i.e. the extent to which assessment results relate to and/or predict behaviors outside the test environment) and representativeness (i.e. the extent to which assessments resemble the everyday contexts in which behaviors will be required) (Dawson and Marcotte, 2017). For example, the Dynamic Inference Task (DIT; Henry et al., 2021) explores the neural correlates of different aspects of ToM (i.e., degree of inference, presence of emotion, gaze direction) identified in everyday-life interactions. In the general population, this task elicits neural activation in temporoparietal and prefrontal regions, internal temporal areas around the amygdala, and the right dorsomedial part of the superior frontal gyrus (Henry et al., 2021).

Besides task-related brain activation analyses, one complementary approach consists in assessing functional connectivity between brain regions (Wang et al., 2021). Task-dependent functional connectivity can shed light on the interactions between regions that subtend ToM (McCormick et al., 2018; Schurz et al., 2020; Zillekens et al., 2019), and could help to delineate the impact of personality traits on ToM processes. A study conducted among individuals with BD and their unaffected relatives revealed reduced frontotemporoparietal connectivity in both groups, whereas increased connectivity between the right middle temporal gyrus and medial prefrontal cortex was only observed in the unaffected relatives (Willert et al., 2015), pointing to the presence of a compensatory process. As difficulties in ToM are already observed in individuals with hypomanic traits, it is possible that the neural network is also modified. To the best of our knowledge, no study has yet conducted a task-dependent functional connectivity analysis to explore the association between the neural correlates of ToM and hypomanic personality traits.

The present study therefore investigated the potential impact of hypomanic traits on brain activation and task-based connectivity strength during a video ToM task in a nonclinical population. For this purpose, we used the Dynamic Inference Task, which requires participants to attribute a mental state to an inferred social situation with different social cues (i.e., emotional expression, gaze direction). As such, it seems an appropriate means of exploring the neural correlates of social interactions (Henry et al., 2021). We hypothesize that activation of temporoparietal regions would be altered, as well as connectivity in a frontotemporoparietal network as described by Willert et al. (2015) in a clinical population.