(PS3-C76) Linking Avoidance Motivation to Treatment Success in Pediatric Anxiety: A Computational Perspective

Friday, November 17, 2023

11:30 AM – 12:30 PM PST

Location: Regency Ballroom, Level 7

Author(s)

SR

Safa Rahman, B.S.

Research Trainee
National Institute of Mental Health
Bethesda, Maryland

Co-Author(s)

PN

Phil Newsome, B.S.

Research Trainee
National Institute of Mental Health
Bethesda, Maryland
SR

Sonia Ruiz, B.S.

Graduate Student
Yale University
New Haven, Connecticut
Daniel S. Pine, M.D.

Chief, Section on Development and Affective Neuroscience
National Institute of Mental Health
Bethesda, Maryland
RA

Rany Abend, Ph.D.

Senior lecturer
Reichman University
Herzliya, HaMerkaz, Israel
SQ

Song Qi, Ph.D.

Post-Doctoral Fellow
National Institute of Mental Health
Bethesda, Maryland

Approach-avoidance conflicts (AAC), encompassing competing motivations to avoid potential threat and to approach desired things, are ubiquitous in life. Consistent maladaptive resolution of AAC towards excessive avoidance is a central feature of pediatric anxiety disorders. Quantification of avoidance motivation in experimental settings could potentially inform on pathological mechanisms of pediatric anxiety and response to treatment. However, very few studies have actually examined the predictive utility of AAC decision-making in youth undergoing exposure-based anxiety treatment. Here, we assessed approach-avoidance motivations as correlates and predictors of anxiety symptoms throughout treatment in 24 youth with anxiety disorders (12 females; Mage=12.86 years). Further, we employed a computational approach to aid our prediction efforts; as recent literature has proposed strong clinical implications for such methods (i.e., computational psychiatry). Avoid and approach motivations towards monetary outcomes were quantified using a novel computer-based AAC task. After completing the task, all participants underwent 12-week exposure-based cognitive behavioral therapy (CBT) and anxiety symptoms were assessed via the Screen for Child Anxiety and Related Disorders (SCARED) and Pediatric Anxiety Rating Scale (PARS). A correlation analysis assessed the relationship between treatment outcome and approach-avoidance motivations. The correlation coefficient between SCARED at CBT-week 12 and avoid motivation was r = 0.575 (p = .039), indicating a worse treatment outcome (more severe anxiety symptoms) with greater avoidance motivation. To validate correlation analyses, a LightGBM model which is known to be efficient and scalable with nonlinear features, decently predicted treatment outcome (assessed via SCARED and PARS) from AAC behavioral measures (R2 = 0.37). In addition, we analyzed the importance of behavioral variables in predicting the treatment outcomes via SHAP values. Consistent with the correlation analysis, avoidance motivation is most prominently associated with treatment outcomes, where patients with greater avoidance tendencies have higher SCARED and PARS scores. Importantly, our observations suggest preliminary evidence for predicting anxiety symptoms at various treatment intervals using task-induced motivational behavior. Ultimately, improving our understanding of pathological avoidance behavior within motivational conflicts in anxiety presents a novel path towards identifying biobehavioral phenotypes. As such, varied motivational sensitivities across individuals – with an emphasis on avoidance – can aid in the development of personalized treatment plans for youth with anxiety disorders.