Your own study has a pretty severe flaw as well, namely confirmation bias. To do this better you need to set a fixed time period in which you will look for gluten consumption and record everything she eats.

Then every so often you (the partner) looks back X days to see if she's eaten gluten. The key is, you can't know ahead of time whether she had a reaction, which is tricky I know.

The way you're doing it, you would find the same results if she she consumes gluten on a semi-regular basis unknowingly but has no allergy. You're only looking when you expect to find something, which is guaranteed to skew the results.

Confirmation bias does not exclude being correct. I have the same experience with a family member having gluten intolerance. We have experimented with diet. On gluten. Off gluten. Eating what is thought to be gluten free and having the usual reaction. And then finding out it wasnt. Having symptoms dissapear after a while of gluten etc. etc.

There might be confirmation bias, and it might be an n=1 test, but it is also 100% repeatable.

And it might not be the gluten, but proteins that are in the same foods as gluten. So when you stop with gluten you also stop with the other proteins. But that is really irellevant when you got the sensitivity.

There's no use in drawing conclusions from a broken experiment, whether it's from a team of scientists or an at-home test. Having no experience with gluten insensitivity myself, I am simply suggesting a way to bring some rigor to this conversation.

Ironically, all the posts calling confirmation bias are, themselves, probably the result of confirmation bias.

>Ironically, all the posts calling confirmation bias are, themselves, probably the result of confirmation bias.

Have a look at my comment (the first reply flagging this as possibly the result of confirmation bias). In what what does it sound to you like it could be the result of confirmation bias?

I really want to know your thinking on that. I try to actively attack such biases as best I can, but as I even mention in that comment, I understand that they affect us all. I'm just not seeing it here, though, so please tell me what you think I'm overlooking.

He's claiming a common lag reaction exists, that does exist. For example, the lectins in oatmeal are known to cause immune reactions with as much as a 3-day lag (common for our immune reactions in general, of course) if the oatmeal is not cooked for long enough. When did you last get a flu only an hour after exposure to a virus? Lag is common (not uniform) in our immune system reactions, which encompass a lot more than IgE reactions.

Ok, but your reply to my question might want to address the question:

>Have a look at my comment (the first reply flagging this as possibly the result of confirmation bias). In what what does it sound to you like it could be the result of confirmation bias?

Did you read the comment that that question​ was referring to? I never attack any notion of lag in a complex system being real possibility, so I'm not sure where you are coming from or why lag excludes the possibility of confirmation bias.

(And only loosely related, you might want to rethink using the example of abscense of influenza symptoms immediately after exposure to the virus as being in any way similar. You seem to be implying something I don't think you mean to).

Confirmation bias because we know most about IgE reactions, so tend to expect all other evidence to conform to that form of reaction - but reality is much more complex. Temporal (lag) characteristics fall under that.

There's no requirement of prior similar errors with confirmation bias - the bias exists as an expectation and if fulfilled either by ignoring, downplaying or discarding other evidence.

The particular bias, that IgE reactions exhaust the category of genuine allergy reactions is very strong - some allergists who don't read much still have it! It's certainly confirmation bias, together with recency. But we know that's false now, it's recent but well established medicine. We know mast cells do a lot more, now.

MCAS = Mast Cell Activation Syndrome MCAD (Mast Cell Activation Disorder) = MCAS or mastocytosis

Great links here: http://strengthflexibilityhealtheds.com/2016/02/04/diagnosin...

http://www.mastattack.org/2014/10/mcad-general-information-p...

http://evilmastcells.com/

http://www.jacionline.org/article/S0091-6749(14)02927-3/full...

Dr. Anne Maitland’s presentation on Allergies & Mast Cell Activation Syndrome in EDS Patients. https://www.youtube.com/watch?v=ktFdr-9rpIM&feature=youtu.be

https://mastcellblog.wordpress.com/mastcellguide/

You're back to trying to justify the plausibility of the parent being true and not answering the question, at least in a coherent way. Perhaps that's my fault in that it may be unclear which comment the question was referring to since this thread has gotten much longer. Here's the comment:

https://news.ycombinator.com/item?id=13901281

And here's the question:

>Have a look at my comment (the first reply flagging this as possibly the result of confirmation bias). In what what does it sound to you like it could be the result of confirmation bias?

