NOTICE: This blog is no longer being updated, so medical information may no longer be accurate.

“In Defense of 20 Parts Per Million” by Alessio Fasano, M.D.

 The following letter is reprinted with permission from The University of Maryland Center for Celiac Research.

“In Defense of 20 Parts Per Million”

A Letter to the FDA

From Alessio Fasano, M.D., and The University of Maryland Center for Celiac Research

In its efforts to establish a safer food environment, the U.S. Food and Drug Administration (FDA) has reopened the period for public comments on its 2007 proposal on defining gluten-free food and establishing a safe level of parts per million of gluten. The 2007 proposal included a proposed rule that defines gluten-free as less than 20 parts per million (ppm) of gluten.

In its recent research, the FDA has taken into consideration two methods for determining a safe gluten threshold for people suffering from celiac disease:  an approach based on the sensibility of assays to detect gluten levels and an approach using evidence-based research from clinical trials and analytical studies. The matter of setting these safety standards is a complex one that involves social, economic and cultural factors as well as health and medical implications. We attempt to address some of these factors below in our discussion of the FDA’s report.

Testing Toxicity of a Substance

In the typical method of establishing the toxicity of a substance (usually a drug), human clinical trials are preceded by animal studies. Since there are no reliable animal models for celiac disease, this cannot be the case for the assessment of a safe gluten threshold. However, we already have a vast “body of evidence” to support the safety of the gluten-free diet in current commercial gluten-free products.

For more than 30 years, millions of people worldwide have been following a gluten-free diet based on a safety level of 20 ppm or more (up to 100 ppm). The vast majority of those consumers have suffered no ill consequences.

Safety-Based Assessments Not the Best Measure

Keeping in mind that a vast number of people have safely followed a gluten-free diet and the fact that the threshold was not extrapolated from animal studies, the safety assessment approach proposed by the FDA (Note: 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) does not apply in this case.

To take it further, let me share a little bit of scientific history. When we first began to measure the ppm level of gluten in products, the assay measurement used was only sensitive to a 100 ppm threshold. As assay measurements became more sensitive, the level of “safe” as defined as gluten-free was lowered.

Currently, the assay measurement is sensitive to a threshold of 5 ppm. However, just because we can measure gluten to the level of 5 ppm (an incredibly miniscule amount), does not mean that we should determine safety levels based on the sensitivity of the assay. This is not the best use of our scientific methods.

Need for Evidence-Based Science

At the University of Maryland Center for Celiac Research, we support the use of evidence-based research to establish a safety level for ppms of gluten. Limits must be established through double-blind, randomized trials such as the one conducted by our center in 2007. The three-month trial showed that a daily intake of 10 mg of gluten (that, translated in ppms, would be equivalent to the daily ingestion of more than a pound of gluten free products containing 20 ppm of gluten!) for three months by adults with celiac disease caused no intestinal damage. However, the threshold of 50 mg was harmful to the majority of patients.

See How Much Gluten is Safe for Me for safe daily levels of gluten consumption.

The best use of science to measure safe levels of any substance is by collecting evidence through double-blind, randomized trials. When it comes to measuring safe levels of gluten, unfortunately there are few such studies. And, as far as I know, there are no evidence-based published studies that demonstrate toxicity with exposure to 20 ppm and safety with 5 ppm exposure.

Margin of Error of ELISA to Measure Gluten in Foodstuff (ELISA stands for Enzyme-linked Immunoassay)

Setting a safe gluten-free threshold below 20 ppm could result in a drastic reduction in the amount and availability of gluten-free products in the U.S. market. Since the assay variability (margin of error of the ELISA) measurement of gluten in safety-based assessments can range from 10 to 20 percent, extremely low thresholds (like 5 ppm) do not give manufacturers enough flexibility to produce good-tasting and safe products.

For instance, a batch of brownies tested today could measure 3 ppm. Keeping in mind the 10 to 20 percent margin of error inherent in assay measurement, that same batch tested tomorrow could measure 7 ppm. If the safety threshold was set at 5 ppm, a manufacturer could test the product as safe at 3 ppm, and it could be measured as unsafe at 7 ppm in a different test.

Under these restrictive limits, manufacturers would either discontinue gluten-free products or be forced to create much more expensive and much less palatable products, resulting in a drastic loss of selection and quality.

If a very low threshold (such as 5 ppm) is chosen, even minimal cross contamination from various sources in the surrounding environment can come into play. In after-the fact contamination, products can pick up extra contamination from the factory to shipping to market, thus raising the level above the restrictively low threshold.

Effects on Gluten-Free Global Village

A final, but very important implication in setting the safety threshold of ppm for gluten is the nature of our gluten-free “global village.” If U.S. manufacturers are forced to abide by an unnecessarily restrictive safety threshold, they will not be able to compete successfully in the gluten-free global marketplace. Conversely, overseas manufacturers with products that have a higher ppm level than the U.S. safety threshold will not be able to sell their products in the U.S.

