Proposed FDA Standard for Gluten-Free Foods (20 ppm) May Not Adequately Protect the Food Supply for Celiacs
By: Peter Olins, PhD and Gillian Olins, PhD on August 18, 2011
The US Food & Drug Administration (FDA) has been developing regulations for the level of gluten in foods to be designated as “gluten-free” (Ref. 1), and opened the documents for public comment on August 3, 2011. The proposed “gluten-free” designation appears to be in sharp contrast to the FDA’s own Health Hazard Assessment (Ref. 2). Based on the scientific literature, we argue that the selection of a 20 ppm (parts per million) gluten limit does not create a sufficient margin of safety for people with a high degree of gluten-sensitivity. Instead, we propose a new, interim labeling of “Low-Gluten”.
- The data on the specific levels of gluten that can be tolerated are ambiguous and preliminary
- Given this uncertainty, a margin of safety should be applied because of the potential toxicity of gluten
- The FDA proposed regulation is based on limits of assay sensitivity, rather than safety
- “Gluten-free” will be mistakenly interpreted as “risk-free”
- Instead, an interim standard of “Low-Gluten” should be adopted, using current assay methods. This would address the needs of manufacturers, while protecting consumers
- Total amount of gluten per serving should be taken into consideration
- More sensitive and validated assays for gluten are urgently needed
- Further clinical research on safe levels of gluten is required
Here, we provide a more detailed analysis of the issues:—
1. Background on the mechanism of celiac disease and gluten-sensitivity
For those not familiar with this field, celiac disease is a chronic inflammatory disease that can take years to develop. It can affect both the small intestine and other organs, and varies in severity from a “silent” form to an “active” form with often severe reaction to consumption of even minute amounts of gluten in foods such as wheat, rye, and barley (Ref. 3). While approximately 30% of the general population have a genetic susceptibility, only a small fraction actually develop the disease (approximately 1% of the population). The environmental or genetic factors that protect most people from developing celiac disease are not understood.
Celiac disease is highly under-diagnosed, because of the frequent lack of clear symptoms. In addition to celiac disease, there is a substantial fraction of the population who have a moderate sensitivity to gluten, but do not go on to develop the disease.
Since there is such a wide spectrum of reaction to gluten, and because celiac disease develops over a period of years, it has been difficult to define a threshold, or “safe”, level of gluten consumption. A detailed review can be found in the May, 2011 FDA internal report on the health hazards of gluten exposure (Ref. 2).
Ref. 1: Food Labeling; Gluten-Free Labeling of Foods; Reopening the Comment Period. Federal Register, Vol. 76 No. 149.
Ref. 2: Health Hazard Assessment for Gluten Exposure in Individuals with Celiac Disease: Determination of Tolerable Daily Intake Levels and Levels of Concern. May 2011. pdf file available for free download.
Ref. 3: Risk of Morbidity in Contemporary Celiac Disease. Expert Rev Gastroenterol Hepatol 2010; 4:767-780. Lewis, NR & Holmes, GKT
2. The ultimate goal: define standards for foods and ingredients which will be safe for people who are highly sensitive to gluten
Unlike mercury or dioxin, which are considered to be toxic to people in general, gluten-containing food products are consumed by almost everyone—but are only toxic to a small fraction of people. Foods labeled “cholesterol free” or “sodium free” provide useful guidance to aid the dietary choices of many people. In contrast, since gluten consumption by a celiac can have a major impact on health, and even life expectancy (Refs. 3, 4), the validity of a “gluten-free” food label carries substantial weight and responsibility.
We believe that gluten should be considered a potentially toxic food component rather than a food ingredient, such as cholesterol or sodium. Since the purpose of a gluten-free designation is to prevent celiacs from food-induced disease, then a stringent standard should be set, with an acceptable margin of safety based on the uncertainty and variability of its toxic effect on different people.
