191. Loneliness, social cognition, and psychotic features in bipolar disorder: the mediating role of social support

Abstract

Abstract Background Bipolar disorder (BD) is associated with persistent interpersonal difficulties, social cognition impairment, and elevated loneliness even during euthymic periods. Individuals with BD who present with psychotic features (BD-P) often demonstrate more severe neurocognitive and functional impairment than those without psychotic features (BD-NP). Social support has been identified as a modifiable protective factor that may buffer the negative effects of loneliness on cognitive and psychosocial outcomes. However, the mechanisms linking loneliness, social support, and social cognition, particularly theory of mind (ToM), remain insufficiently understood across BD subgroups and healthy controls (HC). This study examined group differences and explored whether social support mediates the relationship between loneliness and ToM performance. Aims & Objectives To compare loneliness, social support, and social cognition across BD-P, BD-NP, and HC groups. To examine correlations among loneliness, dimensions of social support, and ToM. To explore whether social support mediates the relationship between loneliness and ToM (total, verbal, and non-verbal) in BD-P, BD-NP, and HC groups. Method A total of 239 participants were included: BD-P (n=59), BD-NP (n=94), and HC (n=86). Loneliness was measured using the UCLA Loneliness Scale; social support was assessed using the Measure of Social Function (MSF), capturing perceived and actual instrumental and emotional support; and ToM was evaluated using validated Chinese Theory of Mind tasks. Clinical assessments included HDRS-17 and YMRS. Group comparisons were conducted using ANOVA, χ2 tests, and post-hoc t-tests. Correlations were examined using Pearson coefficients. Mediation analyses were performed using PROCESS macro (Model 4) to determine indirect effects of social support on the loneliness-ToM relationship across groups. Results Significant group differences emerged. BD-NP reported the highest loneliness, followed by BD-P and HC (p<0.001). Social support was lowest in BD-NP and highest in HC across all dimensions (p<0.001). ToM total scores were reduced in BD-P compared with HC (p=0.03), particularly in verbal ToM (p=0.004). Across the whole sample, loneliness was strongly negatively correlated with all types of social support (r = –0.30 to –0.57, p<0.01). In the BD group, higher loneliness was significantly associated with poorer ToM performance (verbal and non-verbal), and with reduced perceived and actual support. Mediation analyses showed: • All participants: Several social support dimensions partially mediated the relationship between loneliness and ToM. • HC: Emotional and total support mediated loneliness–ToM associations. • BD-P: Significant indirect effects were observed for non-verbal ToM via perceived emotional and instrumental support. • BD-NP: Social support (particularly actual emotional support) significantly mediated the relationship between loneliness and both total and non-verbal ToM (indirect β = 0.03–0.12). Discussion & Conclusions Loneliness and social support differ substantially by BD subtype and contribute uniquely to social cognitive functioning. Mediation patterns indicate that social support serves as a key mechanism linking loneliness to ToM, with particularly strong effects in BD-NP. Strengthening both perceived and actual support may reduce loneliness-related social cognitive impairments in BD populations. These findings highlight social support as a clinically relevant target for psychosocial intervention and recovery-oriented care in BD.