Gluten-Free and Nutritious Too!

This alert is available on the Gluten Free Watchdog site at http://www.glutenfreewatchdog.org/blog.php?id=7

Thanks so much for your support!

Hopefully many of you submitted comments to FDA regarding gluten-free rulemaking. The comment period has now closed.

I am posting both the letter FDALetter and product testing information FDAAttachment2SummaryFindingsGFWRN from Gluten Free Watchdog, LLC (www.glutenfreewatchdog.org) recently submitted to FDA (manufacturer names redacted).

Please see http://www.glutenfreewatchdog.org/blog.php?id=7

Hopefully everyone has taken the time to read the Federal Register Notice. If not it is available at http://www.gpo.gov/fdsys/pkg/FR-2011-08-03/pdf/2011-19620.pdf. Here are a few things to consider: 

• In 2007, the FDA released a proposed rule for labeling of foods as gluten free. The preamble to this rule listed two methods that could be used to establish a gluten threshold level—an analytical methods-based approach and a safety assessment-based approach.
• Under the analytical methods-based approach the gluten threshold would be based on the sensitivity of the validated tests used to assess foods for gluten content (the sensitivity of an assay refers to the level at which gluten can be accurately detected or quantified). Under the safety assessment-based approach the gluten threshold would be based on a daily intake of gluten considered “safe” for the vast majority of individuals with celiac disease.
• The safety assessment found the tolerable daily intake (TDI) for gluten to be 0.4 milligrams per day for adverse effects on the intestinal mucosa and 0.015 milligrams per day for adverse effects on clinical symptoms.
• If these TDI numbers are accurate it means that a person with celiac disease would have to limit themselves to less than a 1 ounce slice of bread containing 20 parts per million* of gluten each day to prevent intestinal damage; far less to prevent clinical effects.

*20 parts per million means 20 milligrams per kilogram of food

• Regarding use of the safety assessment-base approach, The Federal Register states the following, “FDA tentatively concludes that, based on the LOCs (level of concern)* identified in the safety assessment-based approach, the Agency should not use that approach in defining ‘‘gluten-free’’ because the estimation of risk to individuals with celiac disease associated with very low levels of gluten exposure may be conservative and highly uncertain.“

*For more information on LOCs, please see the Health Hazard Assessment for Gluten Exposure in Individuals with Celiac Disease available at: http://tinyurl.com/3b22eto

• In the Federal Register Notice the FDA tentative proposes to use an analytical methods-based approach to adopt a gluten threshold level of <20 parts per million as one of the criteria for labeling food gluten free.
• It is further stated in the Federal Register Notice that, “At the current time, FDA is not aware of any analytical methods that have been validated to reliably and consistently detect gluten below 20 ppm.”

My thoughts

You are not alone if you are confused by the disconnect between the findings of the safety assessment-based approach and the analytical methods-based approach. I have wondered how the FDA was going to reconcile the findings of the safety assessment with the level of gluten listed in the proposed rule (< 20 ppm) for the past three years—ever since I reviewed the safety assessment in 2008. If you would like to read blinded reviewer comments from the External Peer Review of the FDA/CFSAN the Draft Health Hazard Assessment for Gluten Exposure in Individuals with Celiac Disease, please see http://tinyurl.com/3g9exay. Reviewers include David Bellinger, PhD; Donald Kasarda, PhD; Michael Luster, PhD; Joseph Murray, MD; Tricia Thompson, MS, RD; Nga Tran, DrPH.

As I wrote in my review comments at that time, “Many of the TDI (tolerable daily intake) estimates are ridiculously low and do not make sense. The TDI for subchronic gluten exposure for morphological effects is listed as 0.4 mg/d. The TDI for acute, subchronic, and chronic clinical effects is listed as 0.015 mg/d. The FDA’s proposed rule on gluten-free labeling would allow a food to be labeled gluten free if contained less than 20 parts per million of gluten. Codex allows foods to be labeled gluten-free if they contain no more than 20 parts per million of gluten. If a person with CD ate just one ounce of a product containing 20 parts per million of gluten, they would be ingesting 0.57 mg of gluten (greater than the TDI values listed above). If these TDI values are disseminated to the public, persons with celiac disease will be afraid to eat.”

Minds much smarter than mine conducted the safety assessment. While I can point out issues that are confusing or do not make sense, I do not pretend to be an expert on safety assessment methodology. From my perspective, what would be most helpful is clarification from FDA on how they got from the findings of the safety assessment to what is written in the Federal Register notice.

