Thursday, August 15, 2013

< 5ppm gluten: Labeling and Science vs the Real World

Twitter, Facebook and the internet in general are wonderful things. But they are not without problems. This morning, as I normally do, I checked out Twitter mentions and Facebook posts. On Facebook I saw a nice post about our Pizza Crust.  I flipped over to Twitter and saw a mention from someone saying that our pizza crust had caused a gluten reaction and that it wasn't gluten free and that we were lying about our products being gluten free. This caused several of her followers to become very concerned. One was afraid they had "poisoned" her baby by giving her some of our products.

Now, obviously we never want to hear that someone feels unwell after eating one of our products. We have a formal complaint tracking system for this. However I did want to clarify our gluten free status, the strict testing and ingredient sourcing protocols we have in place and our 22 year commitment as a dedicated, family run gluten free company. This did not go well. Twitter's 140 character limit is not the best method to explain something to someone who is feeling unwell and angry. I want to expand on what I posted on Twitter in a form that allows for some continuity and clarity.

One of the things that I tweeted was

"We guarantee our product to be gluten free to < 5ppm (which is the limit of science)"
The reply to this was that this tweet was proof that we were not gluten free and that we shouldn't be allowed to label our products gluten free. I won't refer to this twitter conversation any more but if you feel you'd like to judge our response, our Twitter account is public and we stand behind everything thing we said.

So let's talk about PPM. What does it mean? What's safe? What are the rules? What's the science behind testing?

Warning: Science & Math Ahead

What does PPM mean?
PPM is short for Parts Per Million. OK most everyone knows that. But what does it -really- mean. A million is a pretty big number but it's helpful if we break it down to a real world example.

 So 1ppm is 1 ounce in 1 million ounces or 1 gram in 1 million grams. Hmmm, that's not easy. It becomes a bit easier when you convert that to pounds and kilograms

1ppm = 1 ounce in 62,000 pounds (31.25 tons)
1ppm = 1 gram in 1000 kilograms (1 metric tonne)

Now -that- starts to show you the kind of numbers we are dealing with here. 

Let's give you an actual example. A standard grain car contains about 80 tons (160,000 pounds) or 72.6 metric tonnes (72600 kilograms)*.

To find out what 1ppm of -something- in that grain car would be, all we need to do is run a calculation.

1 ppm      X ppm
31.25tons = 80tons

X = 2.56

If 1 ppm is 1 ounce in 31 tons, 1 ppm in a grain car looks like this:

2.56 ounces per 160,000lb
72.6 grams per 72,600kg

Probably the easiest way to think of what 1ppm means is this:

1ppm = 1 You in a room with a Million people.

or maybe

1ppm = 1 needle in a haystack of a million pieces of hay.

It's very small indeed.

The Rules

For many years, the Codex Alimentarius had a standard of 200ppm and this was widely accepted in Europe and the UK.

Until 13 days ago, the US had -NO- regulations as to what constituted a gluten free product. Let me repeat that. There were no regulations governing gluten free products in the US. Until August 2, 2013.

Up until last year, Canada (where we are based) had an informal limit of 20ppm for product labelled gluten free, although the regulations explicitly said "none". The 20ppm was  what was referred to as the "enforcement threshold". That is, if a product tested higher than 20ppm and was labelled gluten free it was subject to a recall. These guidelines were the strictest in the world and have been in place for at least 15 years and maybe longer. This is the regulatory environment Kinnikinnick "grew up" in.

Fortunately, things have started to become clearer world wide.

Last year Canada released the new guidance for gluten free labeling which basically formalized and clarified the pre-existing rules.

Since 2008, the Codex has an updated standard to set a threshold of 20ppm. (pdf)

In August 2013, after 9 years of deliberation and consultation (and I expect a good deal of lobbying), the FDA finally released the rules for what is allowed to be labeled gluten free. Companies have 1 year to comply.

One note on Oats and gluten free. While the new US regulations allow them, both Canada & the Codex do NOT allow Oats or products containing oats to be labeled gluten free. These agencies recognize that oats can be tolerated by most but not all people who are intolerant to gluten. This is a position we also take and we do not use oats in any of our products.

