Article Text

Are Down syndrome and coeliac disease associated?
  1. M-A MORRIS,
  2. F BIAGI,
  3. H J ELLIS,
  4. P BRETT,
  1. Gastroenterology Unit (UMDS),
  2. Rayne Institute,
  3. St Thomas’s Hospital,
  4. London SE1 7EH, UK
  1. J M DUGGAN,
  2. L GALE
  1. Princeton Medical Centre,
  2. 60 Lindsay Street,
  3. Hamilton, NSW 2303,
  4. Australia

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Editor,—The paper by Gale et al(Gut 1997; 40: 492–6) fails in its aim to investigate the association between coeliac disease and Down syndrome. The authors use methodology for screening they have stated to be outdated. In a vulnerable population this is ethically unsound and inaccurate conclusions are drawn from incomplete data.

Editor,—The paper by Gale et al(Gut 1997; 40: 492–6) fails in its aim to investigate the association between coeliac disease and Down syndrome. The authors use methodology for screening that they have stated is outdated. In a vulnerable population this is ethically unsound and inaccurate conclusions are drawn from incomplete data.

The increased frequency of raised antigliadin levels (and of autoimmune antibodies) in individuals with Down syndrome are already well documented, as is the increased prevalence of coeliac disease.1 ,2 Incidence cannot be assessed by screening.

Serological screening by antigliadin antibodies alone is outdated and the usefulness of antiendomysial antibody levels, which can be easily measured using a commercial kit or human umbilical cord, is well established.3 To compound this error, the authors only regard individuals who have elevated titres of both IgA and IgG as being “gluten antibody” positive. The report does not exclude IgA deficiency in the eight controls positive for IgG alone. By the authors’ own admission, antigliadin IgA is more specific for coeliac disease, so individuals with positive IgA but negative IgG should be considered potentially coeliac and investigated further, with antiendomysial levels or by biopsy.

The failure to biopsy the “gluten antibody” positive controls prevents any conclusions about the prevalence of coeliac disease and the relevance of haematological or biochemical parameters in patients with Down syndrome compared with other institutionalised cohorts. The low serum albumin concentrations may reflect poor diet and feeding problems in an institutionalised setting and the control population values are not given for comparison. It is disappointing that no attempt was made to report the grading of the abnormal biopsy samples and the exclusion of giardia infection in this high risk group, purely on absence from a single duodenal biopsy specimen, is inadequate. Lymphocytic infiltration of the lamina propria can be considered to be “latent coeliac disease” but these changes may not resolve on a gluten free diet as was suggested. Formal gluten challenge, longer term follow up on a gluten containing diet with clinical and serological monitoring or further investigation are required to confirm the diagnosis.4 A diagnosis of coeliac disease has considerable implications and the failure of the authors to discuss EPSGAN criteria is unfortunate.

The importance of the genetic predisposition to coeliac disease is under emphasised and even dismissed. Monozygotic twin concordance approaches 100% with longer follow up of non-affected twins5 and dizygotic twin concordance is 25%, which is highly indicative of a genetic aetiology. HLA class II DQ2 molecules may well be involved in presentation of the toxic peptide(s) but segregation analyses suggest that at least one other non-HLA linked gene confers genetic susceptibility.6 Genome screening has already detected possible loci of interest, including one on chromosome 21.7

The estimate of increased prevalence uses as a comparison an observational report which calculated an incidence 10-fold lower than expected.8 Data from Australia are rare but the population is genetically similar to northern Europe and diet is also similar, so a broadly comparable prevalence of 0.1–2.93/10009 ,10would be expected. A calculated increased risk of 10-fold for individuals with Down syndrome is in keeping with published data.1 ,2

Research into the immune dysfunction found in Down syndrome and searches for genetic linkage may be more informative.



Editor,—We thank Morris et alfor their interest in our article and would like to clarify the points they have raised. Our research commenced in 1992 when only antigliadin antibody testing was available to us. Our study was prompted by a desire to clarify the link between Down syndrome and coeliac disease already reported in numerous anecdotal and inadequate reports. More recent reports documenting this link more strongly appeared after our research was completed. Likewise, at the time of our study an elevated level of both IgA and IgG antibodies was the accepted criterion for biopsy.1-1 Despite this lapse in time between the study and its submission for publication, we felt it was important to proceed because of the continued widespread use of antigliadin antibodies and the importance of reporting their shortcomings.

We did exclude IgA deficiency in the eight controls who were IgG positive only. We did not biopsy the two gluten antibody positive controls as the decision to do this was not ours; moreover, the increases were barely above cut off values for IgA in both and for IgG in one.

The lowered serum albumin concentrations in the Down syndrome group are unlikely to be due to diet. These patients did not have feeding difficulties and low serum albumin is uncommon in this setting, even in a more vulnerable group who do have severe feeding difficulties.1-2 In any case, dietary factors cannot explain a difference in albumin concentrations between those with negative and those with positive antibodies.

As far as giardia is concerned, the routine was to take multiple (three) biopsy specimens per examination and each specimen was closely examined. We did under emphasise the genetic aspects of coeliac disease, but this was motivated by our feeling that these were accepted well enough to not require comment, rather than a dismissal of them.

Concerning EPSGAN criteria, in an institutional setting with severely disabled patients and multiple caregivers, compliance with the diet cannot be guaranteed. This makes it very difficult to draw conclusions about the correctness of the diagnosis. While these difficulties need not preclude adherence to EPSGAN criteria, they render interpretation of subsequent biopsy samples difficult.


  1. 1-1.
  2. 1-2.