(PS13-D76) The Relationship Between Over-attribution of Threat and Interpersonal Risk Factors of Suicide in Serious Mental Illness

Introduction: People with serious mental illness (SMI; schizophrenia, schizoaffective disorder, bipolar disorder) are at an increased risk of suicidal ideation (SI) and behavior. People with SMI have elevated perceived burdensomeness (PB) and thwarted belongingness (TB), core interpersonal risk factors for suicide. Other factors that influence SI in this population may include over-attribution of threat, a social cognitive bias to misattribute threatening emotion to neutral faces. This study examined how over-attribution of threat, measured by a mobile facial affect measure, relates to the risk factors PB and TB.

Method: Two hundred seventy-three participants with a diagnosis of SMI were assessed for current SI and behavior at baseline, and were stratified by two groups (SI vs. non-SI). Participants were instructed to complete short smartphone surveys (via ecological momentary assessments [EMA]) 3x/day for 10 days. They also completed the Mobile Ecological Test of Emotion Recognition (METER) 1x/day. Linear mixed models tested the relationships between SI, over-attribution of threat, METER performance, and PB/TB.

Results: Participants with current SI had higher levels of EMA-measured PB (B=.88, SE=.05, t=18.73, p< .001) and TB (B=-.64, SE=.05, t=-13.84, p< .001). Participants with and without current SI did not differ in over-attribution of threat (B=.02, SE=.02, t=1.19, p=.236) or overall METER performance (B=.05, SE=.08, t=.56, p=.577). In the whole sample, greater momentary over-attribution of threat was related to greater PB (B=.29, SE=.09, t=3.09, p=.002), but not TB (B=.03, SE=.09, t=0.34, p=.731). The relationship between over-attribution of threat and PB remained when controlling for overall performance on the METER (B=.22, SE=.10, t=2.19, p=.029).

Discussion: These results suggest that greater over-attribution of threat is similar across levels of SI and relates more strongly to PB than TB, and that this relationship remains irrespective of facial affect recognition abilities. The results of this study may inform future suicide prevention interventions for people with SMI, specifically those targeting social cognitive abilities and perceived burdensomeness.