What is an acceptable amount of gluten in a ‘gluten-free’ product? Professors Catassi, Ciclitira and Fasano give their thoughts

At the 14th International Coeliac Disease Symposium held in Oslo in June 2011 Professors Carlo Catassi and Paul Ciclitera went ‘head to head’ on what can be considered a ‘safe’ level of gluten in a gluten free diet. These are a few comments drawn from their talk.

Below are also points drawn from a paper by Professor Alessio Fasano, Director, University of Maryland Center for Celiac Research, with reference to the US Food and Drug Administration's (FDA) proposals for establishing levels for gluten-free products in the US. Click here for the full text.

Professor Paul Ciclitera

Professor Ciclitera looked first at the history of gluten free products including his own work, some years ago, on products sold as gluten free. In the process of this work he analysed what was then sold as ‘slimming’ bread (made by a company called Energen) and developed a serological test for measuring gluten in these foods. He was surprised to discover that he could measure, immunologically, the amount of gluten in the basic product that was being used to treat patients with coeliac disease. Until that time there had been no serological assays and therefore the only way of measuring 'gluten' in ‘gluten-free’ products was to measure the Kelgar Nitrogen or wheat starch – a very non-specific method. However, his serological test suggested that there was roughly 300ppm (parts per million) of gluten in this wheat starch.

But although this was, therefore, a relatively contaminated product it is important to realise that at that time it was, essentially, the only food available for making commercial gluten free products.

This level of 300ppm was relevant to a whole series of clinical studies which were undertaken in Oxford at that time and which put people on diets with wheat starch based products containing a significant amount of gluten. He took biopsies at the start and at the end of the study, and was able to show there was actually damage to the mucosa using these products.

Shortly after this the allowed level was reduced to 200ppm which, in Professor Ciclitera’s opinion, was a relatively good, working practical value. Since then the allowed levels have been reduced to 100ppm (for ‘gluten-reduced’ or ‘low gluten’) and to 20ppm for 'gluten free'. However, he was concerned that this was going to cause confusion not only among the patients but among health professionals who would not know what to advise patients in terms of what they can, or cannot, safely eat. He also asked whether this lower limit was going to, effectively, outlaw wheat-starch based products which currently form the basis of nearly 80% of gluten-free products in northern Europe.

He pointed out that the situation is further confused by the fact that regulations differ both across Europe and in the US. Until recently wheat starch products were not allowed in Italy, for example, but now they are, while in the US assessments are done on symptomatology rather that on scientific evaluation. So, for example – a gluten free diet that is low in fibre may induce the symptoms of Irritable Bowel Syndrome, especially in the female patients between 16-40. These symptoms are then attributed to coeliac disease when in fact they are caused by the low fibre content of the diet. 

If wheat starch is outlawed as a basis for gluten-free products, patient choice is going to be restricted. It is already difficult to get patients to adhere to the gluten-free diet – an especial problem with girls between the age of 16-18 as, in his experience, they are terrified that the boyfriend is going to think they're a freak because they are eating a gluten free diet… 

And the regulations themselves present a problem. How do you interpret them?  Gliadin can be measured using the current serological assays.  The only assay that is allowed in America is the R5 assay: you measure gliadin and you multiply by 2 to obtain the gluten values.  The professor’s own group and others have shown that glutenin exacerbates. Glutenines constitute 50% of gluten. If you measure, the gliadin and the gluten content independently, particularly in wheat starch based products, you find that the actual value is up to 10 fold either too high or too low compared to the number that you've got. So that the true gluten content of wheat starch is not, in the professor’s opinion, quantifiable using the methods that are available.  So although these very strict levels of 20 and 100ppm have been introduced, they do not necessarily indicate the true gluten level. 

Several studies have concluded that wheat-starch-based products are well tolerated and well liked by the majority of coeliac patients (hence induce better compliance with the diet) while there is a lack of data to suggest that 100ppm is harmful. Professor Ciclitira would therefore suggest that wheat starch based products and a maximum of 100ppm of gluten should be the rule as he fears that more stringent regulations will not only result in worse patient compliance with the diet but in gluten free product manufacturers pulling out of the market because the financial implications of the stricter regulations will mean that it is not longer financially viable to produce the products.

Professor Carlo Catassi

Professor Catassi started by showing some slides demonstrating that coeliac patients treated with a gluten free diet may have improvement of clinical symptoms of serology but their intestinal mucosa often do not normalise. 
In active coeliac disease the villi disappear. After treatment with the gluten free diet there is improvement of the small intestine mucosa but usually it doesn't go back to a normal; the villi remain shorter than normal so there is still some damage on going. In a second large study in Italy there was improvement of the small intestine mucosa in the vast majority of cases but genuine normalisation occurred in only a few. There were also some subjects who had no change actually and a small percentage who even deteriorated.

So, even on the gluten free diet, there are problems and many believe that it is persistent gluten ingestion that is the cause of this persistent intestinal damage.  There are many studies showing that when the adherence to the treatment is poor the proportion of patients who have still a histological disease is higher than when the compliance is good. 

So, may persistent gluten contamination may be the cause of persistent intestinal damage and, if so, what are the causes of the persistent gluten contamination? 

There are, of course, several different possibilities. There is voluntary transgression and involuntary transgression. We know that there may be gluten in many products (including infant formula) that do not appear to contain gluten. Professor Catassi believes that there is a significant problem with products based on wheat starch made for patients with coeliac disease as regulations governing the amount of gluten they can contain is not strict enough. Although it is difficult to reach a zero level manufacturers really should aim to reach this situation. 

Studies done to establish the level of sensitivity to gluten of coeliac patients show, unfortunately, that it is usually very high. It is also very variable between patients but as little as 50mg of gluten are enough to cause a small amount of  damage to intestinal mucosa. 

The amount of gluten in a normal, non-gluten-free diet this is around 20 grams a day; far higher than that needed to produce a measurable effect on the small intestinal mucosa. Professor Catassi showed the results of a recent study in which, for three months, they gave small amounts of gluten to three groups of celiac patients: one group received no gluten at all (the placebo group), one group received 10mg per day and one group 50mg a day.  

They found was that in the group receiving 50mg per day there was a clear deterioration of the villus height and inflammation. The same deterioration was not found in those consuming 10mg of gluten per day although there was some difference between the placebo and 10mg group which could not be confirmed by this statistical analysis, possibly because the group of patients was too small. 

His group concluded that a dose of 50mg per day of gluten for 3 months was enough to cause a significant change in the small intestinal mucosa. He also pointed out that the variability between patients was very high and that the regulations need to protect those who are most sensitive.

The amount of gluten that a patient is actually consuming depends not only on the so-called gluten free products that he or she is eating but on the amount of wheat substitutes consumed. He quoted data from a European study showing that there is variability of intake of wheat substitutes across Europe: in Italy 10% of the coeliac population eat more than 500g of wheat substitute while in other countries it is nearer 200g. 

So you have to take into consideration both the concentration of gluten in the wheat substitutes and the amount of wheat substitutes that the patient is eating every day. If you take an intake of 50mg of gluten per day as already being risky, and if you eat products that contain 200ppm of gluten you can reach this risk level with just the 250grams of wheat substitute per day; if the products contain 100ppm products the patient is eating 250-500g wheat substitute a day could be in a dangerous situation. 

For this reason, and to allow a margin of safety, Professor Catassi suggests that the 20ppm is the safest level, and is very glad that this is the level that has now been agreed by the Codex commission. However, he wanted to stress that he did not have anything against wheat starch, but only provided that the level was kept very low.

Finally he showed the preliminary results of a study looking at the consumption of wheat starch and gluten-free products across Europe and which suggests that gluten contamination rates in gluten-free products are very low, even though this does, of course effect the rate of gluten consumption through either voluntary or involuntary ingestion.

Professor Alessio Fasano

In its proposals to set a safe limit for gluten levels in foods for those suffering from coeliac disease the FDA is taking into consideration the ability of assays to detect gluten levels in foods and evidence based research from clinical trails and studies.

In a typical assessment, animal studies precede human clinical trials. However, in the case of celiac disease, there are no reliable animal models and for more than 30 years, millions of humans have been following a gluten-free diet based on a safety level of 20-100 part per million of gluten.

Therefore the usual safety levels applied to a drug (‘Abate the safety threshold by a factor of 10 or 100 to take into account the extrapolation of safety data from animal models to humans’) is not relevant as far as coeliac disease is concerned.

Moreover, as the accuracy and sensitivity of assays has improved, so it has been possible to measure lower and lower levels of gluten, from 300ppm to 200ppm and then to 100ppm. It is now possible to measure down to 5ppm – but this does not mean that safety levels should be determined by the sensitivity of the assay.

Evidence based research (see the trial mentioned by Dr Catassi above) suggests that while 50mg per day of gluten is harmful for coeliacs, 20mg is not.

Dr Fasano is concerned that setting a safety threshold below 20ppm will drive  manufacturers out of the market thus drastically reducing the availability of gluten-free food for coeliacs and not only worsening their quality of life but, most likely, worsening their compliance rates with the diet and thus their overall health.

As he points out, assays can vary by as much as 20% thus meaning that, if aiming for a maximum of 5ppm gluten a manufacturer would have no flexibility at all. Moreover, at such a very low threshold level, contamination from the surrounding environment (in shipping, in the retail outlet etc) could become relevant.

He is therefore anxious that the FDA should follow the lead of the Codex Commission and set the limit at 20ppm.

First published November 2011

Further Reading

Fast Facts: Coeliac Disease by Geoffrey Holmes and Carlo Catassi

Gluten Freedom by Alessio Fasano

Click here for more articles on the management of coeliac disease

Back to top