(PS8-36) Dynamic relations among self-injurious urges, affect, and self-injurious behaviors

Abstract Text: Self-injurious urges (SIUs; i.e., urges to engage in self-injurious behaviors (SIBs)) and affect (i.e., discrete positive and negative emotional states) have each been strongly linked to SIB engagement. Although theoretical models suggest that SIUs may mediate associations between momentary affect and engagement in SIBs, no prior empirical work has tested this prediction. Understanding the momentary relations between SIUs, affect, and SIBs could inform ecological momentary interventions that leverage real-time data to predict and intervene on SIBs before they occur. The current study used ecological momentary assessment (EMA) to investigate the temporal relations between SIUs, affect, and SIBs. We hypothesized that: (1) greater SIU intensity at one survey would be associated with increased likelihood of subsequent SIBs; (2) SIU intensity would mediate the association between affect and later SIBs, with higher negative affect indirectly increasing the odds of SIB engagement through elevated SIU intensity, while higher positive affect would indirectly decrease the odds of SIBs through the same pathway; and (3) SIBs would exhibit an autoregressive effect, where SIBs at one timepoint would predict SIBs at later timepoints. Adults (N = 124) who engaged in at least three SIBs (i.e., non-suicidal self-injury and/or disordered eating behaviors) in the past month completed 5 EMA surveys each day for 14 days. At each survey, participants reported on current affect, SIU intensity, and SIB presence. Multilevel dynamic structural equation models with fixed effects for SIU intensity and affect and random intercepts per person were conducted to examine whether SIU intensity (t+1) mediated the relations between latent negative affect and latent positive affect (t+0) and subsequent SIBs (t+2). There was an autoregressive effect of SIBs, with prior engagement in SIBs predicting SIBs at the next survey (OR = 5.701, p < .001). Additionally, SIU intensity at one survey predicted engagement in SIBs at the next timepoint (OR = 1.010, p < .001). However, there were no direct or indirect effects of positive or negative affect on subsequent SIBs. Findings indicate that while SIU intensity significantly predicted future SIBs, affect did not directly or indirectly influence SIBs through SIUs. This challenges theoretical models like the affect regulation model, suggesting that SIUs may be a more acute risk factor for SIBs than affect, though this warrants further research. Future research should 1) aim to replicate these findings to determine whether this holds in other samples and whether different measurements affect demonstrate different results, 2) investigate affect on a longer time scale (e.g., on a day-level), and 3) examine whether other factors—such as interpersonal stressors or cognitive processes (e.g., interoceptive awareness)—better predict SIBs through their influence on SIUs. Given that SIUs were the strongest proximal predictor of SIBs, interventions may consider directly addressing and prioritizing treating SIUs rather than focusing solely on affect. For example, approaches like dialectical behavior therapy, which provide concrete strategies for managing urges, may be particularly important in reducing SIB risk in real time.