Monday, April 4, 2011

How to Properly Interpret Ex Vivo Studies -- Gluten and Leaky Gut As an Example

by Chris Masterjohn

There seems to be some confusion about how to appropriately interpret ex vivo studies, which are studies that are not conducted in a living organism. 

Worse than that, there appears to be a common and rather dramatically misleading presentation of the data in Dr. Allesio Fasano's excellent study linking gluten to zonulin production.  But I'll cover that in my next gluten post.

In any case, some people construe this study as showing that eating wheat or gluten causes leaky gut in people who do not have celiac, when in actuality the authors treated isolated tissues and tissue models with gluten rather than feeding gluten to people and seeing what happened.

A study conducted in isolated human cells or tissues is not a "human study."  A study conducted in isolated human cells or tissues is useful for three things:
  • generating the hypothesis that the effect also occurs in live humans.
  • providing a possible explanation for existing data obtained from live humans. 
  • characterizing the mechanistic details of the molecular biology of some phenomenon that would be impossibly invasive to characterize in living humans.

Whatever such a study is, it is not a test of the hypothesis that the phenomenon occurs in humans.

Let's say we obtain data showing that gluten or wheat causes intestinal permeability in non-celiac patients.  And say we also have a study showing that it does so in an ex vivo model by stimulating zonulin release.  In this case, we can interpret the second study as suggesting a possible mechanism for the phenomenon uncovered by the first.  Our next step would be to try to show that gluten increases zonulin production in live humans under the same conditions under which it causes leaky gut.  

The definitive way to demonstrate the role of zonulin would be to show in a double-blind, randomized, controlled trial that gluten or wheat causes increased zonulin and intestinal permeability in the absence of any co-treatment, but that if the gluten or wheat is fed alongside a zonulin inhibitor, intestinal permeability will remain normal.  

Lest anyone think this is an unrealistic burden of evidence, Dr. Fasano's group has already published such a study conducted in celiac patients.   There is no reason someone couldn't conduct such a study in non-celiac gluten-sensitive patients, and indeed, I'm sure that if Dr. Fasano found that these patients actually have leaky gut, his group would probably perform exactly this type of study on them.

If we did not have any data showing that gluten or wheat causes intestinal permeability in humans, we could interpret an ex vivo study showing that it causes this phenomenon in isolated tissues as a justification to investigate the hypothesis that it also does so in live humans.

Of course, we already have evidence directly refuting such an effect in live humans.  Dr. Fasano's most recent study shows that increased intestinal permeability simply does not exist in non-celiac gluten-sensitive patients.  Yes, this is the same Dr. Fasano who published the 2006 ex vivo study.  

Likewise, a double-blind, randomized, placebo-controlled trial just showed that purified and chemically deamidated gluten (deamidation makes the gluten similar to what would appear in a celiac patient's intestines) does not alter intestinal permeability in non-celiac gluten-sensitive patients (though they provided poorly reported evidence that it may increase some symptoms).

Does that mean no one develops leaky gut in response to gluten except celiacs?  Not necessarily.

Emily Deans left a comment on my last gluten post linking to an interesting study showing very high rates of leaky gut among autistics and their first-degree relatives.  The fact that leaky gut was also seen in their relatives suggests that there is an element of biological inheritance, although without identifying the specific way in which this inheritance is transmitted, it does not necessarily imply anything about genetics because not all biological inheritance is genetic.  
And, in fact, there is some inheritance even in intestinal bacteria, which outnumber human cells 10-to-1 and substantially contribute to our phenotype, even though they for some reason are never considered part of the human organism.

The study identified 55 out of 88 autistic subjects who were not on gluten-free, casein-free (GFCF) diets who did not have leaky gut, and two out of two autistic subjects who were on GFCF diets and also did not have leaky gut.  Perhaps gluten or casein contributes to intestinal permeability in autistics.  Or, perhaps gluten-sensitive autistics, like the gluten-sensitive patients in Dr. Fasano's study, simply have stronger intestinal barriers than non-gluten-sensitive controls regardless of whether they are consuming gluten.  The study sheds no light on this matter.  

Incidentally, Emily Deans rocks not simply because she drives a lot of traffic to this blog,  but because her own blog and her new Psychology Today blog are awesome, and because her kids are so darn cute.

Any biological truth is intrinsically elusive.  Consequently, neither Dr. Fasano's most recent study showing that non-celiac gluten-sensitive patients do not have leaky gut nor the double-blind controlled trial showing that gluten does not cause leaky gut in non-celiac gluten-sensitive patients rule out the possibility that there is some subset of non-celiac subjects, somewhere on the face of the earth, perhaps autistics, perhaps not, who actually do respond to gluten by developing leaky guts.  But to my knowledge there is no solid evidence of this and if anyone knows of some, again, please post it in the comments.

Thus, the situation in which we are faced is one where the available data suggest that people identified as having non-celiac gluten sensitivity do not have leaky guts.  It therefore makes no sense to use a tissue study to explain a phenomenon that does not exist, nor to characterize the molecular biology of a phenomenon that does not exist.  

