Work

Real-time affective predictors of suicidal thinking: A dynamic structural equation modeling study

Year: 2025

Type: article

Abstract: BACKGROUND: Ecological momentary assessment (EMA) has advanced the study of suicidal thinking by capturing moment-to-moment variability. However, prior work is limited by overly conservative analytic approaches that focus too much on prediction of residual variance, a lack of differentiation between passive and active suicidal thinking, and an underrepresentation of high arousal negative emotions. METHODS: In the present study, we address these limitations using Dynamic Structural Equation Modeling (DSEM) to examine proximal affective predictors of suicidal thinking in a sample of 106 adults recently discharged from inpatient care. We assessed the within-person and between-person relation among passive and active suicidal thinking and three affective states: sadness (low arousal negative emotion), anxiety (high arousal negative emotion), and overall emotional distress. RESULTS: Results revealed strong contemporaneous associations between all affective states and both passive and active suicidal thinking. However, only sadness significantly predicted passive suicidal thinking at the next time point, and no affective variables prospectively predicted active suicidal thinking. Additionally, sadness was more strongly associated contemporaneously with suicidal thinking than anxiety. CONCLUSIONS: These findings suggest that although affective states are closely tied to suicidal thinking in the moment, their predictive utility for even near future suicidal thinking is limited. This highlights the challenges of short-term risk prediction and underscores the need for future research to explore dynamic predictors that differentiate passive and active suicidal thoughts over time.

Source: Journal of Affective Disorders

DOI: 10.1016/j.jad.2025.120329