Survey on Thresholds from FARE

[Macabre]

Don't know if this has been posted yet:

http://download.journals.elsevierhealth.com/pdfs/journals/0091-6749/PIIS0091674912017022.pdf

[CMdeux]

Thanks!

The bottom line data in that one is in Table V of the paper-- basically, the ED's are not as solid as they'd need to be in order to base regulations on...

but what they say is that if you want to allow only 1% of the peanut allergic population to experience 95% 'safety' (that is, reaction-free eating of a contaminated food) then you need to be capable of detection BELOW the 0.1 mg per SERVING level.  Per SERVING.  Not per cm³ or per mL.  Now we're down into ppt_{(trillion, to be clear)} measurements for some foods; just. like. that.

Probably.  Now, I say "probably" because the cohort in that study is NOWHERE NEAR large enough for those kind of statistics to be truly useful.  Why not?  Because the variance associated with that value is pretty much huge.

I'm also pretty confident in stating that a 95% confidence interval isn't acceptable to most people in that most sensitive 10% of patients or their families.   



Okay, suppose that we DON'T make the analytical problem so far out as to seem currently daunting.  So suppose that level is set to protect at the ED₁₀ level-- that is, only 10% of allergic people would need to be "worried" below that level...

well, that is down to 5 mg per serving.  So into low PPM, probably.   That is do-able, but probably not in all food matrices at this time, and probably not with the kind of accuracy anyone wants.

It also leaves out the people in that most sensitive 10%.  You know, the people who find the allergy the HARDEST TO LIVE WITH TO START WITH?

Now, I'm not suggesting that we need to work this at some ED_0.0001 level.  I'm not.

But I am saying that throwing several THOUSAND of the most sensitive people under the bus here is probably just wrong.

[spacecanada]

Has anyone factored in that people may eat more than one serving of something, more than ten servings of something, or some obscene number of servings which could put them over the minimum threshold and kill them?  Think of some teenager (sorry to pick on them) eating a whole box of cereal that should be 'safe' only to find out that it is only safe if they ate a maximum of two servings.  And what if they were in an unusually bad pollen season, and their asthma was flaring up, and their allergy cup was pretty full to begin with, pushing their threshold even lower than 'normal'.  Would one serving put them over the edge?  There are simply too many factors that affect allergies and their tolerances.  It's not worth the risk.

For us, it's allergen-free facilities only.  We've learnt the hard way too many times.  Good manufacturing procedures (same facility, allergens on a different line) and thorough cleanings aren't enough for the extremely low tolerance levels in this house.

[GoingNuts]

Good point on the serving size issue, spacecanada, I hadn't even thought of that.  If you responded to the survey, I hope you mentioned that!

[spacecanada]

If you responded to the survey, I hope you mentioned that!

Yes, I made it my biggest point!  I have done this: ate a whole bag of cereal after a half marathon, whilst my system is in overdrive and I'm even more sensitive to allergens... yep.  Recipe for disaster?  If there was some 'acceptable' level of cross contamination per serving, six servings eaten within 15 minutes could have killed me.

[booandbrimom]

Ran across this while reviewing the FARRP site information:

http://www.tno.nl/downloads/food_allergy_in_children.pdf

CM, thought you would like it especially.

[GoingNuts]

Yes, I made it my biggest point!  I have done this: ate a whole bag of cereal after a half marathon, whilst my system is in overdrive and I'm even more sensitive to allergens... yep.  Recipe for disaster?  If there was some 'acceptable' level of cross contamination per serving, six servings eaten within 15 minutes could have killed me.

[CMdeux]

Ran across this while reviewing the FARRP site information:

http://www.tno.nl/downloads/food_allergy_in_children.pdf

CM, thought you would like it especially.

That's a great find, Boo.  You're right, I do like it.

One thing that I wonder about though, is that it is known that apple and hazelnut (and peach, if memory serves) allergies are different (evidently) in geographically diverse populations across Europe-- one thing which is not yet clear (as far as I know) is whether or not those differences are genetically based or if they are dictated by regional exposure to pollens.

This tickled my memory because while I know many hazelnut allergic persons, I don't know anyone with the kind of sensitivity that this study suggests is reasonably normative in that population.  Ergo, I wonder if the people I know have "the other" kind of hazelnut allergy-- the one which is more common in another region of Europe, I mean.  South of the the Alps/Pyrenees, food allergies seem to be different in some fundamental ways.   This is also seen (somewhat) on an East-West boundary line in the middle Mediterranean, as well-- but is that because of exposure?  Cultural food preferences?  Not really clear what all of that means.  It's been assumed that it's based solely on exposure... but if that were true, then rice allergy ought to dominate Asia, and wheat allergy ought to dominate southern Europe.  But clearly that isn't the whole story.   

The vast majority of studies like this one are from extreme Northern Europe  (pretty much all of the studies large enough to have statistics associated are from Norway, Sweden, and Holland).  So just what does that mean when translating those threshold determinations to a more genetically (and geographically) diverse population such as that in North America?

I'll also add that they probably didn't include C.I. data for a good reason in that one-- like the other study, the numbers are small enough that those confidence intervals are probably huge unless they systematically excluded those with very high and very low thresholds.  Such people already probably self-select out of such studies to start with.

How many people with apparent reactivity like my DD's to peanut (or Boo's to milk) would consent to a DBPC challenge for the allergen?  Not many.  Of that population of people, I'm guessing that number would be less than 5%.

Similarly, why would someone with a tolerance like that achieved with desensitization go through one?  Again-- they wouldn't.  (Mostly.)