We can use Dr. Fasano's tissue study as a justification to hypothesize that there may be non-celiac subjects in whom gluten causes leaky gut, but so far the available tests of this hypothesis have refuted it.  

This does not mean that gluten is not problematic, nor that wheat is not problematic.  Nor is it an invalidation of anyone's dietary choices.  It's simply an honest, logical, and straightforward appraisal of the existing evidence.

And of course, if someone identifies wheat as a problem food, they should not be dissuaded from eliminating it from their diet based on a lack of evidence supporting the leaky gut hypothesis.  

Ultimately the purpose of understanding why a food is a problem is to understand how to properly identify other problem foods, whether it is possible or even desirable to recover tolerance for a given problem food, and if so, how to do it.  Thus understanding the why's and how's can often help people resolve residual health problems, and enjoy a healthier life to a fuller extent.  There will always, however, be uncertainty in the mechanistic details, and this uncertainty should never stop someone from acting in their best judgment to improve their health.  

Read more about the author, Chris Masterjohn, PhD, here.


  1. Do you believe that eating wheat is healthy?

  2. Mike,

    I'm not sure there's a straightforward answer to that question. It's very clear that it isn't necessary to eat wheat to be healthy. It's also clear that some people do not tolerate it well and improve their health when they remove it.

    It is, on the other hand, also clear that many people eat it with impunity and live a long and healthy life. Whether they do so in spite of wheat or because of it is anyone's guess.

    If someone wants to hedge their bets with wheat being unhealthy and best avoided, then I wouldn't blame them. However, addressing this question is particularly difficult in light of the fact that most existing commercial wheat products are quite definitely bad for you.


  3. Just added a paragraph to clarify my conclusion.


  4. Understood. Damn it is hard to find in vivo studies for this stuff. Paleos are all pretty convinced that wheat is pathological but it is indeed a bit of a leap to say it's because of permeability. You're right it might be WGA.

    I'm not so sure that anyone eats grains with impunity. Certainly not always with complete death, especially when they are prepared properly, but they are designed to harm animals with toxins, that's their evolutionary strategy after all. Still needs data, though.

    An interesting study would be a controlled diet with rice and wheat as the difference and then measure c-reactive protein. Or even other grains and soaked/fermented.

    But that's neither here nor there, we're concerned with permeability here, not cytokines.

    You should pose the question to Chris at The Healthy Skeptic to ask Robb and Mat the paleo gurus on the upcoming podcast. If anything it will get them to straighten out their terminology.

  5. Stabby,

    Actually, I'm quite enthusiastic about the possibility of in vivo studies precisely because they *are* being done.

    Dr. Fasano's group, for example, published a double-blind placebo-controlled trial in 2007 where they fed gluten to celiacs short-term either with a zonulin inhibitor or a placebo and showed that the zonulin inhibitor prevented wheat-induced leaky gut.

    The rice/wheat study would be interesting, but the Australian study I've cited that just came out essentially did exactly that. They fed a group of non-celiac gluten-sensitive patients who had been gluten-free a diet of either gluten-free baked goods or the same exact baked goods with deamidated gluten added. They found that gluten had no effect on C-reactive protein or intestinal permeability.

    I agree it is almost impossible to show anyone eats grains with impunity because there is no way to know what their health would be like without them. What I meant is that it is clear that many people eat grains yet live to what we understand to be close to a reasonable maximum of a healthy life.

    I think we're all reading each other's blogs -- I at least know Robb follows me on twitter and Chris K reads this and sometimes posts in the comments, so I look forward to us figuring this diet thing out together, even if we approach certain issues from different perspectives sometimes.


  6. Hey, I finally figured out why they deamidate gluten commercially-- it improves the baking qualities and makes it more soluble in water. I think a lot of the commercial gluten that's isolated and used in baked products etc. is deamidated. So the gluten they used in the IBS study may have actually been pretty similar to what is added to certain processed baked goods.

  7. Thanks Chris! Yes, my kids are very cute :-)

  8. Holy cute kids is right. :)

    Just another thought, perhaps there is a "perfect nutritional storm" that enables permeability but not necessarily each factor individually. Vitamin d deficiency, poor flora, superfluous eicosanoid activity, nutrient-deficiency and other factors tend to work together to produce NAFLD after all. Our gut isn't just inert and every antagonist destroys a little more until we have no more mucosal barrier, it also heals when we have much going for us. More healing than damage would mean no permeability.

  9. @Stabby; wow, if that isn't Catholic trying to explain purgatory, I'm not sure what is

    Plenty of reasons to avoid grains from a weight issue. Not that many from health issues.

  10. I'm not sure if I understand what you mean by that but I think it is a perfectly reasonable hypotheses since many factors play into disease and you're not necessarily going to get something like NAFLD from sugar without nutrient deficiencies and inflammation, although sugar is still sub-optimal after a point. Keeping in mind that I do terribly with grains and will never eat them again no matter what, I just want to arrive at the raw truth like cartoon Chris says: to be able to best treat patients. If we understand the mechanisms behind leaky gut then we can do much more than just say "eliminate grains".