Ref. 4: Increased prevalence and mortality in undiagnosed celiac disease. Gastroenterology. 2009 July; 137(1): 88–93. Rubio-Tapia, A et al.
3. Problems in the measurement of gluten
“Gluten” is not a single substance: it is an umbrella term used to describe a set of different proteins. This complexity is increased further since, for example, each wheat plant has multiple genomes, and many different varieties of wheat are grown around the world. This results in a complex mixture of proteins which differ from variety to variety (Ref. 5).
In addition, food processing affects both the physical and chemical structure of gluten, and the validity of the ELISA for processed foods has not been proven.
Unlike a single hazardous substance like mercury, the relative safety of the different components in a complex mixture of proteins in “gluten” has not been established. In an ideal world, we would be able to measure “gluten”, but at present, the best that science can provide is to use a test which will give us an indication of the important toxic components in a food, not an accurate measurement.
Ref. 5: Celiac disease: how complicated can it get? Immunogenetics 2010; 62:641-651. May-Tjon, J, et al.
4. Biological activity of different gluten components differ between patients
Numerous peptide fragments of the different proteins in “gluten” induce an immune response that is different for different people (Refs. 6, 7). While some peptides appear to be more reactive than others in stimulating an immune response, there is no single peptide fragment that has been proven to be representative of how gluten consumption provokes an immune reaction or leads to celiac disease. Indeed, the R5 ELISA assay commonly used for detection of gluten in food has never been validated as a predictor of disease severity (Ref. 8).
Ref. 6: Intestinal T Cell Responses to Gluten Peptides are Largely Herogeneous: Implications for Peptide-Based Therapy in Celiac Disease. J Immunol 2009, 4158-4166. Camarca, et al.
Ref. 7: The gluten response in children with celiac disease is directed towards multiple gliadin and gluten peptides. Gastroenterology 2002, 122:1279-37. Vader, W, et al.
Ref. 8: Gluten measurement and its relationship to food toxicity for celiac disease patients Plant Methods, 2008, 4, 26. Lester, DR
5. Different human populations differ in their sensitivity to gluten
The prevalence of celiac disease varies widely from country to country, resulting from a combination of genetic differences and environmental factors (Ref. 9). Within a group, individual celiac patients differ in their response to trace amounts of gluten (Ref. 10). It is therefore difficult to extrapolate from, say, the results from 12 subjects in a clinical trial, to the diverse population of 3 million people in the US with celiac disease.
Ref. 9: The prevalence of celiac disease in Europe: results of a centralized, international mass screening project Ann Med 2010; 42:587-95. Mustalahti, MK et al.
Ref. 10: A prospective, double-blind, placebo-controlled trial to establish a safe gluten threshold for patients with celiac disease. Am J Clin Nutr 2007; 85:160-166. Catassi, C et al.
6. Recovery from celiac disease
It is well known that some people, especially older adults, recover from celiac disease very slowly, or only partially (Ref. 11). It is still not clear if their intestines have suffered irreversible damage, or whether the trace levels of gluten present in the diet continue to stimulate the immune system. Since recovery can take years, celiacs need to err on the side of caution.
Ref. 11: Mucosal recovery in adults with celiac disease after treatment with a gluten free diet Am J Gastroenterol 2010; 105:1412-20. Rubio-Tapia, A, et al.
7. FDA health hazard assessment; margin of safety
The May, 2011 FDA internal report by the Office of Food Safety (Ref. 2) is a 93-page analysis of the available information on the clinical safety of gluten, in an attempt to define a safe level of gluten consumption. The FDA report is a revised version of a 2008 draft, and includes input from a December 2010 review by an external expert panel (Ref. 12). All the published studies involved small numbers of patients, using different study designs, making it impossible to draw a definitive conclusion about a safe threshold level of gluten consumption for celiacs.
(Unlike the drug and therapeutic sections of the FDA, where the agency can insist on rigorous clinical trials, the Office of Food Safety has to rely on existing published information.)