[LinksEtc]

https://www.foodallergy.org/laws-and-regulations/current-issues

FARE submitted its official comment to the FDA's public docket on a risk assessment for establishing threshold levels for major food allergens on May 13.

Responses to FARE's survey may be viewed

[CMdeux]

OH my.

It's stunning the number of responses which indicate severe reactions....

and also the sheer, overwhelming number of responses which indicate that a diagnosis was made at least in part by a naturopath or dietician.

Started with acupuncturist 18 years ago for myself, allergist finised helping me
recently. Allergist for kids. Husband self-diagnosed

[twinturbo]

Until the quack detritus is filtered out I have no idea how this branch of medicine is going to truly gain traction.

[CMdeux]

My take-away from this is that once you remove the "I think that I have an intolerance" group from things, the IgE-mediated food allergic people who answered this survey fall into two groups:

1) have seen terrifyingly severe reactions and want NO part of poking the dragon with a stick.  Ever.

2) are patting themselves on the back for their mad skilz at avoidance.  (Many of those seem to report RAST levels in the process).

The former group is about 50% of the genuinely allergic responders.

I was surprised that so few people keyed in on the notion that there is no practical way on earth that an industry that can't even track infectious agents in source materials could possibly be relied upon to offer allergen loads in a tight confidence interval.

I found my comments and Boo's quite easily.  Actually, over the years, I know some people's writing styles so well, I identified about 20 people via their comments. 


(ETA:  Thanks, in-house grammar-and-usage police...    )

[LinksEtc]

An interesting threshold comment

Food Allergy Research & Resource Program, University of Nebraska

FDA-2012-N-0711-0053

FARRP is an industry-funded consortium with more than 70 member companies.

FARRP would further note that processing and manufacturing processes likely affect the prevalence, severity and potency of allergenic foods in some cases. In the passage of FALCPA, Congress wisely exempted highly refined oils likely because of good evidence that highly refined peanut and soybean oils were safe for ingestion by peanut- and soybean-allergic individuals (Hourihane et al., 1997; Bush et al., 1985). Highly refined oils contain virtually no detectable protein with the protein being largely removed by the refining process (Rigby et al., 2011). However, the manufacturing of other ingredients derived from commonly allergenic sources as defined by FALCPA also greatly reduces the amount of protein or allergen from the source. Many examples exist including butter ester, butter acid, glucose syrup from wheat, ethanol from wheat, wheat starch, and shea nut oil or butter. FARRP would urge FDA to consider the use of quantitative risk assessment together with data on the protein levels of such ingredients to develop a transparent approach to source labeling exemptions for specific ingredients. Other ingredients, such as soy lecithin, contain trace levels of protein that may merit source labeling under some but not all uses. FARRP likewise encourages FDA to use quantitative risk assessment to make decisions regarding uses of such ingredients which should have source labeling and others which would not. Separately, FARRP has worked with a major soy lecithin manufacturer (Solae) to conduct a quantitative risk assessment on soy lecithin which will be shared as a separate response to this Notice. FARRP would further note that protein-containing ingredients from commonly allergenic sources can occasionally contain only trace levels of the allergenic protein. That appears to be the case with extensively washed fish gelatin (Koppelman et al., 2012). Fish gelatin is primarily collagen obtained from fish skins; washing during fish gelatin manufacturing removes parvalbumin, the primary fish allergen. Clinical evidence indicates that fish gelatin derived from cod is safe for ingestion by cod-allergic individuals except at rather high doses (Hansen et al., 2004). FARRP encourages FDA to take this sort of evidence into consideration when making a source labeling exemption decision on fish gelatin.

FARRP would assert that regulatory thresholds could be appropriately based upon use of the ED01
estimate where sufficient data exist to allow that estimate to be confidently made (for peanut, milk, egg,
and hazelnut) and use of the 95% lower confidence interval of ED05 estimate in other cases (for soybean,
wheat, and crustacean shellfish).

------------------------------------

General Mills Inc

FDA-2012-N-0711-0062

In addition to our direct comments, we fully endorse comments submitted by the Food Allergy Research and Resource Program.

[LinksEtc]

Another interesting comment

The International Dairy Foods Association (IDFA)

ID: FDA-2012-N-0711-0046

Definition of Tree Nuts Should be Reviewed
We would also request that FDA review comments regarding the list of tree nuts included in the Q&A document, previously submitted by Grocery Manufacturers Association to Docket No. 2005D-0490 on May 14, 2007. These comments identified some concerns with categorizing ten nuts including coconut as a major food allergen, due to the botanical classification and the lack of incidence of severe allergic reactions. IDFA strongly agrees with GMA's comments that review of the scientific literature establishes there are insufficient data to support the inclusion of beech nut, butternut, chinquapin, ginkgo nut, hickory nut, pili nut, sheanut, chestnut, coconut, and lichee nut in a list of major food allergens. Inclusions of "tree nuts" that have either no history of sensitization and elicitation of allergic reactions (beech nut, butternut, chinquapin, ginkgo nut, hickory nut, pili nut, and sheanut), or only a few cases of mild and non-life threatening reactions (chestnut, coconut, and lichee nut) contradict the intent of FALCPA and leads to an unnecessary elimination of food choices that are enjoyable, nutritious and convenient to allergic consumers

[LinksEtc]

Grocery Manufacturers Association GMA

FDA-2012-N-0711-0063

Moreover, GMA and the food associations who have joined this letter believe that the Agency should not require any recalls for packaged food products that may contain trace amounts of an undeclared allergen at or below the established threshold as the risk to human health would be extremely low.