Amin et al from Oxford reported their findings of longitudinal growth characteristics and glycemic control in children with type 1 diabetes along with celiac disease (CD).  Annually, from 1994 and 1998, 230 children with type 1 diabetes were screened starting in the first year after the onset for the presence of IgA and anti-endomysial antibodies (EMA).  A total of 10 children were EMA positive and another one was AGA positive, which was 4.8% of the clinic population.  Only one patient demonstrated symptoms typical of CD, including failure to thrive and steatorrhea; four complained of some mild abdominal discomfort.  Jejunal biopsy showed classical histopathology of CD in all eleven patients.  These subjects were matched for age, sex, and diabetes duration with two control diabetic children who were negative for EMA.  Height, weight, and HbA_1c were measured at the time of diagnosis of CD and every 3 months.  Antibody levels were tested every 3 months until negative, and then yearly. The ANOVA model was used to determine the influence of CD on both HbA_1c and BMI SDS.  The data are presented as mean + SEM.   

Mean BMI SDS in the CD group was significantly lower (-1.2+ 0.1 vs. –0.1 + 0.1, P=0.005), as was mean weight SDS (-0.7+ 0.3 vs. 0.5 + 0.3, P=0.002) than in those without CD.  However, there was no difference between the two groups mean height or C-peptide level. Mean age of diagnosis of CD was 11.2 years (2.2-17.3).  The mean duration of diabetes at diagnosis was 3.8 years (0.9-7.2). Mean HbA_1c was significantly lower at diagnosis in the children with CD (8.9%+ 0.3% vs. 9.8% + 0.3%, P=0.002), but there was no difference in the mean daily insulin dose in the two groups.  The difference in mean BMI SDS between the subjects and the controls was eliminated by 12 months of gluten-free diet (1.1+ 0.13 vs. 1.0+ 0.1, P= 0.11).  HbA_1c (Figure 1) levels were lower than in the controls during the period of gluten-free diet (8.3+ 0.2 vs.10.0+ 0.2, P=0.002). Insulin requirements increased in both groups, but no difference in those requirements developed between the two groups.  Using a general factorial linear model, CD was associated with lower BMI SDS and lower HbA_1c across time, independent of other factors such as insulin dose and regime. Also, while on a gluten-free diet, the children with CD had lower HbA_1c which was independent of BMI SDS or the insulin dose or regimen.  The EMA antibodies tended to disappear while the patients were on the gluten-free diets.

The authors reviewed recent reports regarding the association in children between type 1 diabetes and CD. Prevalence rates range between 1.7 to 10%.  However the data on whether intervention with gluten-free diet would be of benefit remain controversial.  This is, in part, because there are few longitudinal follow-up data and few age and sex matched controlled studies.  The authors note that their findings could have been influenced by the small sample size or the increased input by dieticians which was received by case subjects.  They stress, that because the long-term complications of CD include gastrointestinal malignancy, lymphoma, infertility, and osteoporosis, the screening of children with type 1 diabetes at a young age may be cost effective and warranted.

Editor’s Comment:  These findings are very intriguing.  Many pediatric endocrine clinics are now screening children with type 1 diabetes for EMA or tissue transglutaminase IGA to identify CD. There is controversy as to whether or not children who are asymptomatic with their CD will benefit from a gluten-free diet, and whether or not there is any effect of a gluten-free diet on the management of their diabetes. Amin and co-workers have demonstrated that indeed children with CD and type 1 diabetes are anthropometrically different from those children without CD, and that treatment reverses this finding.  In addition, there appears to be a treatment benefit on overall glucose control. The authors noted that their data could have been influenced by the frequent visits to the dietician by case subjects.  It will be important to determine whether gluten-free diet is of benefit in all children with diabetes, and or whether similar nutritional input to all type 1 diabetic children could improve HbA_1c to the extent observed in this study.

William L. Clarke, MD