(PS8-28) Decision Making in Loss- and Reward-contexts: Impacts on Self-injury

Abstract Text: Purpose
Self-injurious thoughts and behaviors (SITBs) are hypothesized to emerge when individuals struggle to implement less-costly emotion regulation (ER) strategies while facing aversive situations that require a high cognitive load to resolve. During these situations, individuals with SITBs may experience decision-making impairments that limit the ability to search for solutions, learn from previous actions, and make adaptive choices that resolve the negative experience. Potential explanations for these decision-making deficits include: 1) difficulty causally linking actions with consequences, impairing learning in emotion regulation contexts, and 2) reward valuation problems, affecting appropriate outcome valuation. These impairments may lead individuals to choose SITBs as ER strategies, as these behaviors are highly salient, typically rapidly reduce negative affect, and their cost may not be adequately weighted. Self-harm (SH), commonly endorsed as an ER strategy, may thus be selected as a maladaptive but effective strategy to reduce a negative state. Despite theoretical accounts, our understanding of these mechanisms remains limited, likely due to most studies on the topic being cross-sectional while SITBs are highly dynamic. This study addresses this gap by investigating relationships between daily SH urges, negative affect, and decision-making parameters using a drift diffusion model (DDM), providing insight into real-time choice processes underpinning SITBs.

Methods
Participants (n = 66, Mage = 38.2 ± 9.4; 36% non-White, 58% male) recruited online completed a 7-day daily diary, including a behavioral task where they responded to reward- and loss-associated images with monetary outcomes, and daily ratings of SH urges and negative affect. Hierarchical Bayesian DDM was used to estimate boundary separation, learning rate, and the drift rates for reward- and loss contexts. Bayesian mixed effects models examined the relationship between drift diffusion parameters, negative affect, and SH urges at the within- and between-subjects level in reward and loss-contexts.

Results
Within-subjects, daily SH urges were associated with greater loss-associated drift rates (β = 0.02 [.01, 0.03]) and negative affect (β = .05 [0.03, 0.06]), but not reward-associated drift rates, learning rate, or boundary separation. Between-subjects, SH urges were associated with negative affect (β = 0.07 [0.04, 0.09]) and inversely associated with loss-associated drift rates (β = -0.10 [-0.15, -0.05]). We did not find interactions between negative affect and decision-making parameters in either model.

Discussion
These findings suggest that in loss-contexts, individuals with higher SH urges generally engage in a slower decision-making process, however, make faster decisions on days where they experience relatively higher SH urges. This suggests a dynamic relationship between decision-making and SSITBs, where daily fluctuations in loss-associated information processing may contribute to momentary SH risk. Moreover, the difference in within- vs. between-subject findings highlights the importance of dynamically assessing the relationship between decision-making and SITBs.