For example, it would be helpful to know how FDA feels about the tolerable daily intake values in the safety assessment. Like the LOCs, do they also believe the TDI levels to be “conservative and highly uncertain?”  If so, it would be very helpful for them to explain why to help calm fears among individuals with celiac disease.  

If, on the other hand, FDA believes the tolerable daily intake values to be accurate, it would be helpful for them to explain why they are continuing to propose a threshold level of gluten of < 20 ppm. If they truly believe that gluten can not be reliably and consistently detected at levels below 20 ppm* then they should not allow certifying agencies to have lower standards of 10 ppm and 5 ppm.

*It is worth noting that when I discussed the level at which the R5 ELISA had been validated with Dr. Enrique Mendez (developer of the R5 ELISA) he referred me to the sensitivity levels. The sandwich R5 ELISA has a limit of detection for gluten of 3 ppm and a limit of quantification for gluten of 5 ppm. 

If they believe lowering the standard to 15, 10, or 5 ppm will prevent industry from being able to manufacture gluten-free food, this belief appears to be unfounded. Based on testing done by Gluten Free Watchdog (www.glutenfreewatchdog.org), the majority of food is testing below 5 ppm gluten* and most of the products testing this low do not carry any specialty certification.

*As stated above, 5 ppm is the lower limit of quantification for gluten. 3 ppm is the lower limit of detection. There are no validated tests that test to zero.

It is the case that the methodology utilized in the safety assessment is generally used to assess the safety of substances that are toxic to all individuals not just certain individuals with a given medical condition. For example lead is toxic to all individuals; gluten is not. As stated in the Federal Register Notice, “FDA’s assessment of the adverse health effects of gluten exposure in individuals with celiac disease presented in the Gluten Report followed established hazard assessment components used within the Center for Food Safety and Applied Nutrition to determine TDIs for chemical and natural toxin contaminants in food.”

As a result, it may simply be the case that the safety assessment as currently conducted is not the best approach for determining safe levels of gluten for the approximately 1% of the US population with celiac disease. 

Ideally, the upper limit of gluten allowed in foods labeled gluten free would be based on what is considered safe for the majority of individuals with celiac disease. However, one of the more important points to stress is just because a labeled gluten-free food is allowed to contain up to (but not including) 20 ppm gluten does not mean it contains that much.  

What I would like to see

• It would be extremely helpful if the FDA would develop a Q & A or consumer/healthcare practitioner bulletin to address issues surrounding the safety assessment.
• It also would be helpful to have this available before the comment period is closed (or to perhaps extend the comment period).

The following is from http://tinyurl.com/3ok27gd

The FDA encourages members of the food industry, state and local governments, consumers, and other interested parties to offer comments and suggestions about gluten-free labeling in docket number FDA-2005-N-0404 at www.regulations.gov¹. The docket will officially open for comments after noon on Aug 3, 2011 and will remain open for 60 days.

To submit your comments electronically to the docket go to www.regulations.gov²

1. Choose “Submit a Comment” from the top task bar
2. Enter the docket number FDA-2005-N-0404 in the “Keyword” space
3. Select “Search”

To submit your comments to the docket by mail, use the following address:

The Division of Dockets Management
HFA-305
Food and Drug Administration
5630 Fishers Lane, Room 1061
Rockville, MD 20852

Include docket number FDA-2005-N-0404 on each page of your written comments.

Individuals with celiac disease have long worried about ingredients that are sometimes (rarely in the US) made from wheat starch, including wheat-based caramel color, wheat-based glucose syrup, wheat-based maltodextrin, and wheat-based dextrin. Under proposed FDA regulations for labeling food gluten free, wheat starch and ingredients made from wheat starch can be included in labeled gluten-free foods as long as the final food product contains less than 20 parts per million of gluten.

But what about foods not labeled gluten free that contain these ingredients? For example, if wheat-based maltodextrin is the only “suspect” ingredient in a food product not labeled gluten free is it okay to eat? While many (and I know not all!) dietitians and people following gluten-free diets have moved the ingredients caramel, glucose syrup, and maltodextrin into the “safe” column regardless of what they are made from, wheat-based dextrin (in foods not labeled gluten free) remains in the “unsafe” category. The question is whether this is necessary.

Background

Wheat starch is not wheat grain and it is not wheat protein. It is not intended to contain any gluten. Nonetheless it is difficult to completely separate the starch and protein components of wheat and small amounts of gluten remain.