Because of the major uncertainty, especially in the US, about what gluten free actually meant, several celiac support groups drew up their own guidelines and created gluten free certification programs. The allowable PPM for certification from these groups ranges from 10-20ppm. While this filled a void, we choose to certify our own products.

Some of you may say "well, why would we trust you over a certification organization?" The answer is simple, if a company does not comply with a certification organization standards (ie: tests over 10 or 20ppm) then they risk being dropped by the certification organization (although I expect there is a process to "fix things" to retain status). Because we label our products with our own logo which claims <5ppm we are legally bound to provide that. There are no certification organization bodies that can claim that. There are a couple of companies who claim 100% gluten free but we'll get to why that is unsupportable shortly.

What's Safe.

Here's a good question. How much gluten is safe for the general celiac population. There's not an easy answer. It used to be extraordinarily confusing for consumers,but in recent years things have started to become clearer.
As of August 2013, at last, the consensus of governments, science, support groups and certification organizations has set the safe maximum level to be somewhere between 10-20ppm.

These numbers come with one caveat, especially at the 20ppm level. This is safe for the -average- celiac to consume and cause no damage. This does not mean -you- won't react to 20ppm. This is why some certification organizations have chosen to have a 10ppm standard. And remember, 20ppm per product is additive. This is up to 20ppm -per product- you eat per day. If you eat a lot of products in a day that are at or near 20ppm, even "average" celiacs may be getting enough to cause a problem.

This is why we have chosen the < 5ppm standard.

The Science of testing for gluten (and why claims of 100% gluten free are unsupportable)
Before we get to the < 5ppm standard, it might be helpful for a quick refresher on our standards. For a complete overview please see our post Gluten Testing and Our New Lab Tested Gluten Free Logo

I'll summarize some important points from that post:

  1. In general, all the tests on the market today are looking for gliadin, which is a protein in wheat. This protein is most commonly referred to as gluten. 
  2. It's  important to realize that for testing, gliadin & gluten are not the same thing. A gliadin result of 5ppm means it contains 10ppm gluten.
  3. The Level of detection of gluten is different than the level of quantification (translation: you can tell gluten is in something at a lower level than you can accurately say how much there is)
  4. The only accepted & verified test for determination of gluten is Enzyme-linked Immunoassay (ELISA) R5 Mendez Method.
  5. The most sensitive test available to science (that is recognized) is the R5 ELISA Quantitative test, with a Limit of detection of 1.5ppm gliadin/3ppm gluten.
  6. We use both the R5 ELISA Quantitative (how much) test and the R5 ELISA Qualitative (yes/no) test which has a limit of detection of 2.5ppm gliadin/5ppm gluten.
  7. The large majority of our suppliers are dedicated gluten free. For those that aren't, where alternatives are not possible, we have extensive documentation on their procedures and rigorous testing of every lot.
  8. We test both incoming ingredients and finished products using both the qualitative and quantitative tests depending on the ingredient.
  9. All our finished products must test below 2.5ppm gliadin/5ppm gluten.
The most important point for this blog entry are points 4 & 5. The testing we use is the only one that is universally accepted to be accurate. The very smallest amount that it can identify is 3ppm gluten. Not 1ppm and certainly not 0ppm. Can a company who claims to be 100% gluten free (which is 0ppm) have absolutely no gluten in their product. Yes. Can they prove it? No. Since the most accurate test in the world can only test to 3ppm, the most they can say is they have 3ppm or less or perhaps < 5ppm (see where I'm going with this?)

< 5 ppm and the real world.

Let take an example of a gluten free bread by Brand X. We'll say that 1 slice of our example bread is approximately 1 ounce or 28 grams. (your brand may be heavier & denser ;)

How much gluten would be in the product under the 20, 10 & 5ppm standards.

I'm only going to work in grams because there's really no useful measure below ounce.

So using our example above we know that 1ppm = 1 gram in 1000 kilograms.

20ppm = 20grams in 1000 kilograms or 20000mg in 1000000000mg = 0.00002
10ppm =  10grams in 1000 kilograms or 10000mg in 1000000000mg = 0.00001
5ppm =  5grams in 1000 kilograms or 5000mg in 1000000000mg = 0.000005

So for a 28 gram sample (28000 millgrams) we get the following totals.