Consequently, U.S. consumers of gluten-free products will find themselves in a conundrum while traveling overseas: they either bring their own gluten-free items or limit their diet while traveling outside the U.S. These developments would certainly worsen the quality of life for people with celiac disease and gluten sensitivity and deliver the opposite outcome that this law is intended to obtain for these patients.

We believe that establishing a restrictively low threshold of parts per million of gluten will complicate the lives of people with celiac disease and do nothing to improve their levels of safety and comfort. We urge the FDA to establish 20 ppm as the level that defines “gluten-free” in the U.S. marketplace.

Sincerely yours,
Alessio Fasano, M.D.
University of Maryland Center for Celiac Research

Fasano FDA from UM news on Vimeo.

Send Your Comments to the FDA Today

In 2003, Dr. Fasano and others from the CFCR created a task force that brought together patient support groups, researchers, clinicians and food manufacturers to advocate for changes in food labeling of products containing gluten.  The American Celiac Disease Alliance (ACDA) emerged from those efforts and is now the leading advocacy organization for celiac disease in the U.S.

If you’d like to join the efforts of the CFCR and the ACDA, please visit HERE to sign your name to the letter we will be sending to the FDA in October.

If you’d like to submit your comments to the FDA individually, please visit the FDA website for instructions.

You can stay up to date with The Center for Celiac Research on Facebook and Twitter.


  1. Since not everyone in the world uses Facebook, I want to be sure that your post on this gets as much attention as it deserves. For those of us who are celiac, the FDA's rules have tremendous importance. After reading Dr. Fasano's letter above, I signed up to join the efforts of the CFCR as mentioned above. Then, I posted my reaction to your question on your Facebook page: "While I would prefer that "gluten-free" meant exactly that — NO gluten — I do understand that it may be one of those things where close is close enough…maybe. Apparently some people are not as sensitive as others to small amounts of gluten. Dr. Fasano's letter mentions a 3-month study done in 2007 that indicated that 10ppm showed no intestinal damage in the subjects, while 50ppm did show some damage. That's fine as far as it went, but it was only 3 months. Furthermore, some people apparently do suffer symptoms with such small amounts, and some of them suffer a lot. Possibly it would work for manufacturers to mark what level their product is safe to — 20ppm, 10 or even 5. That way the consumer could make his own decision about what he thinks is necessary for his own health. I know, that might be confusing for some of the consumers to learn about, but…well, it's awfully confusing now, and I think that might be better. In my opinion, at least. Thanks, Heidi, for bringing this up here" …and on Facebook.

  2. Tara Marshall says

    20 ppm may work as a threshold-level for some people… but not for me. I react to anything higher than about 10 ppm, and the reaction is NOT mild. Small amounts of "contamination" like that give me neurological symptoms of extreme hyperactivity and OCD, which a friend has compared to me acting like I was on Meth. More than contamination gives me the reactions of vomiting, diarrhea, pain, nausea, and resultant dehydration that can keep me out of work for several days to a week.

    My reaction is so immediate that I have a friend who uses me to test products and restaurants for her son (I'm more sensitive than the testing kit she bought), who has almost the same food allergies and sensitivities that I have (he tolerate nuts, I do not, but I tolerate tomatoes and grapes, and he does not, but neither of us tolerate soy or corn). I've apparently saved them a lot of missed school days and sleepless nights.

    I already spend more than most people by buying products exclusively from gluten-free facilities, and this practice, apparently, will have to continue if the threshold is set at 20 ppm. I am very disappointed that this is considered a "safe" level.

    • I know all about being a celiac with very low tolerances 🙁

      I recently bought a bottle of Mug Rootbeer and after having one small sip, happened to look at the ingrediant list and saw "Modified Food Starch". I was sick for over a week with severe symptoms.

      Gluten free should mean just that – free from all glutens. If it were a peanut allergy they would not get away with 20 parts per million and this is why I pretty much have to avoid all the so called Gluten Free products out there because it is a joke. Why should I pay extra for the lable when it makes me just as sick as if I gone ahead and eaten a whole plate of spaghetti?

  3. Australia and NZ have laws limiting the use of the “gluten free” label to those containing less than 5ppm gluten.  Having shopped grocery stores in Canada, USA, and Australia in the last few years, I can say that there appear to be more products labeled gluten free under these more restrictive laws than in the US (where there are no laws) and Canada, where the new law allows up to 20ppm. 

    As mentioned in this article, there has been very few thorough tests of the minimum amount of gluten that cause a reaction in celiac sufferers, and none, that I’m aware of that included a large number people.  So why not, using the precautionary principle, start with a law that requires that only products with NO detectable gluten be able to be labeled gluten free, as it is in other countries.  If large scale definitive testing later show that 20ppm can be harmlessly tolerated by all celiac sufferers, then by all means change the law then.


  1. […] letter written by Dr. Fasano from the University of Maryland Center for Celiac Research titled “In defense of 20 Parts Per Million”.  Twenty parts per million translates in to twenty milligrams per kilogram of food.  He notes […]