The Effect of Semaglutide on Disordered Eating Behaviour in Type 2 Diabetic Patients

This randomized controlled trial will be conducted in Clinical Hospital Center Sestre milosrdnice and University of Zagreb School of medicine. Patients will be randomized after completing the EAT-26 questionnaire (1:1), in group with disordered eating behaviour (n=30) if participants scored >20 points, and group without disordered eating behaviour (n=30) if participants scored ≤20 points. Group of patients with disordered eating behaviour will be further randomized whether participants will receive semaglutide (n=15) with standard care or if participants will be treated with standard care (n=15). The investigators will evaluate baseline EAT-26 score, GLP-1 and GIP blood levels in a mixed meal test, anthropometric measurements (weight, BMI, waist circumference), biochemical parameters (complete blood count, glucose, insulin, c-peptide, HbA1c, urea, creatinine, uric acid, aspartate aminotransferase (AST), alanine aminotransferase (ALT), gamma-glutamyltransferase (GGT), international normalized ratio (INR), albumin, serum lipid levels (cholesterol. HDL, LDL, triglyceride), C reactive protein (CRP), urine sediment and their changes at the end of the trial (except mixed meal test). Also, through 14 days at the beginning and 14 days before the end of the trial, patients will keep a food diary and have intermittently scanned continuous glucose monitor (FreeStyle Libre Pro CGM device, Abbott).

A factor analysis of EAT-26 scores will be conducted to form 3 (expected) latent variables (components). To assess the effect on dietary habits, the difference in EAT-26 total score and latent variables between subjects randomized to semaglutide and control subjects (randomized to standard care) will be assessed after 3 months of treatment. The data will be analysed in a general linear model, with basal covariates: age, sex, BMI and score at the beginning of treatment. In the same way, the effect of semaglutide on GLP1 concentrations (basal covariate: basal concentration next to the location of the EAT-26 score) will be evaluated, separately for the condition before and after the mixed meal test. GLP-1 and GIP concentrations will be compared between subjects with ("exposed") and without (control) eating disorders, in a generalized linear model for repeated measurements (GLP-1 concentration before and after mixed meal test) with fixed covariates: age, sex, BMI, time ( before or after mixed meal test) and exposure*time interactions. Type I (α) error=0.05.