The European Food Safety Authority (EFSA) has investigated wheat starch hydrolysates (e.g., wheat-based glucose syrup, wheat-based maltodextrin) as part of its allergen labeling program. According to the EFSA, testing done on wheat starch using the R5 ELISA found gluten in amounts up to 279 ppm. Wheat starch hydrolysates undergo many purification steps designed to remove protein. While wheat starch hydrolysates were found to contain intact gliadin and gluten peptides, the amounts were very low (< 5 ppm gluten for maltodextrin and up to 25 ppm in the glucose/dextrose samples).

In 2007, As a result of these findings along with a commitment from the starch industry that wheat starch hydrolysates would be purified to a maximum level of 20 parts per million of gluten, these ingredients were given permanent exemption from allergen labeling in the European Union (this is not the case in the US—in the US, under FALCPA, if any ingredient in a packaged food product regulated by the FDA contains wheat protein, the word “wheat” must be included in the ingredients list or Contains statement). Wheat starch and modified wheat starch were not exempted from allergen labeling in the European Union.

Okay, So What about Dextrin?

Dextrin also may be derived from wheat starch and if so would be considered a wheat starch hydrolysate. As many of you know, Benefiber sold in the United States uses wheat dextrin as its fiber source. Nonetheless, the product is labeled gluten free and product packaging states that it is tested to below 20 parts per million of gluten. This is the only information on the gluten content of dextrin I have been able to find. This ingredient does not appear to have been tested by EFSA.

Based on a subscriber request, Benefiber was recently tested by Gluten Free Watchdog, LLC (www.glutenfreewatchdog.org). The product was analyzed for gluten using both the sandwich and competitive R5 ELISAs. The competitive ELISA was used to detect any gluten peptide fragments in the product that might not be measured using the sandwich ELISA. Results of this testing indicate that wheat dextrin, especially when it is one of many ingredients in a food product, would be exceedingly unlikely to cause an otherwise gluten-free product to contain 20 parts per million or more gluten. Granted this is limited testing but it was done on a product containing one ingredient; 100% wheat dextrin.

Just a little food for thought…

Copyright © Tricia Thompson, Gluten Free Dietitan, July 2011. All Rights Reserved.

If you would like more information on testing done by Gluten Free Watchdog please read the FAQ at http://www.glutenfreewatchdog.org/faq.php

For more information on EFSA and maltodextrin, please see:

http://www.efsa.europa.eu/fr/scdocs/doc/487.pdf

For more information on EFSA and glucose syrup, please see:

http://tinyurl.com/3jzbb5m

For more information on the use of wheat starch hydrolysates in gluten-free foods in the EU, please see:

http://www.aaf-eu.org/pdf/AAF_02-2009_Communication_on_EU_allergen_labelling.pdf

There are now four gluten-free certification organizations in North America, namely programs run by the Celiac Sprue Association, the Gluten Intolerance Group, the National Foundation for Celiac Awareness, and the Canadian Celiac Association.  Click on the link to view a chart comparing these programs. The chart was reviewed for accuracy by each of the certifying organizations. The various seals and marks included below are from left to right: NFCA, CCA, GIG, CSA

BLOGGlutenFreeCertificationsTABLEJuly14             

Gluten Free Watchdog, LLC has tested a cross section of gluten-free foods since its debut on May 16. Flours, mixes, hot cereals, ready-to-eat cereals, cookies, snack foods, breads, pastas, soups, and tortillas from Arrowhead Mills, Authentic Foods, Bob’s Red Mill, Eco-Planet, General Mills, Holly’s Oatmeal, Lundberg Family Farms, Mary’s Gone Crackers, Namaste Foods, Orgran, Pamela’s Products, Quinoa Corporation, Thai Kitchen, Think Thin, and Trader Joe’s have been tested. These products have been purchased in triplicate from grocery and natural foods stores, through Amazon, and directly from manufacturers. All products have been sent unopened to Bia Diagnostics in Burlington Vermont. Samples have been tested in duplicate (two extractions) using the standard sandwich R5 ELISA (R7001 Ridascreen Gliadin) with cocktail extraction.

As of the end of June test results for seventeen products have been posted. All products tested were labeled gluten free. Two products also carried the Certified Gluten Free mark from the Gluten Free Certification Organization. Sixteen products tested below 20 parts per million of gluten, including 15 that tested below 5 parts per million. One product tested above 20 parts per million and also carried the Certified Gluten Free mark.

