(PS5-67) Prospective effects of mobile weight-loss and lifestyle interventions on binge eating

Background: Behavioral weight loss and lifestyle interventions (BWL-LI) have a long history of short-term efficacy and typically result in 5-10% reductions in BMI. Mobile adaptations of these interventions are more scalable, cost-effective, and have introduced virtual coaching and peer group support. Despite these benefits, there is a question about the risk of these interventions worsening or causing eating disorders, particularly binge eating disorders. We prospectively characterized the effect of mobile BWL-LI on binge eating in a secondary analysis of a randomized clinical trial comparing a coach-assisted weight loss program versus digital self-directed.


Methods: We hypothesized that a digital BWL-LI program with peer-groups and coaching would lead to greater reductions in binge eating when compared to a digital self-directed BWL-LI program. We also conducted exploratory analyses to determine what baseline characteristics predicted binge eating at baseline and over time. The primary study design can be found in Sysko et al. (2022). A total of 600 adults with ≥ 27.9 body mass index (BMI) were randomized to a 6-month intervention phase and 24-month maintenance phase comparing active treatment (peer group/coach-assisted mobile app) to control (self-directed mobile app). Frequency of binge eating was assessed by Eating Disorder Examination Questionnaire (EDE-Q) at 7 time points including 1-, 4-, 6-, 12-, 18-, 24-, and 30-month assessments. We used a zero-inflated negative binomial piecewise latent growth curve model (treatment 0-6 months; follow-up 6-30 months) to test our primary hypothesis.

Results: 600 participants were randomized with N = 444 completing follow-up EDE-Q assessments. Binge eating was reported in 20.6% at baseline and improved over the course of the 6-month treatment (β = -1.15, SE = .066, p = 0.08) and returned to baseline levels over the course of follow-up (β = 1.53, SE = 1.01, p = 0.13). There was no significant effect of treatment on either group’s within-treatment change. Additionally, no significant change was found for follow-up or baseline binge-eating X treatment interactions (ps > .05). Several significant predictors of baseline binge eating frequency and change over treatment were found. Those who identified as Hispanic (vs. Not Hispanic) and Never Married (vs. Married) reported more binge episodes at baseline. Less educated individuals reported greater reductions in binge eating within-treatment and those identified as Black/African American, Multiracial, or higher income level reported greater reductions in binge eating over the follow-up period (ps ≤ .01). Those reporting being separated (vs. Married) reported significantly greater increase in binge eating during follow-up (ps ≤ .01).

Discussion: Mobile BWL-LI have the potential to reduce binge eating symptoms in the short-term, but this effect is not durable. The inclusion of peer group support and coaching didn’t alter these effects relative to a self-guided program. Specific binge eating interventions may be necessary to improve outcomes for BWL-LI interventions.