Article Text

Download PDFPDF

IgG antibodies to foods in IBS
  1. P J Whorwell,
  2. K J Bentley,
  3. W Atkinson,
  4. T A Sheldon
  1. University Hospital of South Manchester, Manchester, UK
  1. Correspondence to:
    Professor P J Whorwell
    Department of Medicine, Education and Research Centre, Southmoor Road, Manchester M23 9LT, UK;

Statistics from

Mawdsley et al raise the important question as to whether patients with irritable bowel syndrome (IBS) would gain as much symptomatic improvement if recommended to exclude the top four foods (yeast, milk, whole egg, and wheat) compared with an IgG antibody test based diet.1 In other words, does the test add specificity? This requires a trial which compares patients receiving an IgG antibody test based diet to those advised to eliminate some or all of the top four foods. We are currently seeking funding for such a trial.

There is some evidence however from our trial that the IgG antibody test based diet may provide a better response than simply eliminating a standard set of foods. When the change in IBS symptom severity score was compared for fully adherent true and sham diet patients who were advised to eliminate one or more of the top four foods, it was found that the true diet patients experienced a significantly greater reduction than the sham diet patients (difference = 94; 95% confidence interval 18, 170; p = 0.017).

We agree with Sewell’s comment that the food elimination diets in the true and sham groups were not similar in terms of content,2 although they were for numbers of food types excluded. This was to some extent inevitable given the high prevalence of IgG antibodies to certain foods, such as yeast (86.7%) and milk (84.3%). However, exclusion was not quite as unbalanced as implied as the so-called sugar foods were allowed in the “yeast positive” patients. While we accept that a more balanced comparison would have been desirable, the principal point of the sham diet was to control for placebo effect. In future, more care needs to be taken to match diets not just for number of food types excluded but also for types of food. We are still confident, however, that the difference in symptom improvement observed in our study for the true and sham diet groups is a real one. This is evidenced by the highly significant difference in worsening of symptoms between the true and sham groups when patients reintroduced foods they had been asked to exclude (p = 0.003).


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.