The authors developed different estimates of safety, including a Lowest Observable Adverse Effect Level and the Tolerable Daily Intake (TDI) (see Ref. 2, Appendix page B-14). A TDI of 0.4 mg gluten per day was derived using published data on measurable intestinal damage caused by gluten, and would correspond to the consumption of 400 g of food at a gluten level of 1 ppm.
The final sentence of the FDA Health Hazard Assessment report (Ref. 2, p. 46) states that:
In sum, these findings indicate that a less than 1 ppm level of gluten in foods is the level of exposure for individuals with CD on a GFD that protects the most sensitive individuals with CD and thus, also protects the most number of individuals with CD from experiencing any detrimental health effects from extended to long-term exposure to gluten.
It is important to note that a 1 ppm threshold would be beyond the capabilities of current assay, which creates a quandary for the FDA. The FDA has requested feedback on this proposal (comments should be submitted by October 3, 2011, see Ref. 1).
Ref. 12: External Peer Review of the FDA/CFSAN Draft Health Hazard Assessment for Gluten in Individuals with Celiac Disease: Determination of Tolerable Daily Intake Levels and Levels of Concern for Gluten pdf file available for free download
8. Examples of studies attempting to determine a safe dietary intake of gluten for celiacs
As explained earlier, there are remarkably few clinical studies which actually address the question of the specific dose of gluten that is safe for a celiac. Two publications were of particular concern to the FDA. One widely cited paper by Catassi et. al. (Ref. 10) was a small pilot study which attempted to define a safe intake of gluten. Thirteen celiac patients were tested with 50 mg gluten per day and showed a statistically significant level of mucosal damage compared to placebo. At the lower dose of 10 mg per day (which would correspond to 400 g food containing 25 ppm gluten), no statistically significant effects were observed. However, the authors noted:
The gluten microchallenge disclosed large interpatient variability in the sensitivity to gluten traces. Some CD patients showed a clear-cut worsening of the small-intestinal architecture after ingesting only 10 mg gluten/d[ay]…
and cautioned that:
Because of the limited number of patients, we were not able to reach firm conclusions about the potential toxicity of 10 mg gluten/d[ay], which remained a “gray” area.
Another paper by Chartrand et. al. (Ref. 13) showed that 11 out of 17 celiac patients developed symptoms while consuming processed wheat starch containing 7.5 ppm gluten. They concluded that:
The innocuousness of long-term ingestion of “gluten-free” products containing wheat starch is still unproven…
Ref. 13: Wheat starch intolerance in patients with celiac disease. J Am Diet Assoc 1997; 97:612-618. Chartrand, LJ et al.
9. Proposed FDA regulation disregards the available safety data
The proposed FDA regulation focuses on the current limit of detection of “gluten” (approximately 20 ppm using current assays). The FDA argues that the Hazard Assessment values that it developed are too uncertain to be taken into consideration (Ref. 1). In our opinion, this would be analogous to taking a known toxic chemical, and setting a high threshold for food contamination just because the available assay cannot detect lower levels of the toxin. Would the general public want to consume such a food with a label of “toxin-free”? We think not.
10. Risk of setting a “gluten-safe” threshold too high
- What is an acceptable level of risk? We all face countless risks every day, and modify our behavior accordingly. Obviously, there is no way of guaranteeing that a food is safe for everyone.
- For some unseen risks, such as toxins in the food supply, we rely on our government to provide guidance. If the FDA says that a food is “free” of a potential toxin, it sends a strong message to the consumer that the product is “safe”. Unfortunately, such a definitive position is not supported by the FDA’s own report.
- The damage caused by gluten is at the microscopic level and often does not manifest as obvious symptoms, so we also need to be concerned about protecting celiacs from undetected damage. A margin of safety is therefore required to protect the public.
- A margin of safety is also required, given the lack of validation that the current gluten assay is a predictor of biological activity.