20ppm = 0.56mg gluten/28 gram serving
10ppm = 0.28mg gluten/28 gram serving
5ppm = 0.14mg gluten/28 gram serving****

 (ok for those of you who still want ounces - 5 ppm  = 0.0000128oz/1 oz serving)

Now even at 20ppm, 0.56mg is not very much. About the weight of half of an eyelash. Or perhaps 1 crystal of granulated sugar. Or a grain of sand. It's a truly tiny amount. But our bodies ("our" being us celiacs) are truly remarkable and this tiny amount can cause issues for some people. Based on our current knowledge ("our" being Science's) very few would react to 0.28mg and virtually none to 0.14mg.

But here's the thing. You'll notice I've added ****  to the 5ppm amount calculation.

When we do the testing on our finished products we use the 5ppm Quick Test (qualitative yes/no). A Pass mean no detection. A fail means detection. (we've never had a fail by the way) So a pass means that there is, yes, you guessed it < 5ppm.

Update: I just received and email from our QA lab manager saying "hey, we haven't used quick tests for finished products for quite a while."

Here's what he had to say:

For finished products we use microwell strip format ELISA with heated cocktail extraction based on AOAC Official First Action Method No. 2012.01.

As a routine procedure we use the kit in a screening mode by running a 5 ppm standard alongside the samples and then comparing optical densities of the sample wells to that of the 5 ppm well using a spectrophotometer. If we were to encounter a positive sample that tests above 5 ppm we would build a full scale calibration curve with 5 standards.

It’s more time consuming (1.5 hrs just for the assay), but ensures better detection and/or quantification of glidain in heat treated foods.

(I actually knew this but yesterday had me linking to the post from 18 months ago and that's how we -used- to do it)

 Now I'll translate. Turns out that heat treated products (ie: any baked goods) are much harder to test because heat treatment changes the protein structure. Quick tests can't properly do the job. Even the sensitive test has challenges and the samples require extra, specialized processing (the heated cocktail extraction thing)

Another point to make. Every test we do takes an minimum 90 minutes.This stuff is not easy or cheap folks.

Here's the really important thing. The reason for this post. If we say our products are guaranteed to < 5ppm, this does -not- mean there is 0.14mg of gluten in it. (or rather less than 0.14mg) It means this:
  • We have researched our suppliers products, procedures and standards.
  • We have tested incoming ingredients.
  • We have made our products in our 2 dedicated facilities (where staff are forbidden to even bring gluten containing bread to lunch).
  • We have regularly verified our testing results with outside labs.
  • We have tested our finished products to have < 5ppm gluten.
 It would be awesome if I could tell you that our products had 0ppm gluten, but to do so would be inaccurate because the Science won't let me.

What I can tell you is that our products are the gluten free-est gluten free products around*.

* If some other manufacturer wants to dispute that I'm fine with that. Just post your testing and protocols for everyone to see, like we do. As a celiac of 16 years and gluten free consumer, I really wish you would.

Jay Bigam
Executive VP



Cathy Lauer said...


Jo said...

Fantastic information thanks

Molly (Sprue Story) said...

What a great, educational, not-too-defensive post. Conceptualizing ppm is tough, but your examples break it down nicely. Thanks, Kinnikinnick.

Ellen Bayens said...

Bravo, Jay! Kinnikinnick has been operating to standard and beyond before most people had even heard of gluten. Thanks for your leadership in the industry!

Warm regards,

Ellen Bayens

Vice-President, Victoria Chapter
Canadian Celiac Association

k said...


Anonymous said...

Brilliant post. Thanks!

Anonymous said...


chairex said...

What a great article - thank you.

I was diagnosed as coeliac 3 years ago (bit of a blow approaching my 60th birthday!) and live in the UK. Although I now feel much better than I used to, I find I am constantly unwell in a low-key kind of way. It's just depressing.

I had a 6-week holiday in New Zealand recently - and I felt great the whole time! No reaction to anything at all. But on returning to the UK the symptoms reappeared immediately.

So, more research - and I find that the standard in NZ is 3ppm, as against the UK's 20ppm.

Is this the cause of my problems? I don't know. But this article (and it had led me to lots of other things you have posted) has been really helpful in explaining things to me.

So, again, thanks!