Real-Time Monitoring of Relapse in Eating Disorders - Project Summary/Abstract Eating disorders (EDs) are severe mental illnesses with high mortality,1 few effective treatments,2–4 and high relapse rates.5 Current evidence-based treatments for EDs often result in only 50% of individuals reaching full remission of their symptoms.4 Further, even among those who do recover, many relapse. Up to 50% of those who recover experience a relapse after treatment, about 25% of individuals go on to experience a severe and persistent course of illness.5–13 Despite these high rates of relapse, the past decades of research have culminated in very little improvement to prevent relapse within the EDs. To date, we still know very little about why relapse occurs; in part, because ED relapse has been studied primarily as an outcome (i.e., relapsed or not) rather than a dynamic process can be intervened upon.9,14 Reconceptualizing and studying ED relapse as a dynamic process will aid the identification of mechanisms of relapse and inform the development and refinement of interventions and relapse prevention efforts. A well-established model of relapse in substance use disorders (i.e., the cognitive behavioral relapse model; CBRM)15 can inform the initial formation of a similar model in EDs. With research design informed by the CBRM, the proposed study will use micro-longitudinal (i.e., ecological momentary assessment) and longitudinal (i.e., 3-month follow-up) methods to capture lapses and relapse among those with EDs. The use of theory-driven, fine-grained assessment will move the field toward the elucidation of distal and proximal mechanisms of relapse as it unfolds in real-time. First, we will capture the presence and dynamics of lapses (i.e., momentary engagement in ED behaviors), and whether features of lapses (e.g., frequency, time between lapses) predict eventual relapse (Aim 1). The CBRM specifies that both tonic processes (i.e., stable, chronic factors) and phasic responses (i.e., immediate precipitants) increase the risk for relapse. As such, we will test the associations between tonic processes and the propensity for lapsing (i.e., higher urge valence and stability and lower urge variability; Aim 2). Additionally, we will use generalized estimating equation models to examine whether each proposed phasic response (e.g., affect, emotion regulation skill use) predicts a lapse at the next survey (Aim 3). Lastly, the CBRM posits that “high-risk” situations increase the likelihood of an individual lapsing. This study will aid in the initial identification of said high-risk situations in EDs by using mixed-method (i.e., qualitative and quantitative) analyses to explore contextual factors contributing to individuals’ reported lapses (Exploratory Aim 4). Overall, the proposed project will implement advanced statistics and micro- and macro-longitudinal design to provide an initial mechanistic understanding of relapse as it occurs in the EDs. Findings from this study will inform the refinement of relapse prevention interventions and just-in-time adaptive interventions, and ultimately, promote lasting, long-term recovery for those with this impairing and persistent disorder.