For more information, please see www.glutenfreewatchdog.org. Please note that complete test results are available to subscribers only.

Thank you to Thomas Grace of Bia Diagnostics, www.biadiagnostics.com for sharing his expertise on Lateral Flow Devices

What are lateral flow devices (LFDs)?

LFDs are used to qualitatively (is gluten present) or semi-quantitatively (what amount of gluten is present) determine whether gluten is present in a food product. There are several LFDs on the market for testing food for gluten, including the Ridascreen Quick Gliadin manufactured by R-biopharm, EZ Gluten manufactured by Elisa Technologies, AgraStrip Gluten G12 Test Kit manufactured by Romer Labs, and Gluten Tox manufactured by Biomedal Diagnostics. Specifics on each of these tests are included at the end of the Q & A section.

How do LFDs work?

LFDs are usually what we think of as “dipstick” tests. Most employ sandwich type methodologies. They utilize a line of fixed antibody on a surface strip and a second antibody conjugated with colored “nano” size particles (on the sample addition end of the strip). When a liquid sample extract is applied to the strip, the conjugate and the sample start to migrate across the surface of the strip together. If the sample extract has the protein or compound present (gluten) and the conjugate can recognize its epitope (binding site), under the right conditions they will bind together. Now that they are “hooked” together as they come in contact with the line of antibodies that are fixed to the strip, these antibodies will also bind to the protein forming a sandwich complex, “sandwiching” the protein (gluten) between the two antibodies. As the conjugate complex starts to accumulate on the surface of the strip the “nano” particles start to become visible. The more protein (gluten) in the sample the more antibody-protein-conjugate binding will take place and the darker the line becomes.

It should be noted that unless the LFD has been shown to reliably detect pure wheat, barley or rye cultivars (e.g. undiluted wheat flour), it should have a hook line to avoid false negative results with high concentrations of gluten.  A hook line is simply a small amount of analyte protein on the surface of the strip that will capture any “free” antibody that has not been bound with the sample. A positive hook line shows that the strip is not over loaded and the assay is valid. The lack of any hook line alerts the user that the test is over loaded and the sample should be diluted further.

Under what circumstances is it okay for manufacturers to use LFDs?

LFDs can be great tools for helping manufacturers in defining their Hazard Analysis Critical Control Point programs and determining if they are in compliance with their specific programs. Manufacturers can use LFDs to test raw ingredients, to check surfaces between product runs to be sure they are clean, and in some cases to test finished products. In every case, LFDs require validation to 1) demonstrate compatibility with the manufacturer’s product, 2) demonstrate the ability to detect known amounts of gluten in the product (spiked/recovery), and 3) prove similar results can be found with the R5 ELISA method which most people consider the “gold standard” for gluten testing.

How would a manufacturer go about validating an LFD?

Validation usually would consist of taking an ingredient, ingredient mix, and/or finished product negative control and running the test on the sample. If the test is negative, the manufacturer would then take the samples or sample mixes and add a 20 part per million gluten spike (a proportional mix of wheat gliadin, barley hordein, and rye secalin) into each sample (the gluten spike should be less than 10% of ingredient) and run the test again to be sure they get a positive result for each of the spiked samples.

If all samples are positive at the level expected, they would then take pure wheat flour and test it to be sure they don’t get false negative results with an over abundance of contaminant. Also, finished products should be periodically checked (every batch if produced on shared equipment or every few months if not) via third party labs using CODEX/AOACRI approved methods.

What are the limitations of using a LFD to test food for gluten?

Manufacturers utilizing LFDs must keep in mind that these tests:

1. Are qualitative methods used to screen for specific proteins or compounds.
2. Require two antibody binding sites to be present on the gluten protein or peptide in order for the methods to detect gluten. If only one antibody binding site is present (as may be the case with highly hydrolyzed foods and ingredients) than no sandwich complex can be formed and no test line will be visible and a false negative result will be reported.
3. Are only meant for detecting trace amounts of protein. When more than trace amounts are present, false negative results may occur.
4. Should never be used for finished product validation in and of themselves. Only a fully validated method such as ELISA or PCR or HPLC should be used for validating finished products.

Should consumers use LFDs to test products for gluten?

Consumers generally should not use these devices for home use because of the extreme variance in food products–each LFD should be validated for each specific product. Why? The burger they make at home will be different from the burger they get at McDonalds. The fat content will be different. The salt content will be different. The sauces will be different. All of these variables might make one burger compatible with the LFD and the other not. Something as “innocent” as a pinch of salt or a little vinegar can be all it takes for an LFD to produce a false negative or false positive result. For most consumers it is impractical or impossible to validate each product they would want to test before they actually test it.

