LAUSR

OBJECTIVE The prevalence of emotional eating (EE) has increased in the general population over past decades. There is limited information on how common EE is among people seeking obesity treatment. We aimed to estimate the proportion of people with EE, and strength of associations between a predefined set of factors and EE in people referred for obesity treatment. METHODS Cross-sectional study recruiting 387 adults from a hospital obesity service. "Emotional eating" was defined as Emotional Eating Scale (EES) score ≥25. Strength of associations were estimated by boot-strapped quantile regression analysis. Results are presented as quantile difference (QD) of EES scores at the 25th, 50th or 75th quantile, and 95% confidence intervals (95%CI). RESULTS The study population consisted of 71% women, with a median age of 52 years (interquartile range [IQR]: 42, 61), and a median body mass index of 42 kg/m2 (IQR: 37, 49). 187 participants were managed with lifestyle modification alone, 103 with the addition of obesity pharmacotherapy, 79 with bariatric surgery, and 18 with both bariatric surgery and medications. EE was reported by an estimated 58% (95%CI: 53, 63) of participants. Factors with the largest and most consistent magnitude of association with EES differences include age, sex, use of glucagon-like peptide-1 (GLP-1) agonists, history of sleeve gastrectomy and recent bariatric surgery. CONCLUSION Emotional eating affected more than half of people referred for obesity treatment. Age, sex, use of GLP-1 agonists, history of sleeve gastrectomy and recent bariatric surgery had the strongest associations with EE. These findings allow hypothesis generation about the underlying physiological mechanisms behind emotional eating for investigation in future research.