Temporal associations between experiential avoidance and disordered eating behaviors in adolescents and young adults: findings from an epidemiological cohort study with ecological momentary assessment

Abstract

Abstract Purpose Previous studies linking experiential avoidance (EA) and eating pathology have largely relied on single measurements based on traditional retrospective questionnaires. Taking advantage of available repeated assessments of EA and disordered eating behaviors (DEBs) in the everyday lives of young people from an epidemiological cohort, we aimed to investigate ecologically valid temporal associations between these constructs. Methods A random population sample of N = 1180 14–21-year-olds from Dresden, Germany, participated at baseline (2015/2016). As part of smartphone-based ecological momentary assessment (EMA), participants reported on engagement in EA and four DEBs (skipping eating, eating large amounts of food, loss-of-control eating, and restrained eating) up to eight times per day for four days. Multilevel modeling of concurrent and time-lagged associations between EA and DEBs, was conducted among those with at least 50% EMA-compliance (n = 1069). Results EA was associated with higher concurrent levels of all four types of concurrent DEBs. In addition, EA significantly predicted subsequent levels of restrained eating. Only loss-of-control eating significantly predicted subsequent EA, and this effect depended on the timespan between consecutive assessments. When this timespan was short, higher Loss-of-control eating predicted lower subsequent EA, while it predicted higher subsequent EA when the timespan was longer. Conclusion The present findings suggest that EA is temporally closely linked to greater engagement in DEBs, supporting theoretical assumptions that DEBs may serve an attempted avoidance function in the context of unpleasant inner experiences. Future studies may benefit from examining samples with more pronounced eating pathology. Level of evidence Level IV: Evidence obtained from multiple time series with or without the intervention, such as case studies.