Article Text

Download PDFPDF

Oats and coeliac disease
  1. R Dor1,
  2. D J Shanahan1
  1. 1Academic Unit of Medical and Surgical Gastroenterology, Homerton University Hospital, London E9 6SR, UK.
  1. Correspondence to:
    R Dor;
  1. R Julkunen2,
  2. M Uusitupa3
  1. 2Gastroenterological Unit, Department of Medicine, Kuopio University Hospital, Kuopio, Finland
  2. 3Department of Clinical Nutrition, University of Kuopio, PO Box 1627 70211, Kuopio, Finland

Statistics from

We read the study by Janatuinen et al (Gut 2002;50:332–5) with great interest. However, we would like to highlight some concerns.

The initial study1 included 92 patients randomised to two groups—oats and gluten free diet, 45 and 47 patients respectively—however, these numbers do not correspond to those in figure 1 of their paper.

Patients were verbally consenting volunteers, thus introducing selection bias to compliance. The number of dropouts, especially in the control group, was surprisingly high (41 in total). Were there so many dropouts because of the unpalatability of oats or concerns over their safety? If the latter, then surely this would be greater in the oats group.

It is not clear whether patients were followed up in the interim period between 12 months and five years. It would be important to ascertain objectively whether the oats group were in fact including oats in their diet, as omission would not address their long term safety. Assessment of compliance and food diary are very subjective, introducing bias towards dietary compliance. How was this done? We were surprised to see that one third of the oats group did not in fact take oats at all, so only one third of patients were ingesting oats on a daily basis. The proportion adhering to a gluten free diet was paradoxically greater in the control group. If oats were allowed then this could be interpreted by patients as acceptance of other (gluten containing) foods also. The purity and amount of oats ingested in the first year was regulated but this was not monitored thereafter and contamination of oat products could lead to small bowel architectural changes.

Histological changes in coeliac disease can be subjective but no mention was made as to whether blind reporting was carried out.

Also, no explanation was given for the greater numbers of high values for ARA, AGA, and EMA in both groups at five years. This needs to be quantified and statistically analysed.

The safety of oats in coeliac disease is extremely important. This and other studies have shown that compliance with a strict gluten free diet is difficult, reflecting its uncompromising limitations as well its relative unpalatability. This paper suggests that oats are safe but the small numbers involved could mask subtle differences between the groups which may result in greater morbidity. In order to address this, further larger multicentre trials are required.


Authors’ reply

Dor and Shanahan point out the inconsistency in the number of study patients in the flow chart in figure 1 of our study on oats ingestion by coeliac patients.1 Indeed, there is an error. The number of patients in the oats group in remission in the original study (Gut 2002;50:332–5) should be 26 and not 12. The figures given in the text are however correct.

Dor and Shanahan also emphasise the total number of dropouts (41) which covers both the original study of 6–12 months1 as well as the follow up period of five years. They ask whether this could be due to the possible unpalatability of oats. However, the number of dropouts was similar among controls (19) and those consuming oats (22), and there was no evidence that the oats products were unpalatable.1 Probably the number of dropouts in many cases reflects the lack of motivation to participate, especially in the unpleasant part of the study—that is, gastroscopy to obtain small intestinal biopsies. As expressed in the text, many patients also felt uncertain about the long term safety of oats. Our results show that such fear was not justified.

Patients had no strict follow up visits to the outpatient clinic of the University Hospital during the time period between 12 months and five years. However, they were advised to have regular follow up clinical and laboratory examinations in the local health care centres. Some also had regular visits to the outpatient clinic of Kuopio University Hospital.

As stated in the article, patients had free choice regarding the amount of oats they wished to consume and which they felt convenient. In this respect the study also reports on the reality and usefulness of oat products. Information on the quantity of oats in the diet and the degree of compliance was based on an interview and a questionnaire carried out by a clinical nutritionist.

Dor and Shanahan raise the question of whether patients interpreted the use of oats as comparable with consuming gluten containing foods. They base this on the degree of compliance in our study. Compliance was 71.4% in the oats group and 78.6% among controls. In our opinion this does not justify such a conclusion. As stated in the article, the oats group also consumed a strict gluten free diet but part of their gluten free products were substituted by oats.

After 12 months, patients used oat products from major Finnish mills. These products have been found to be free of contamination. If oats by itself or oat products had any deteriorating effect on the duodenal mucosa or stimulated immune mechanisms, the results in the oats group would have differed from those of controls. This was not the case. The results indicate that oats were well tolerated.

As in our original study (Gut 2002;50:332–5), histological examinations were carried out blindly by an experienced pathologist.

The values for the various antibody levels were statistically analysed comparing them at the end of the original study and at the five year time point. There were no statistically significant differences between time points or between the oats and control groups. The high values occurred only in a few patients in both the control and oats groups without any significant effect. As stated, those patients showed poor adherence to a gluten free diet.

We do agree with the final conclusion of Dor and Shanahan that larger multicentre trials could be worthwhile. However, to conduct a similar controlled study with larger numbers of patients and continuous frequent control of intake of oats, including multinational purity analyses of freely available oat supplements, would be an enormous effort. To date, our investigation is the largest controlled randomised study on oats in coeliac patients. Furthermore, it represents the first attempt in showing the long term safety of oats in coeliac patients.


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.