- We believe that a designation of “gluten-free” at 20 ppm will lull some people into a false sense of security, potentially resulting in an increased health risk.
- In our opinion, a designation of Low-Gluten would be much more meaningful: it implies that there is a certain level of a toxin present, and that there is a potential risk to those most sensitive to the toxin.
11. Social and business impact
While many have protested the delay in developing a final FDA regulation for “gluten-free” food designation, given the complexity of the issues involved, the FDA has not been dragging its feet, but has been justifiably cautious. Having no standard at all is better than having a poor one that potentially could cause harm. We think that this caution and delay has actually been a good thing.
Even though celiac consumers want to have their food products clearly labelled “gluten-free”, this may not be a reasonable or achievable goal currently, or in the near future.
There is currently a boom in the consumption of “gluten-free” processed foods in the US, probably fueled by the (unproven) use of gluten-free products as a weight-loss aid. However, the increased availability of “gluten-free” foods is excellent news for celiacs, and food manufacturers are eager to have a more unified labeling standard, both to guide their manufacturing processes and to aid their marketing. We suspect that a designation of “low gluten” might have less marketing impact than “gluten-free”, but such a compromise would provide evidence for the manufacturers’ good faith intent to protect the consumer.
12. Gluten-free concentration standard versus dose per serving
One secondary point that hasn’t been discussed is that toxicity is proportional to the amount of gluten consumed, not the concentration. In other words, while a single threshold of 20 ppm may be convenient for manufacturers, the consumer is more concerned about amount per serving. The public has become comfortable with reading food labels and % Daily Value, so it should not be hard for the FDA to implement a similar system for gluten content. Since some foods are consumed in small amounts, this might have the added benefit of increasing the range of processed foods that can be available to the consumer on a low-gluten diet.
13. Our proposal
The FDA should abandon the hope of defining a true “gluten-free” standard (implying zero gluten) for now, but use the existing assays to define a Low-Gluten standard that will help the majority of people who are gluten-sensitive, while also providing some guidance to those who are extremely sensitive. Such a compromise would provide an interim solution, until a more rigorous definition of “gluten-free” can be obtained, and more extensive clinical trials are undertaken.
- The FDA has developed an internal report which recommends a “Tolerable Daily Intake” as 0.4 mg “gluten”, which corresponds to a gluten concentration of 1 ppm.
- This is in sharp contrast to the proposed 20 ppm FDA standard for labeling a food as “gluten-free”.
- The FDA has chosen to define “gluten-free” with respect to the current limits of detection, rather than on the basis of safety.
- There is a substantial degree of uncertainty as to whether the current ELISA test for “gluten” is an accurate measure of gluten’s clinical effects in different individuals.
- The current state of clinical science is insufficient to confidently define what is biologically gluten-free.
- FDA standards should create a margin of safety, and err on the side of caution.
- Setting a lax standard of “gluten-free” may actually pose a greater health risk than no standard at all.
- FDA should instead create an interim designation of “Low-Gluten”, using currently available assays. This would be valuable today to people with mild gluten sensitivity, while balancing the needs of manufacturers.
- “Low Gluten” foods may still be unsafe for many highly sensitive celiacs.
- FDA regulations should be geared to amount per serving rather than a simple ppm concentration.
- We recommend that the celiac community argue against the FDA’s proposal as it stands, but support a “Low-Gluten” labeling, instead.
- More sensitive, validated assays need to be developed in order to support a future labeling standard of “gluten-free” that is based on predicted health risk
- Larger clinical trials of celiac patients are required in order to establish the threshold of gluten sensitivity, and the variation from person to person.
We realize that this opinion may be controversial, but we believe that it would be better for the FDA to adopt no labeling standard for “gluten-free” foods than a poorly-designed one. Please help develop the public dialog by using the Comment Section below. (We can also be reached privately).
If you choose to give your feedback to the FDA, we encourage you to read the relevant reports first. Instructions for public comment can be found at: