Our New Cookbook!

Gluten Free, Deliciously! Party Food Recipes

Our Blogs

Assay for Gluten in Barley is Flawed—Implications for Gluten-Free Labeling of Foods and Gluten-Free Beer

By: Peter Olins, PhD on July 27, 2013.

In case you saw my recent article on gluten-free beer (Ref. 1), about whether Omission Beer has succeeded in removing gluten from their beer, let me say that I’m not obsessed with this topic! However, I just came across two very important research reports from a group of laboratories in Australia, both focusing on the challenges of measuring the gluten content in barley (Ref. 2, 3). Their results extend far beyond beer, because they affect how we measure levels of barley contamination in gluten-free foods. This is vital information for people with celiac disease. The studies are also critically relevant to brewers who start with barley malt, and then treat the beer with an enzyme preparation that degrades gluten.

Take-Home Messages From This Article:

  • The assay that is commonly used for measuring gluten in foods and beverages relies on an antibody called “R5″. However, this antibody did not give a consistent result with eight different varieties of barley, meaning that some assays for gluten could be highly suspect. In fact, hordein proteins from some varieties of barley can give readings that differ by 1000-fold!
  • This means that current tests could underestimate the level of barley contamination, but still meet the common standard of “less than 20 ppm gluten”.
  • In my opinion, based on these new results, gluten tests based on the “R5″ antibody are not a valid way to predict the potential toxicity of reduced-gluten beers in celiacs.

The “R5 Sandwich ELISA” for Measuring Gluten in Food

What’s the difference between gluten and hordein?

In this article, I will use the simple umbrella term “gluten” to describe all the different storage proteins in the wheat/barley/rye family of grains, since wheat gluten is what most people are familiar with. But to be precise, the corresponding proteins in barley are actually called “hordeins” and “barley glutelins”. The term  “prolamin” is a generic term and includes hordein.

Worldwide, the most common assay used for measuring gluten contamination in food is the “R5 Sandwich ELISA”, which has been standardized using gluten from wheat. It also detects related proteins from rye and barley, and it has been assumed that this single assay can be used to estimate gluten in all three grains. However, the two new Australian studies (Ref. 2, 3) show that the R5 antibody is very poor at detecting “gluten” derived from certain varieties of barley. This means that “gluten” values could be as much as 1000-fold off.

This observation is relevant to how barley contamination levels are measured, in order to comply with the common cutoff requiring gluten-free foods to have less than 20 ppm “gluten”.

The “R5″ Antibody is Very Poor at Detecting Gluten (Hordein) in Certain Varieties of Barley

R5 ELISA for Detecting Gluten in Barley

R5 ELISA for Detecting Gluten in Eight Varieties of Barley

Don’t be intimidated by the diagram! It’s quite easy to understand. It’s simply a graph of how a particular variety of barley produces a detectable signal for “gluten” content (A450). In an ideal assay, all the lines for all 8 varieties of barley would overlap with the black line in the middle line (which is wheat gliadin). Ideally one should get exactly the same value, regardless of which variety of barley was tested. However, as you can see, there is a huge variation between these results. In the worst case, two varieties of barley hordein proteins that should give an identical signal actually vary by at least 1000-fold!

In some cases, substantial contamination levels of certain barley varieties would give a low signal, but could be legally labeled as “gluten-free” in many countries of the world. In other cases, levels of barley contamination could be an overestimate.

Implications for Measuring Gluten (Hordein) in Beer Made From Barley Malt

Over the past few years, several breweries have made considerable progress in developing beers that are based on a variety of different grains and seeds (e.g. rice, millet, sorghum, quinoa, buckwheat). These are gluten-free because they start with gluten-free ingredients (Ref. 4). So, brewers of beer that rely on ingredients other than wheat or barley should have no concern.

The problems with using the R5 antibody for detecting gluten are extremely relevant to the brewers who start with barley malt, but then treat the beer in an attempt to reduce the “gluten” content. Currently, there are at least five beers which use this basic concept:

Omission Beer (Widmer Brothers)
Estrella-Damm Daura (Estrella Damm)
Ambre Ale and Blonde Ale (Brunehaut)
Prairie Path - Golden Ale (Two Bros. Brewing Co.)

This is also relevant for brewers who start with a supposedly “low-gluten” variety of barley malt to produce their gluten-free beer, since the level of gluten may be an underestimate. Brewers who treat their barley-based beers to remove residual gluten (hordein) will now have a greater challenge to satisfy the public that their products are safe.

Recently, a new assay based on the R5 antibody has been developed (called the “R5 Competitive ELISA”), which is more sensitive to the gluten (hordein) present in beer (Ref. 1). However, this has not been validated by the regulatory authorities. My main problem with this approach is that it still relies on the same R5 antibody, which we now know is very poor at detecting gluten in certain barley varieties.

Of course, the key issue is that all these assays are really “surrogates”, rather than telling us what we really want to know, which is “Is this food safe for celiacs?”.

What’s the Solution?

It is still early days, and there is no obvious way to address these issues, but one approach might be to use a completely different assay technique, called “LC-MS” (Ref. 3). However, this approach is still in the research phase, so it will probably be a long time before we have enough data to conclude that we have a better alternative than the current ELISA methods.

The regulatory agencies responsible for monitoring the levels of gluten in foods are, I’m sure, following this topic closely.

Closing Comments on Detection of Barley in Gluten-Free Foods

A good food safety assay needs to be both accurate and relevant. In the case of barley, I would argue that the currently-used assays are not accurate, and their relevance has not been demonstrated.


Ref. 1: Gluten-Free Beer: Does Omission Beer Deliver the Goods? — A Simple Guide for the Non-Biochemist

Ref. 2: Tanner, GJ, et al. Quantification of Hordeins by ELISA: The Correct Standard Makes a Magnitude of Difference  http://www.plosone.org/article/fetchObject.action?uri=info%3Adoi%2F10.1371%2Fjournal.pone.0056456&representation=PDF

Ref. 3: Tanner, GJ, et al. Measuring Hordein (Gluten) in Beer – A Comparison of ELISA and Mass Spectrometry  http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0056452

Ref. 4: Is Gluten-Free Beer Made From Barley Malt Safe for Celiacs?

12 comments to Assay for Gluten in Barley is Flawed—Implications for Gluten-Free Labeling of Foods and Gluten-Free Beer

  • Lew Black

    Thanks for this valuable information. I have recently been diagnosed and have cut gluten from my diet. I used EZ gluten strips to test St. Pauli Girl and Warsteiner beers and they easily passed. Back to Bards I guess… I did email EZ gluten and gave them a link to this article. I will be curious to see how they respond. Lew Black

    • Peter

      Hi Lew,
      Thanks for the information.
      I don’t know enough about the details of this product to comment fully. However, your observation is consistent with the case that I have been making, that current antibody-based assays are a poor tool for monitoring the safety of fermented or hydrolyzed products that are based on barley malt. This consumer diagnostic product may be great for normal foods, but may be unsuited for the beers that you mention.

      Let us know if you hear back from EZ gluten.

  • Lew Black

    I heard back from Laura Allred at EZ Gluten. She is implying that it is filtration in commercial breweries that removes the gluten, though i don’t know why this works with German lagers, but not most commercial American lagers. This is her response: I am impressed by how quick and thorough she was:

    Dear Mr. Black,

    The EZ Gluten test is not based on the R5 antibody, and is a different test method (lateral flow vs 96-well plate) than the sandwich ELISA.

    We have just completed a large validation of the EZ Gluten and our quantitative method (AllerTek Gluten) at various stages of beer production, and both tests performed quite well for detecting gluten in traditionally brewed beer. Contrary to speculation, the processes of malting, kilning and fermentation do not break barley proteins down to a level where they are undetectable, and if we spiked low levels of barley malt into a sorghum based wort, we could continue to detect them through the production of beer and into the finished product. Only very fine filtration would remove these proteins from the finished beer, and commercial breweries often use this type of filtration. We had these results confirmed by LC/MS to show that there were not remaining small fragments that were missed by our antibodies.

    But products like Omission beer are a very different story. These are beers brewed entirely from barley that then use a digestion process to hydrolyze remaining barley proteins. Their process does seem able to break proteins down to the point where they are no longer recognized by the antibodies in test kits. Since the disease mechanism in celiac disease is also antibody mediated, the argument could be that the proteins are also too small for the antibodies produced by someone with celiac disease to recognize. But that’s where you are asking people to take a leap of faith.

    So I would say that for your purpose, testing traditional beers or gluten-free beers with the EZ Gluten should be sufficient. I can’t, however, recommend it for products like Omission without further study.

    I hope that helps. Let us know if you have any other questions.


    I will read your source articles to see if there is any reference to the “lateral flow vs 96-well plate” test. Does it mean anything to you?

    • Peter

      Hi Lew,
      “Lateral Flow” devices typically have a test strip which changes color if a substance is detected. These typically use an antibody in a “sandwich” configuration, as I described my previous blog on beer.

      If I am not mistaken, the EZ Gluten strips use the “Skerrit” antibody, which is not very sensitive in detecting barley, but I am surprised that you got a negative signal with conventional beer.

  • Lew Black

    Peter, I had read on another website that German beers, as well as Heineken, were essentially gluten free, so I got some EZ Gluten test strips and tested two German beers as well as Batch 19, a micro brew from Coors. The Batch 19 tested positive while the St. Pauli and Warsteiner were both negative.

    You are correct that EZ Gluten is a test strip that changes color. It claims sensitivity of 10ppm. You are also correct that it uses the Skerrit and Hill antibody.

    Laura states that “We have just completed a large validation of the EZ Gluten and our quantitative method (AllerTek Gluten) at various stages of beer production, and both tests performed quite well for detecting gluten in traditionally brewed beer..We had these results confirmed by LC/MS to show that there were not remaining small fragments that were missed by our antibodies.” Note that they spiked sorghum based wort with low levels of barley malt and were able to accurately detect it in the finished product. From your article above I gather that they need to test many varieties of barley in the same way to be sure that they get the same results. Is that a correct interpretation?

    If she is correct that the filtering of the beer German beers is the reason for their clean results, than the type of barley wouldn’t matter. I may buy some more German beers by the bottle and do more testing… I would sure like to be able to test beers accurately so I can have alternatives to Bards and the like..

    • Peter

      Hi Lew,
      It’s hard to comment on the statement from EZ Gluten without seeing the actual data. However, since their tests rely on a “sandwich” antibody configuration, I don’t think it will ever be appropriate for detecting low levels of “gluten” in barley-based beverages. The FDA specifically excluded hydrolyzed foods from its regulation precisely because of this problem.

      Regarding filtration, it seems unlikely to me that a commercial process could efficiently filter out small fragments of proteins.

      Did you read my previous blog? It describes in detail why it is impossible to get accurate test results for “hydrolyzed” foods such as barley-based beers:

  • Lew Black

    Peter, I had not read your follow up article that you linked to. That helped, as my last chemistry class was in 1965! I will give up on my plan to use EZ gluten to test beers and stick with the non-barley based beers. Thanks for the follow up. Lew

  • Is Omission Beer Gluten Free? | Gluten Free: The Celiac Site

    […] allegedly gluten free beer. I will not attempt to summarize them for you. Click here and here. This article deals with the specific situation involving Omission […]

  • Steffen

    Hi all, I have just started a new website about the gluten level in beer. So far I tested GlutenTox Home which uses the G12 antibody and Gluten-in-Food Kit from Imutest.

    If you have made experience with EZ gluten I would be grateful if you could contribute with your results - no matter if Skerrit antibody is useful or not.

    In my tests Corona was tested gluten-free and Heineken has below 40 ppm. Please find some pictures and the test result here: http://www.lowgluten.org

    As I am from Germany I will test a number of German beers obviously. I could test Warsteiner using GlutenTox for instance to double check with your EZ Gluten result.

    Regards, Steffen

  • Steffen

    Hi Peter,

    Thanks for your feedback. The site is definitely determined for people with CD or who are sensitive to gluten like me. But I don’t have a scientific approach as I simply don’t have the expertise.

    People with CD should just not drink regular beer but for those who are “just” sensitive the level of gluten in beer can be important. For instance I tested my favourite beer above 100 ppm so I consider changing it. Such tests can lower the risks for some people but I do not declare to carry out scientific analysis.

    Reading you articles I realised that the detection of gluten in beer is very complex and you don’t consider these methods reliable. Also the FDA statement is very interesting but again I am not a scientist. I just started with my website and I am in the learning process so the information you share are actually very useful – even when it questions the whole concept of my site.

    What is your opinion on the G12 antibody compared to R5?



    • Peter

      Hi Steffen,

      Thanks for your comments. I fully agree that people will differ in their degree of sensitivity to gluten. However, for celiacs, it is well known the lack of physical symptoms is a poor indication of whether there is actual damage taking place to the lining of the small intestine. Many other effects of untreated celiac disease occur without symptoms at all, such as anemia and osteopenia. Even after five years of an apparently gluten-free diet, about one third of celiacs still have intestinal damage. While there will be a number of reasons for this, it seems likely that low level gluten contamination of food is probably a contributor to this:
      Therefore, in my opinion, based on current information, celiacs who drink beer that is based on barley or wheat malt are taking an unnecessary risk.

      Regarding the use of different antibodies for detecting gluten, I don’t think there is much known about how well the G12 antibody can detect gluten in beer. However, I would expect that any of the currently available antibodies would probably suffer from the same detection problems that I described in the article that I mentioned earlier:

      [As an aside, I would appreciate any feedback you could offer about whether I was successful in explaining these complex issues to someone who is not a biochemist!]

      Good luck with your website.

      Regards, Peter.

Leave a Reply