Anthropometry, Metabolic Control, and Thyroid Autoimmunity in Type 1 Diabetes with Celiac Disease: A Multicenter Survey
© 2005 Prime Health Consultants, Inc.
The DPV-Wiss database is a central registry for the documentation of treatment processes and outcomes in children with type 1 diabetes (T1DM) in Germany and Austria. Data are gathered and available for analysis from 150 pediatric departments and 19,796 subjects (ages 0.1 to 19.9), representing approximately half the children with T1DM in Germany. Kaspers et al report their findings regarding the association between T1DM and celiac disease (CD) and anthropometrics and metabolic control. Although there is no consensus for measuring celiac and thyroid antibodies in children with diabetes, in Germany these determinations are commonly performed every 1 to 2 years.
Three groups of diabetic children were identified: Group 1 comprised no clinical or biochemical signs of CD (n=18 470); Group 2 had 1 or more significantly elevated CD-associated antibodies (IgG and IgA antibodies to gliadin, endomysium, or tissue transglutaminase–IgA antibodies), but no jejunal biopsy performed or recorded (n = 1 119); Group 3 exhibited biopsy-proven CD (n = 127). At least one CD associated antibody elevation was present in 6.7% of the cohort, while 0.6% had histologically confirmed CD. The mean age of diagnosis of CD was lower than that of the Group 1 subjects (12.2 ± 4.6 years vs 13.4 ± 4.3 years, P< 0.05) and the onset of T1DM was at an earlier age in Group 3 children (5.8 ± 4.0 vs 8.2 ± 4.0 years, P<0.001). The average age of diagnosis of CD was 4.3 ± 3.8 years. In 13 children, CD was diagnosed prior to the diagnosis of T1DM; 57% of Group 3 children were female.
Standing height was significantly reduced in Group 3 children (–49 ± 1.1 vs –0.06 ± 1.0 height-SDS, P<0.001), and this difference increased over time. Growth of children less than 11 years of age was more affected than that of older children. BMI was also lower in the CD children and did not improve over time. Daily insulin requirements, number of daily injections, and number of severe hypoglycemic episodes did not differ among groups, but HbA1c levels were significantly lower in the Group 3 children at CD diagnosis (8.1% ±1.8% vs 8.8% ± 2.4%, P<0.001) and remained lower during the observation period. The incidence of thyroid disease was greater in the group with T1DM and CD (6.3% vs 2.7%, P<0.02).
The authors noted that the actual incidence of CD in their population may be greater than recorded, since all children with positive antibodies did not undergo jejunal biopsy. The failure of children with CD to exhibit catch-up growth may have been the result of poor compliance with the gluten-free diet. The growth reductions seen in these children would account for a loss of 5.2 cm in final height. The lower HbA1c values in the Group 3 children could have been the result of malabsorption of nutrients or perhaps more meticulous attention to carbohydrate amounts and source when using the gluten-free diet. The authors concluded that the recommendations for regular screening of children with T1DM for CD beginning at disease onset is both rational and important, as untreated CD is associated with significant long-term health risks such as osteoporosis and intestinal lymphoma.
First Editor’s Comment: This paper presents important data confirming the relationship between CD and T1DM; it lends evidence for the need to screen children with T1DM for celiac associated antibodies at the time of diagnosis and throughout childhood. The findings of reduced height and BMI, not related to poor glycemic control, are of concern. The implementation of a gluten-free diet is difficult and compliance is even more difficult to assess than that of the more routine carbohydrate-counting caloric recommendation diet usually prescribed for children with T1DM. It will be important for studies such as the current one to continue in order to assess the long-term impact of living with both diseases. Information regarding the development of osteoporosis and lymphoma in asymptomatic children is especially important. The current database may assist in the collection of such information.
Currently, some third-party payors in the United States are reluctant to reimburse for the laboratory determinations of celiac associated antibodies in children with T1DM who do not exhibit typical symptoms and signs of CD. The data from this German database can be used to help justify such screenings.
William L. Clarke, MD
Second Editor’s Comment: It is important to distinguish CD from celiac autoimmunity (CA). The former presents with symptoms and abnormalities on the intestinal mucosa and function which improve with a gluten-free diet; the later presents no alterations other than positive IgA transglutaminase antibodies and enodmysial antibody immunofluorescence IgA. The natural history of CD is well known, including chronic intestinal malabsorption and long-term risks of osteoporosis and lymphoma. However, no such data exist for CA patients who have no symptoms or alterations in jejunal morphology and function; however, monitoring height and weight progression is warranted. Since adherence to a gluten-free diet is poor even in the CD patients, the added burden to diabetic individuals would need to be considered before recommending it when CA is encountered. The reader is referred to GGH1 for a review and commentary of a paper by Hofferberg2 on the natural history of CA.
Fima Lifshitz, MD
Reference - (linked to 
)