Merely alerting someone that based on the language they used in their comment that there may be confirmation bias at play does not represent confirmation bias on my part. That's what question you were theoretically answering...

You could swap some words in the parent comment about the specific situation for some other completely ordinary and uncontroversial possibility (nothing to do with gluten or even health) and the concern still stands based on how the situation is being investigated. (See my other post about "remembering hits" and "forgetting misses" for the primary flaw expressed in the way he's going about investigating it).

(Note that two people can be wrong at the same time. When I agree that the B responders may indeed be motivated by confirmation bias that doesn't rule out the possibility that the A responder was also motivated by (an opposite) confirmation bias based on quite different previous observations. So, the parent may be interested in a genuine possibility due to a bias and pointing to a real but not the most likely possibility (it's hard to judge how committed he - the parent concerned about delayed reactions - is, so we are bound to interpret his comments in the way that makes the most sense and assume he is making only the rational point that is there to be made); and yet at the same time those who assume he's speaking due to confirmation do so because they don't recognize how common such delayed reactions are and assume he must be stretching a point or imagining such long delays because he's desperate to preserve a prior belief. Note this study was small and quick - 37 people and 9 days, apparently.)

As of now, I don't think you can get away from confirmation bias on the part of the B responders, at least, though. If the parent is true (and yes, those delayed reactions do exist), then those who accuse that poster of confirmation bias can only get there by assuming (consciously or unconsciously) that the reactions they are most familiar with (intolerance and IgE) are exhaustive, since they can't read his mind and know that he would reject future results from better experiments that are consistent with this experiment. He (the parent) raised a real possibility therefore the knock-down proof fails, as for now. We don't know that non-celiac gluten-specific reactions don't exist. That assumption that reactions to gluten or anything else have to be similar to IgE reactions or intolerance is confirmation bias (of the common "what I've seen so far is what there is" kind) - but of course our brains are associative, and we like to save thought and energy, so there's no massive shame in that; confirmation bias is a part of everybody's daily experience in at least small ways because that's how our brains work and must work. Overlooking even the possibility of disconfirming evidence (that in fact exists in similar contexts, such as oats) is quite a strong form of confirmation bias.

It's worth noting that the study doesn't show that there aren't reactions to gluten that aren't celiac reactions; in fact it shows that there are people who have such reactions, it's just that they also react to a lot of foods (this is actually common with MCAS - Mast Cell Activation Syndrome, for example - some people with MCAS end up restricted to a handful of foods.) Such patients turn out to have a helluva lot more problems than just gluten, so they benefit from avoiding a helluva lot more than gluten.

You may to have to disambiguate "concern" "situation" and "he" if this reply doesn't seem suitable to you. But if you meant the original scientist, he didn't allege confirmation bias, even by implication so far as I can tell.

I do mean to compare the lag re toxins and the lag re infections, because these response systems are complex and not severed from one another (both are mediated by mast cells); our system often has to react to foreign particles with a general defense, since it can't clearly distinguish in every case (or immediately) whether it's encountering a toxin or invader, or bacteria vs virus. No doubt errors are made both ways.

Celiac reactions also often have that 3-day lag, btw, very understandable because Celiac is a unique white-cell mediated autoimmune disease.

I meant that in the sense that posts claiming the parent was showing confirmation bias were in themselves likely trying to fault the parent's experiments in a way that reaffirmed their own existing beliefs.

I'm not saying that they were wrong in pointing out confirmation bias, but that they aren't coming from a much more objective perspective themselves.

Gotcha. I appreciate the reply.

I understand what you were thinking now, but I reject that assertion, at least for my comment. In it, I explicitly accept the possibility that he's right, after warning him that based (only) on the text of his comment, he might not have sufficiently guarded himself against a confirmation bias.

I replied because the way he described it is classic of a "remember this hits, forget the misses" cognitive problem. There's no prospective testing described (other than the initial elimination of gluten from her diet). Their confidence seems to mostly rest on post-hoc examinations of diet once something is noticed (so you're missing most of the data and potentially most of the misses) and even gives himself an out if they do an examination and it does miss (the reference to needing a "deep enough" investigation, implying that he open to chalking up a miss as that this time they just didn't look deep enough).

