The prevalence of eating disorders is particularly high among people with type 1 diabetes (T1D). These abnormalities, such as "diabulimia," are frequently responsible for poor insulin therapy management and, consequently, chronic glycemic imbalance, exposing them to an increased risk of complications. Their detection and management unfortunately remain insufficient in current practice.

However, to date, no study has addressed the question of the prevalence and impact of eating disorders in this context. Our research hypotheses are therefore as follows:

1. The existence of an eating disorder is likely frequently overlooked and, therefore, not taken into account in patients with T1D initiating semi-automated insulin therapy with BF.
2. The existence of an eating disorder could impair the performance of the BF device in terms of improving glycemic control in patients with T1D initiating this treatment method.

Patients clinical characteristics, glycemic monitoring parameters ad questionnaires answers will be recorded. The prevalence of eating disorders will be calculated, and the association between the presence of these abnormalities and baseline clinical characteristics and glycemic control parameters will be analyzed.

The prevalence of eating disorders (EDs) and problematic eating behaviors (PEBs) is particularly high among people with type 1 diabetes (T1D), affecting 20 to 50% of women and 5 to 25% of men among young patients. These abnormalities, such as "diabulimia," are frequently responsible for poor insulin therapy management and, consequently, chronic glycemic imbalance, exposing them to an increased risk of complications. Their detection and management unfortunately remain insufficient in current practice.

The treatment of T1D is currently undergoing a veritable technological revolution linked to the arrival of the first semi-automated insulin therapy devices (or closed loop, LF). The use of these devices most often makes it possible to achieve recommended glycemic control targets (HbA1c, continuous glucose monitoring [CGM] data), reduce glycemic variability, and avoid hypoglycemia. Insulin administration is based on the data transmitted: food intake and quantity of carbohydrates consumed. The role of dietitians is therefore crucial in the therapeutic education of patients and the care pathway imposed by the initiation of this new treatment method. It is likely that many patients initiating automated insulin therapy with BF present an eating disorder or CAP that may interfere with the management of the device.

However, to date, no study has addressed the question of the prevalence and impact of eating disorders and CAP in this context. In the absence of available data, national and international recommendation texts dedicated to insulin therapy with BF do not specify appropriate conduct for patients presenting with these eating disorders. Our research hypotheses are therefore as follows:

1. The existence of an eating disorder (ED and/or CAP) is likely frequently overlooked and, therefore, not taken into account in patients with T1D initiating semi-automated insulin therapy with BF.
2. The existence of an eating disorder (ED and/or CAP) could impair the performance of the BF device in terms of improving glycemic control in patients with T1D initiating this treatment method.

Patients clinical characteristics and glycemic monitoring parameters (HbA1c and CGM data), available as part of routine care, will be recorded in an electronic case report (REDCap software). The mSCOFF and QACD questionnaires will be administered during the dietary interview preceding BF device implantation, and responses will be collected. The prevalence of eating disorders will be calculated, and the association between the presence of these abnormalities and baseline clinical characteristics and glycemic control parameters will be analyzed.

The care pathway established for initiating BF treatment requires a follow-up medical consultation at 3 months, followed by a consultation at 6 months. As part of the study, visits at 3 months (12 ± 2 weeks) and 6 months (24 ± 4 weeks) will be scheduled and will include an interview with a dietitian, to collect glycemic control parameters (at 3 and 6 months), measure body weight (at 3 and 6 months), and repeat the mSCOFF and QACD questionnaires (at 6 months). Data on the evolution of glycemic control and variability parameters, weight and mSCOFF and QACD scores will be analyzed in the overall study population, then according to the existence or not of disturbances in eating behavior at inclusion.

Inclusion Criteria:

• Patient diagnosed with type 1 diabetes.
• Patient with a CGM device for more than 3 months with > 80% of data captured in the last month.
• Indication for the implementation of a BF device as part of routine care, in the Diabetology, Metabolic Diseases and Nutrition Department of the Toulouse University Hospital (initiating center).

Exclusion Criteria:

• Pregnant or breastfeeding women, or those planning to become pregnant within 6 months of inclusion.
• History of obesity surgery or a preparatory program for obesity surgery planned during the study period.
• Pharmacological treatment likely to significantly influence food intake and body weight, whether ongoing or discontinued within the last 3 months (e.g., GLP-1R agonist): Liraglutide (Victoza®, Saxenda®), Semaglutide (Ozempic®, Wegovy®), Dulaglutide (Trulicity®), Tirzepatde (Mounjaro®), Orlistat (Xenical®)
• Eating disorder requiring specific treatment in the previous 12 months.
• Refusal or inability to consent to participate.
• Persons under legal protection (guardianship, curatorship, or court-ordered protection).

patients with type 1 diabetes for whom closed-loop insulin therapy has been recommended