Empirically validated digital interventions for recurrent binge eating typically target numerous hypothesized change mechanisms via the delivery of different modules, skills, and techniques. Emerging evidence suggests that interventions designed to target and isolate one key change mechanism may also produce meaningful change in core symptoms. Although both ‘broad’ and ‘focused’ digital programs have demonstrated efficacy, no study has performed a direct, head-to-head comparison of the two approaches. We addressed this through a randomized non-inferiority trial.

Participants with recurrent binge eating were randomly assigned to a broad (n = 199) or focused digital intervention (n = 199), or a waitlist (n = 202). The broad program targeted dietary restraint, mood intolerance, and body image disturbances, while the focused program exclusively targeted dietary restraint. Primary outcomes were eating disorder psychopathology and binge eating frequency.

In intention-to-treat analyses, both intervention groups reported greater improvements in primary and secondary outcomes than the waitlist, which were sustained at an 8-week follow-up. The focused intervention was not inferior to the broad intervention on all but one outcome, but was associated with higher rates of attrition and non-compliance.

Focused digital interventions that are designed to target one key change mechanism may produce comparable symptom improvements to broader digital interventions, but appear to be associated with lower engagement.

Although effective treatments exist for diagnostic and subthreshold-level eating disorders (EDs), a significant proportion of affected individuals do not receive help. Interventions translated for delivery through smartphone apps may be one solution towards reducing this treatment gap. However, evidence for the efficacy of smartphones apps for EDs is lacking. We developed a smartphone app based on the principles and techniques of transdiagnostic cognitive-behavioral therapy for EDs and evaluated it through a pre-registered randomized controlled trial.

Symptomatic individuals (those who reported the presence of binge eating) were randomly assigned to the app (n = 197) or waiting list (n = 195). Of the total sample, 42 and 31% exhibited diagnostic-level bulimia nervosa and binge-eating disorder symptoms, respectively. Assessments took place at baseline, 4 weeks, and 8 weeks post-randomization. Analyses were intention-to-treat. The primary outcome was global levels of ED psychopathology. Secondary outcomes were other ED symptoms, impairment, and distress.

Intervention participants reported greater reductions in global ED psychopathology than the control group at post-test (d = −0.80). Significant effects were also observed for secondary outcomes (d's = −0.30 to −0.74), except compensatory behavior frequency. Symptom levels remained stable at follow-up. Participants were largely satisfied with the app, although the overall post-test attrition rate was 35%.

Findings highlight the potential for this app to serve as a cost-effective and easily accessible intervention for those who cannot receive standard treatment. The capacity for apps to be flexibly integrated within current models of mental health care delivery may prove vital for addressing the unmet needs of people with EDs.

We built an app to help clients of food pantries. The app offers vegetable-based recipes, food tips and no-cost strategies for making mealtimes healthier and for bargain-conscious grocery shopping, among other themes. Users customize materials to meet their own preferences. The app, available in English and Spanish, has been tested in a randomized field trial.

A randomized controlled trial with repeated measures across 10 weeks.

Clients of fifteen community food pantry distributions in Los Angeles County, USA.

Distributions were randomized to control and experimental conditions, and 289 household cooks and one of their 9–14-year-old children were enrolled as participants. Experimental dyads were given a smartphone with our app and a phone use-plan, then trained to use the app. ‘Test vegetables’ were added to the foods that both control and experimental participants received at their pantries.

After 3–4 weeks of additional ‘test vegetables’, cooks at experimental pantries had made 38 % more preparations with these items than control cooks (P = 0·03). Ten weeks following baseline, experimental pantries also scored greater gains in using a wider assortment of vegetables than control pantries (P = 0·003). Use of the app increased between mid-experiment and final measurement (P = 0·0001).

The app appears to encourage household cooks to try new preparation methods and widen their incorporation of vegetables into family diets. Further research is needed to identify specific app features that contributed most to outcomes and to test ways in which to disseminate the app widely.

Adolescents’ snacking habits are driven by both explicit reflective and implicit hedonic processes. Hedonic pathways and differences in sensitivity to food rewards in addition to reflective determinants should be considered. The present study evaluated the feasibility and impact of a mobile phone-delivered intervention, incorporating explicit reflective and implicit rewarding strategies, on adolescents’ snack intake.

Adolescents (n 988; mean age 14·9 (sd 0·70) years, 59·4 % boys) completed a non-randomized clustered controlled trial. Adolescents (n 416) in the intervention schools (n 3) were provided with the intervention application for four weeks, while adolescents (n 572) in the control schools (n 3) followed the regular curriculum. Outcomes were differences in healthy snacking ratio and key determinants (awareness, intention, attitude, self-efficacy, habits and knowledge). Process evaluation data were collected via questionnaires and through log data of the app.

No significant positive intervention effects on the healthy snack ratio (b=−3·52 (se 1·82), P>0·05) or targeted determinants were observed. Only 268 adolescents started using the app, of whom only fifty-five (20·5 %) still logged in after four weeks. Within the group of users, higher exposure to the app was not significantly associated with positive intervention effects. App satisfaction ratings were low in both high and low user groups. Moderation analyses revealed small positive intervention effects on the healthy snack ratio in high compared with low reward-sensitive boys (b=1·38 (se 0·59), P<0·05).

The intervention was not able to improve adolescents’ snack choices, due to low reach and exposure. Future interventions should consider multicomponent interventions, teacher engagement, exhaustive participatory app content development and tailoring.