(DCP042) FACTORS ASSOCIATED WITH INTENTIONAL INSULIN OMISSION TO LOSE WEIGHT IN PEOPLE WITH TYPE 1 DIABETES: A BETTER REGISTRY ANALYSIS

Background: Intentional insulin omission (IIO) to lose weight is a compensatory behaviour related to eating disorders, specific to people with type 1 diabetes (PwT1D). This behaviour can lead to serious medical complications. It is not fully known what factors may predict this behaviour. More evidence on predictors of this behaviour is needed to form targeted interventions to prevent IIO.

Objectives: 1) to compare demographic, physical, dietary, and diabetes-related outcomes by those who have versus have not reported IIO. 2) Explore possible predictors of IIO.

METHODS AND RESULTS: This was a cross-sectional study of PwT1D over the age of 14 years in the Behaviours, Therapies, Technologies, and Hypoglycemic Risk in Type 1 Diabetes (BETTER) registry. People in the BETTER registry were included if they were on an insulin regimen, were not pregnant, and answered the question ‘have you ever intentionally omitted insulin to lose weight?’. T-tests (for means), and chi-square tests (for proportions) were used to determine the difference in demographic, physical, dietary, and diabetes-related outcomes in PwT1D. Multivariable logistic regression model was used to determine predictors of IIO. A total of 751 PwT1D were included in the analysis; 5.6% (n=42) reported IIO. On average, those who endorsed to IIO were more likely to report nephropathy (21.4% v. 10.3%; p=0.05), and neuropathy (31.0% v. 16.8%; p=0.03), but not retinopathy (26.2% v. 22.4%; p=0.71). People who reported IIO and completed a 24-hour food recall had higher caloric (2,783 kcal (1,334 SD) v. 2,072 kcal (704 SD); p< 0.001), and alcohol intake (46g (81 SD) v. 7g (16 SD); p< 0.001). Women, compared to men, had higher odds of IIO (OR:1.34 95%CI: 0.36, 2.59). Additionally, diabetes distress score (0.51, 95%CI: 0.18, 0.83) and weight cycling occurrences (0.14, 95%CI: 0.04, 0.19) were associated with IIO, while BMI 0.016 (95%CI: -0.05, 0.08) was not.

Conclusion: PwT1D who acknowledged IIO were more likely to also report diabetes-related microvascular complications and higher caloric and alcohol intake compared to those who did not endorse IIO. Status as a woman, diabetes distress, and weight cycling were predictors of IIO, while BMI was not in PwT1D.