Ridascreen Quick Gliadin: From the Manufacturer*

• “RIDA® QUICK Gliadin can be used as a swab test for the gluten determination on surfaces in the hygiene control and for the qualitative detection of gliadin/gluten in raw material. The test should only be used for the detection of small amounts of gluten (contaminations). The gliadin/gluten determination can be carried out for numerous raw materials after a simple ethanol extraction. Further applications, e.g. for processed foods and chocolate, are available upon request.”
• “The immunochromatographic test employs the monoclonal R5-antibody which is specific for the detection of gliadin from wheat and prolamins from rye and barley.”
• “Results are read visually. Generally, the higher the analyte level in the sample, the stronger the red color of the test band will be.”

*From manufacturer website

EZ Gluten: From the Manufacturer*

• “The EZ Gluten Test is an easy to use home test that will quickly detect the presence of gluten in foods. It is sensitive enough to detect gluten as low as 10 parts per million (ppm). This simple test is also small and portable enough for use in restaurants or when traveling. It can be used to test individual ingredients in foods, or to test finished and cooked products.”
• “The EZ Gluten Test uses the anti-omega gliadin antibody developed by Skerritt and Hill for detection.”
• “The test strip can be read visually for the presence of gluten in the sample.”

Tricia’s note: The EZ Gluten uses the omega-gliadin antibody that has very low cross-reactivity with barley (4 to 8 percent). If barley is a contaminant in the food product it will not be detected (or severely underestimated) using this test.

*From product package insert

AgraStrip Gluten G12 Test Kit: From the Manufacturer*

• “The AgraStrip Gluten G12 Test Kit is a lateral flow immunochromatographic assay that determines a semi-quantitative level for the presence of gluten.”
• “The Test Kit uses a new monoclonal antibody called G12 that specifically recognizes the pathogenic fragment of the gliadin protein present in gluten. This fragment is called 33-mer and triggers the auto-immune reaction in celiac patients.”
• “The G12 antibody used in the test kit reacts with some oat varieties.”
• Results can be read visually.

*From product package insert

Gluten Tox: From the Manufacturer*

• “GlutenTox Home is a rapid and user-friendly test for the detection of gluten in food and drinks which is harmful for people who suffer from celiac disease.”
• “GlutenTox Home contains a new antibody called G12. It was developed to specifically recognize the toxic fraction of gluten present in wheat, barley, rye, and oat that triggers the auto-immune response in celiac patients.”

*From manufacturer website

Tricia’s note: In the US oats not contaminated with wheat, barley, or rye are allowed in gluten-free diets. However, controversy over oats still exists.

Dietitians attending FNCE 2011: Please make plans to join us at the pre-FNCE workshop, “Celiac Disease Toolkit: Guiding Your Patients through the Gluten-Free Diet” on Saturday, September 24 from 9 am to 3:30 pm. Thomas Grace of Bia Diagnostics will be in attendance and will be discussing testing methods for gluten, including lateral flow devices.

Reminder: I have started a new website Gluten Free Watchdog (www.glutenfreewatchdog.org) to test products for gluten. Products from Arrowhead Mills, Authentic Foods, Bob’s Red Mill, General Mills, Holly’s Oatmeal, Mary’s Gone Crackers, Namaste Foods, Orgran, and Pamela’s have been tested. Products to come include those made by Lundberg Farms, Think Thin, Twin Valley Mills, and Novartis (Benefiber). Please visit the site and browse the specific products tested and share your thoughts about what you would like to see tested next.

I am thrilled to announce that www.glutenfreewatchdog.org is fully operational and taking subscriptions. This website was started to make gluten-free food testing data available to consumers and other interested parties. Ten product reports are currently available and four new reports will be added each month. The monthly subscription cost is $4.99. Please keep in mind that the cost of testing each food product fairly and correctly (e.g. 3 separate samples tested in duplicate using the standard sandwich R5 ELISA) is approximately $500. And this does not include the cost of product or shipping. As long as the gluten-free community feels the site is useful and is willing to pool their resources–contributing a relatively small amount of money in return for access to expensive test results—it will continue to operate.

Please help spread the word so we can test a lot of products. And thank you so very much for your support!

Inquiries: Please contact Tricia Thompson at tricia_s_thompson@hotmail.com