Abstract
Background: There is a complex overlap and high prevalence rate between type 1 diabetes (T1D) and eating disorders. However, screening for disordered eating in young people with T1D may be avoided by diabetes professionals and deterred by parents. More ways of effectively screening for disordered eating in this population are needed and parent reports can be beneficial for early detection of symptoms and engagement in treatment.
Aims: This study aimed to validate a parent-reported version of a pre-existing and validated disordered eating screening tool for adolescents with T1D (the Diabetes Eating Problems Survey-Revised; DEPS-R).
Methods: The existing DEPS-R was adapted for parent use. During baseline assessment of the wider PRIORITY trial project, 73 parents of adolescents with T1D aged 11-14 years completed the parent-reported DEPS-R and other questionnaires related to demographics, child eating behaviours and quality of life. Adolescents of parent participants were also offered the opportunity to participate, and 40 adolescents completed demographic questions and the original DEPS-R measure.
Results: The parent-reported DEPS-R demonstrated good internal consistency (Cronbach’s = 0.827). Moderate to good inter-rater reliability was found between parent-reported DEPS-R and adolescent self-reported DEPS-R measurements (ICC was 0.8, 95% CI = 0.623-0.894, p <0.001), indicating good convergent validity. Construct validity with specific hypothesised variables, including certain eating behaviours, diabetes-related distress, overall parent quality of life, adolescent BMI, parent gender and parental worry about adolescent disordered eating, suggested that the measure might be valid. However, some hypothesised variables did not significantly correlate with the parent-reported DEPS-R, as expected.
Conclusions: The parent-reported DEPS-R has demonstrated good reliability and validity and it may provide clinical benefits of increasing screening and early detection of disordered eating in T1D adolescents. These findings are tentative, due to missing data and small adolescent sample size, and therefore further validation is needed.