2 - (SYM 37) Associations Between Affect and Suicidality After Psychiatric Inpatient Discharge: An Application of Dynamic Structural Equation Modeling

Rates of suicide are increasing, and there is a critical need to identify clear warning signs for suicide that can be amenable to timely interventions. Changes in affective states may be a salient imminent risk factor for suicide. Previous research suggests higher mean-level negative affect and lower mean-level positive affect predict subsequent suicidality. Recently, increased research has focused on affect dynamics that capture more nuanced emotional experiences, including emotional inertia (i.e., how persistent emotional states remain over time) and emotional variability (i.e., how much emotional states fluctuate over time). However, momentary changes in affect and affect dynamics have rarely been examined simultaneously in suicide research. It is unclear if some affective markers have better clinical utility in predicting suicidality, particularly among high-risk clinical samples. In the present study, we address this gap in the literature by evaluating how both within-person positive and negative affect intensity, as well as emotional inertia and variability, influence risk for suicidal ideation (SI). Our sample includes 158 adult psychiatric inpatients who completed a daily ecological momentary assessment protocol for two months after discharge, a period of notably elevated risk for suicide (M_age = 35.5 ± 14.5 years, 63.9% female, 65.2% heterosexual, 82.2% white). We modeled associations between affect and SI using Dynamic Structural Equation Modeling (DSEM), a cutting-edge analysis approach that integrates the advantages of structural equation, multi-level, and time-series modeling. At the within-person level, momentary higher negative affect and lower positive affect were associated with higher SI intensity. At the between-person level, the data did not support significant associations between emotional inertia, emotional variability, and SI intensity. Thus, results suggest momentary affect intensity may be a stronger correlate of suicidality than individual differences in broader affect dynamics (e.g., persistent negative affect). Results support the use of intensive longitudinal methodologies to help identify personalized intervention targets that could improve clinical care during periods of increased risk for suicide. This presentation will provide an overview of making analysis decisions within a DSEM framework. Considerations for conducting intensive longitudinal research with high-risk clinical samples will also be discussed (e.g., sparseness in data due to low rates of EMA compliance).