The Diabetes Eating Problem Survey-Revised (DEPS-R) in a Greek Adult Population with Type 1 Diabetes Mellitus: Model Comparison Supporting a Single Factor Structure.

Calliope Karastogiannidou,Evangelia Kotzakioulafi,Triantafyllos Didangelos,Ioannis Samaras,Emilia Vassilopoulou,Parthena Giannoulaki,Ioana Corina Bocsan

doi:10.3390/nu13072375

Abstract

Type 1 diabetes mellitus (T1DM) patients occasionally develop disordered eating behaviors, leading to insulin manipulation without medical consultation, targeting to achieve weight control. In clinical practice, the Diabetes Eating Problem Survey-Revised Version (DEPS-R) questionnaire has been used to evaluate eating disorders in T1DM patients. This study was conducted to validate the factor structure of the Greek version of DEPS-R using Confirmatory Factor Analysis (CFA), to investigate its reliability and convergent validity in Greek T1DM adults and to compare a single factor DEPS-R model with multiple factor models. Participants were 103 T1DM adults receiving insulin, who responded to DEPS-R. Their anthropometric, biochemical and clinical history data were evaluated. The sample presented good glycemic control and 30.1% scored above the established DEPS-R cut-off score for disturbed eating behavior. CFA results revealed that the data fit well to the factor models. The DEPS-R scale had good reliability and was positively linked to BMI, HbA1c, total daily dose and time in range. Model comparison supported the superiority of the 1-factor model, implying that Greek clinicians and practitioners might not have to consider individualized treatment based on various scores across different subscales but they can adopt a single DEPS-R score for an easy and efficient screening for disordered eating.

Highlights

The women had borderline significantly lower time in range (TIR) compared with the men (Median = 68.5 vs. 79.5, p = 0.058), the TAR, time below mg/dL (TBR) and hypoglycemic events showed no difference between the two groups
Through confirmatory factor analysis (CFA), we focused primarily on comparison of the three-factor solution describing “Eating Habits”, “Thinness” and “High Blood Glucose” identified in previous research [9,10,16], with the single factor model originally proposed by Markowitz and colleagues (2010) [8]
We considered two additional alternative models: (1) a fourfactor model capturing the three aforementioned factors plus the “avoidance” factor that emerged in the Spanish study [14], (2) a two-factor model in which the “Eating Habits” and “Thinness” factors were combined in one factor, with “High Blood Glucose” as the second factor

Summary

Introduction

Diabulimia is a term, derived from diabetes mellitus (DM) and bulimia, for an eating disorder (ED) characterized by the intentional reduction or omission of a dose of insulin without consulting a doctor, in order to control weight [1]. By reducing the prescribed insulin dosage, the person with T1DM tries to achieve weight loss through acidosis, which causes urination and elimination of excess glucose (glucosuria). This process drives the body to burn fat to counteract the lack of available energy, which leads to the release of ketones, resulting in ketoacidosis [3,4]. Regardless of the underlying reasons, withholding insulin results in ketoacidosis, a condition that causes various complications in the body, especially when persisting for a long time [5]

Methods

Results

Discussion

Conclusion