Prevalence of Eating Disorders in Young Patients with Type 1 Diabetes Mellitus from two Different Italian Cities

The prevalence of eating disorders and behaviors was evaluated in two populations of adolescents with type 1 diabetes from two cities in Italy. In this report, we will establish the relationship of these disorders with sex, BMI, socioeconomic status, metabolic control, and compliance to therapy. A total of 193 patients with type 1 diabetes aged 8–18 years (mean 13.6 ± 2.7 years; 92 female and 101 male subjects) were recruited from the pediatric diabetology units at the Second University of Naples (n = 118, 56 female and 62 male subjects) and at the University of Parma (n = 75, 36 female and 39 male subjects). All patients were affected by type 1 diabetes without evidence or history of other autoimmune diseases (thyroiditis, celiac disease, etc.). The distribution of sex, age at the onset of diabetes, disease duration, and BMI was statistically similar in the two groups. A total of 63.1% of the patients (64.6% from Naples and 35.4% from Parma) were from families with low socioeconomic status (according to the annual income and the parents’ level of schooling). The mean BMI was 21.45 ± 3.45 kg/m2. Almost all patients were on intensive insulin therapy (43% on three administrations/day and 54.8% on four administrations/day). All patients completed the Eating Disorder Examination Questionnaire (EDE-Q) (2,3), which was modified for diabetes (2,3) and the diabetes compliance scale (4). A total of 131 healthy control subjects from Naples and Parma, matched for age and sex, completed the EDE-Q. No major eating disorders, such as anorexia and bulimia nervosa, were found in patients with diabetes or in healthy control subjects. Otherwise unspecified minor eating disorders that do not meet the DSM-IV criteria for anorexia and bulimia nervosa, such as binge eating, overeating (with and without loss of control), and inappropriate compensatory behavior were more frequent in patients with diabetes than in control subjects (9 of 181 patients who answered this specific item vs. 1 of 131 control subjects, χ2 = 2.883, P = 0.09). Binge eating episodes were reported by 49.7% of diabetic patients and by only 24% of control subjects (P = 0.002). The presence of this disturbance was found more frequently in patients with low social status (P = 0.003). Objective overeating was present in 41.9% of patients and only in 16.9% of control subjects (P = 0.0001), while the difference between the report of subjective overeating was not significant between patients and control subjects. The prevalence of inappropriate compensatory behaviors such as voluntary vomiting, self-administration of diuretics and laxatives, and excess physical exercise have been found to be slightly more frequent in diabetic patients (9 of 181 patients who answered this specific item vs. 1 of 131 control subjects’ answers, χ2 P = 0.074). If we consider the skipping or manipulating of insulin dosage to lose weight as a sign of body dissatisfaction and therefore as an otherwise unspecified eating disorder, the total prevalence of eating disturbances in diabetic patients is significantly higher (χ2 P = 0.002) than in control subjects. We recognize that this may not be an entirely valid comparison because control subjects do not have the opportunity to manifest episodes of otherwise unspecified eating disorders by manipulating insulin doses. On the other hand, insulin omission and/or dose manipulation offers a unique resource to patients with diabetes to manifest his/her concern about body image. All together, the behaviors in our study were reported by 25 of 192 patients and 1 of 131 control subjects (χ2 = 12.273, P = 0.0001). The presence of eating disturbances was only slightly correlated to the reported compliance. In people with the highest mean score on the diabetes compliance scale (mean = 8), the prevalence of eating disturbances was 11%. When the mean score was lowest (mean = 4), it increased to 17% (χ2 = 5.331, P = 0.021). In conclusion, in our study, anorexia and bulimia nervosa are not common in adolescents and young adults with type 1 diabetes, while otherwise unspecified eating disorders seem to be more common than in healthy control subjects. There was no difference in eating disorder prevalence between control subjects and patients from two different cities and eating habits, but the prevalence appears correlated only to socioeconomic status and low compliance to therapy.