Trials / Recruiting

RecruitingNCT07348432

The diabEAT Study: Insulin dElivery Technologies And eaTing Behaviours in People With Type 1 Diabetes

Status: Recruiting
Phase: —
Study type: Observational
Enrollment: 106 (estimated)

Sponsor: McGill University · Academic / Other
Sex: All
Age: 12 Years
Healthy volunteers: Accepted

Summary

Type 1 diabetes is an autoimmune health condition that requires daily injections of insulin. Insulin allows the body to use energy from carbohydrates in food. Disordered eating behaviours, like restricting food intake to lose body weight, are more common in women and people with type 1 diabetes, compared to those without because they must practice carbohydrate counting. Carbohydrate counting means identifying, measuring, and planning carbohydrate intake to match insulin dosage. New technologies, such as automated insulin delivery (AID) systems adjust insulin delivery in a blood sugar responsive manner. AID is rapidly replacing conventional insulin delivery like injections or non-automated insulin pumps since it reduces management burden and improves blood sugar levels. It is not known if AID reduces food management and disordered eating behaviours. This study aims to: 1. investigate the relationship between AID and eating behaviours according to gender for youth (12 to 17 years), and adults (18 years and older). 2. Determine the limit of carbohydrate counting inaccuracy to maintain stable blood sugar levels according to insulin delivery method (AID, injections, or pumps). It is hypothesized that those who use AID will have lower disordered eating behaviours and will maintain stable blood sugar levels while allowing for higher carbohydrate counting inaccuracy. This will be a cross-sectional cohort study of people with type 1 diabetes who are 12 years of age or over. Participants will be recruited through the BETTER registry and social medias across Canada. This research is needed to improve nutrition guidelines for type 1 diabetes in the context of new technologies like AID. Evidence from this study may reduce food management burden, lower the risk of disordered eating behaviours, and prevent eating disorders and medical complications.

Detailed description

Introduction: Type 1 diabetes occurs when the pancreas cannot produce insulin and a person must be given insulin exogenously. Managing this health condition involves strict diet planning including carbohydrate counting (i.e. identifying and counting carbohydrates to match insulin dose). In addition, to high food management burden, frequent bodyweight monitoring, and weight gain related to insulin usage, makes people living with type 1 diabetes more susceptible to disordered eating behaviours (like intentional food restriction), compared to those without. Automated insulin delivery systems (AID), which automatically adjusts insulin as a response to continuously measured blood glucose levels has shown to improve quality of life, and improve glycemic levels, however it's impact on eating behaviours and diet is unknown. There is also a need to determine the carbohydrate counting inaccuracy threshold to maintain glycemic stability depending on the type of delivery system used (AID vs. Not). Objectives: 1. Determine the relationship between AID and eating behaviours 2. Determine the carbohydrate counting inaccuracy threshold to maintain glycemic stability and understand whether AID use modifies this relationship. Methods: This is an observational cross-section analysis of people with type 1 diabetes. Eligibility criteria included: those living with type 1 diabetes for more than 1 year, using at least 2 insulin injections per day or using an insulin pump, who were living in Canada, 12 years of age or older, and using their current insulin delivery system for 3 months or more. Participants are excluded if they were pregnant/currently breastfeeding and did not speak English or French. Demographic, and diabetes-related information (including AID use), are determined through an initial questionnaire, which takes about 15 minutes to complete. Disordered eating behaviours were determined through validated questionnaires. The Three Factor Eating Questionnaire (TFEQR-21) identified behaviours such as cognitive restraint (score ranged from 6 to 24), emotional eating (score ranged from 6 to 24), and uncontrolled eating (score ranged from 9 to 36). The Diabetes Eating Problem Survey Revised (DEPS-R) identified DEBs specific to diabetes (score ranged from 0 to 80 with \> 20 representing those at risk of DEB). The Teruel Orthorexia Scale (TOS) identified orthorexia nervosa behaviours defined as an obsession with healthy eating which may lead to emotional impairments (score ranged from 0 to 24). Dietary intake information will be collected through a 4 day picture food journal (3 weekdays and 1 weekend) application called Keenoa and analyzed by a Registered Dietitian. Glycemic outcomes such as glucose time in range (TIR), which measures the amount of time glucose levels are between 3.9-10.0mmol/L, and Coefficient Variation, were reported through a 14 day CGM report (Clarity, Dexcom, Medtronic). Physical activity will be measured through an Actigraph GT3X for 8 days. Descriptive analysis will be completed at enrollment, to determine the mean (SD) socio-demographic information, eating behaviour scores and dietary intake (macro/micronutrient profile) by insulin delivery system (AID and injections/insulin pumps). Multivariate linear regression will be used to determine the relationship between AID compared to injections/insulin pumps and disordered eating behaviour scores. Secondly, carbohydrate counting inaccuracy will be determined by comparing participant reported carbohydrate counts measured in mean (SD) grams per meal to RD measured carbohydrate counts by analyzing 4-day dietary reports, as generated by Keenoa. Carbohydrate counting inaccuracy threshold will be determined through multivariate linear regression by exploring the relationship between percent carbohydrate counting inaccuracy and % of glucose Time in Range and Coefficient Variation. Type of insulin delivery will be used as an effect modifier to determine how this relationship is modified by AID and injections/insulin pumps.

Conditions

Interventions

Type	Name	Description
DEVICE	Automated Insulin Delivery (AID) Systems	AID automatically adjusts insulin delivery by using continuously measured blood glucose levels. AID use will be determined through the initial questionnaire through the following questions: Do you currently use the pump as an automated insulin delivery system (connected to a CGM with automated insulin adjustments)? Yes, a commercial AID with control IQ (Tandem) or SmartGuard (Medtronic) Yes, a non-commercial open-source do-it yourself AID (e.g., Loop) No, they use it as a manual (non-automated) pump or with a suspend on low functionality (e.g., Basal IQ) I prefer not to answer I don't know The type of AID system (hybrid, advanced hybrid, etc.) will also be confirmed. The exposure variable will be coded as a binary-categorical variable of AID use (yes or no) with no representing all other non-AID insulin pumps or injections.
BEHAVIORAL	Carbohydrate Counting Inaccuracy Percentage	Carbohydrate counting inaccuracy: will be determined by subtracting the estimated carbohydrates (by participant) by the actual amount of carbohydrate (through diet analysis) divided by the actual amount of carbohydrate, multiplied by 100, to determined the percentage. Estimated carbohydrate counts will be entered at each meal and snack by the participant in a daily log provided to the participant. Carbohydrate amounts will be collected through a 4-day food diary through the phone application Keenoa (carb count from the app will be blinded to the participant), and reviewed by a research assistant with education in dietetics.

Timeline

Start date: 2024-07-29
Primary completion: 2026-01-31
Completion: 2026-05-01
First posted: 2026-01-16
Last updated: 2026-01-16

Locations

1 site across 1 country: Canada

Source: ClinicalTrials.gov record NCT07348432. Inclusion in this directory is not an endorsement.