An intensive weight loss programme with behavioural support for people with type 2 diabetes at risk of eating disorders in England (ARIADNE): a randomised, controlled, non-inferiority trial

Background: There are concerns that low-energy total diet replacement (TDR) programmes could trigger eating disorders, given their focus on weight and rigid dietary rules. We aimed to assess the effect of a TDR programme on eating disorder symptoms in people living with overweight or obesity and type 2 diabetes at high risk of developing an eating disorder. Methods: In this randomised, controlled, non-inferiority trial, participants with type 2 diabetes, overweight, and eating disorder symptoms across England were randomly assigned (1:1) to a low-energy TDR programme with formula products and behavioural support delivered remotely, or usual care. In brief, the intervention comprised 12 weeks of low-energy TDR in a nutritionally complete package of soups, shakes, and bars. After the 12 weeks, the intervention continued with stepped food reintroduction (around 8 weeks) based on a low-energy, nutrient-rich diet, followed by weight maintenance advice (around 4 weeks), personalised to an individual participant's circumstances and preferences. Participants allocated to the control group received usual care for their diabetes. The primary outcome was the change in eating disorder symptoms using the Eating Disorders Examination Questionnaire (EDE-Q) global score at 6 months (programme end). Safety was determined by the incidence of cases with high suspicion of a new eating disorder. The primary outcome analysis had an upper non-inferiority margin for EDE-Q of +1 SD (0·72). People with lived experience were involved throughout the trial and provided input on study conceptualisation, protocol development, delivery of the intervention, and intervention materials. The study was registered with ClinicalTrials.gov, NCT05744232. Findings: Between March 8, 2023, and Sept 12, 2023, 56 participants were randomly assigned to the intervention group (28 participants) or control group (28 participants). Participants had a mean age of 49·9 years (SD 8·1). 35 (63%) of 56 participants were women, 20 (36%) were men, and one (2%) was non-binary. 54 (96%) of participants were White and two (4%) were Asian. Participants had a mean BMI of 39·6 kg/m2 (SD 7·8) and a mean EDE-Q global score of 3·3 (0·4). 49 (88%) of 56 participants provided outcome data at 6 months and 45 (80%) at 1 year. At completion of the programme at 6 months, the mean weight loss was –13·9 kg (SD 11·2) in the intervention group and –3·7 kg (7·9) in the control group, with a between-group difference of –10·2 kg (95% CI –14·2 to –6·2). The between-group difference in the EDE-Q score was –0·8 points (–1·4 to –0·3) at 6 months, indicating non-inferiority. At 12 months, weight change was not different between groups, but non-inferiority and superiority in EDE-Q remained. No participants were suspected of having developed an eating disorder. 13 adverse events were documented, of which one, a cholecystectomy, was serious. Interpretation: Participation in a supported TDR programme did not worsen eating disorder symptoms in people with overweight or obesity and type 2 diabetes at high risk of developing an eating disorder. We found no evidence these programmes cause harm and a suggestion of benefit on eating disorder symptoms, independent of weight loss. Funding: Novo Nordisk UK Research Foundation.