Your criticism of a randomized, double-blind, placebo-controlled study is essentially "but it doesn't fit my own (unblinded, uncontrolled) anecdote".

You offer an edge case that was not tested by the study.

Your criticism is constructive/relevant, but there is always some area of our understanding that the experiment doesn't reach... basically saying "here is a different but similar hypothesis that was not tested by the trial".

Ok; fair enough. But in the eyes of most readers, this data set generated by an insanely tight scientific design suggests that the optimal allocation might lie in a different direction of research for the individuals suffering these symptoms.

Interestingly: you can perform your own blinded experiment. If I were in your position, I would want to know whether my wife and I were planning our lives around a superstition. The way to answer that: an n-of-1 blinded clinical trial. It's a fair amount of work, but not as much as gluten avoidance, and would answer the question of your own circumstances better than any generalized scientific result from the literature.

>if we look deep enough

That could be the problem. Also youll always notice an improvement due to placebo.

Yes, 3 days is way to short...

I find it infuriating how arrogant and counter-logical American nutritionists are on this issue. They believe that any allergy will show up if you merely put a piece of food in your mouth, and do so almost instantly. For them, the oral challenge is a golden test which always works: https://www.foodallergy.org/diagnosis-and-testing/oral-food-...

"If you have no symptoms, food allergy can be ruled out."

Some of this is obstinate pendanticism about what the word allergy means, but mostly it is just arrogance and refusal to understand reality and logic.

I for example, will throw up in the evening if I eat a large portion of barley for lunch. Obviously, I shouldn't eat barley and my European doctors have no trouble understanding that. But according to American medicine, there is no problem since I obviously pass the oral challenge for barley.

Do you know what the words "allergy" and "sensitivity" mean? They are being used as jargon by medical professionals, and it's not clear to me that you're using the words the way they do.

A 3 day delay doesn't make much sense. The gluten would be digested and out of her system by then.

Has she been to a doctor?

It's called the IgG intolerance and shows up after 24h-72h of eating food that one doesn't tolerate. The difference between intolerance and allergy (IgE) is that allergy shows up immediately.

IgE allergies show up quickly - but many allergy (mast cell) reactions, are not IgE reactions and may take considerably longer to show up (for example, MCAS reactions generally peak in the middle of the night, unlike IgE reactions.)

Our immune system is complex, IgE is just a tiny part of it. Some introductory reading:

MCAS = Mast Cell Activation Syndrome MCAD (Mast Cell Activation Disorder) = MCAS or mastocytosis

Good MCAS links here: http://strengthflexibilityhealtheds.com/2016/02/04/diagnosin...

http://www.mastattack.org/2014/10/mcad-general-information-p...

http://evilmastcells.com/

http://www.jacionline.org/article/S0091-6749(14)02927-3/full...

Dr. Anne Maitland’s presentation on Allergies & Mast Cell Activation Syndrome in EDS Patients. https://www.youtube.com/watch?v=ktFdr-9rpIM&feature=youtu.be

https://mastcellblog.wordpress.com/mastcellguide/

It isn't necessarilly that unresonable. For example, when a person takes allergy shots (alergens injected directly into the veins) there will be soreness and swelling days after the injection.

By similar logic, a buffer over-run that doesn't cause your program to crash should cause no issues once the data that was being processed at the time has been freed from memory.

Confirmation bias can be a strong thing. And it affects us all. I'd suggest you rexamine to be sure that it or a similar cognitive bias is not at play, particularly given the wording you used above which is very suggestive of putting yourself in an ideal situation where such biases can work:

>we sometimes are able to sort out ingredients after the fact, and if she has the reaction, it will show up if we look deep enough.

Not saying she doesn't have an issue with gluten (I don't know you two so how the heck would I know?) but given she has symptoms of something you owe it to yourself to make sure you are addressing the correct root cause.

EDIT: looks like cpncrunch's sibling comment is getting at the same thing, only I would not use the term "placebo" for what could be at play here.

It's interesting to consider that occasions of feeling good may not receive the same post hoc dietary scrutiny. Gluten is really hard to avoid entirely.