    And what do you mean plenty of weight issues but not from health issues? How do you think we get fat other than metabolic dysregulation?

  11. Chris,

    I really appreciate all of the work you're doing on this topic and I'm enjoying the series. As I mentioned in my previous comment, I think it's important to investigate the mechanisms involved in this stuff and dispel any common myths that have emerged. I, for one, clearly misunderstood the research and assumed that gluten was upregulating zonulin production and contributing to leaky gut in non-celiacs.

    I view continuing education as a big part of my job as a clinician, so I'm always happy to learn more even if it means having to amend a position I've taken on a topic. I've always felt that as much as we've learned about human health, there's a lot more that we don't know. I try not to get too stuck on a particular point of view, because if there's one thing that history has shown is that our understanding of the body is always evolving.

    That said, as a clinician my primary concern is how my patients react to a particular food or class of foods. And it is my experience that most people feel better when they don't eat gluten-containing foods. Perhaps it's the WGA, or the lectins, or the gluteomorphins, or some other constituent. Who knows? In the end, I think it's the individual outcome that matters most. If someone removes wheat/gluten and they feel better, that's all the "proof" they need.

  12. Hey Stephan,

    Yes, I know that deamidation improves the baking qualities, and in fact tissue transglutaminase itself is often used in food processing! However, I did some looking around and couldn't find much evidence for wheat being processed for gluten deamidation except in a few select products like Japanese noodles. That said, when I looked I was expected tTG to be the main culprit and I didn't realize that chemical deamidation was a potential problem. Have you uncovered anything about how commonly wheat proteins are deamidated for commercial baked goods?

    Stabby, I agree that a "perfect storm" is often necessary for a particular disease to develop and NAFLD is a great example of that. I'm not sure "leaky gut" qualifies as a disease in the same sense as NAFLD or celiac, though. I think maybe it often is one of the contributors to the storm.

    Chris K, Thanks for the comments, and I definitely think you are doing great work too. You strike me as a very research-based but very open-minded clinician and I imagine your patients are very lucky to have you. And of course many of us are grateful for your contribution on the web too! And I agree that you don't need to understand the why of something to take action on it. Understanding the why, however, is very helpful for future directions. Sometimes you change something and it works for a while, then the honemoon ends. Or sometimes you change something and it improves, but some problems remain. So I think sorting these things out is important. I've spent a lot of time as part of GFCF communities online and from what I have seen GFCF is rarely the end of it.


  13. Chris, how do you think fermented what products play into all of this. I have the studies laying around somewhere and i'll oink them tommorow when I have time, but what I see is that even celiacs don't have nearly as strong as a reaction to lactobacilii fermented dough as they do to regular bread. If I recall correctly, scientists went so far as to engineer a lactobacilii so effective at reducing gluten content that almost all celiacs tested could eat the bread without negative effects.

    I will admit I am biased, because I enjoy the occasional homemade pulled pork sandwich and I don't like the pulled pork without slightly toasted sourdough.

  14. What explanation is there for "However, anti-gliadin antibodies lack disease specificity being found in 10% of healthy blood donors."?

    Does the presence of anti-gliadin antibodies mean that excessively-long peptide chains have got into the circulation? The above also excludes people who are not healthy enough to be blood donors.

  15. As a side note I recently saw an article suggesting that deamidated gluten is frequently used as a clarifying agent in red wine.

  16. Do you think PUFA might cause leaky gut which then allows gluten to become more of a problem?
    I have an autistic child and I am looking at ways to improve his health and would really appreciate any insights you might have.

  17. There may be more to gluten problems than just villi damage. Are you familiar with the work of Dr. M. Hadjivassiliou? He links gluten sensitivity to certain brain disorders. In fact, it was googling for causes of migraines that led me to his work and some by Italian researchers on the effects of gluten on the brain. This led me to eliminating gluten from my diet for the last 3 years which has virtually eliminated the migraines I have had for my entire life (52 years) as well several other nagging problems - joint aches, restless legs, etc. I'm not saying that everyone needs to eliminate gluten but it has certainly been a great benefit to me (n=1 data!). I've listed some articles I found below - two reliable, one more of a review source, but maybe they will give you some more data to pursue. Keep up the good work!

  18. Just stumbled accross this while looking into zonulin... have you heard of GAPS (Gut And Psychology Syndrome)? I can only take it on face value as I am not an academic, but it certainly seems to answer a lot of questions that are posted on here... very interesting blog by the way, nice to find a source of information simplified for those of us who want to know but don't know where to look... I would be interested to hear what you think of GAPS, if you have the inclination to check it out.

  19. With regard to high rates of leaky gut among autistics and their first-degree relatives, the theory is that the microbiomes of those who cohabitate (including family members) come resemble one another. There's that fabulous study, "Variation in Gut Microbiota Strongly Influences Individual Rodent Phenotypes", whereby the microbiota and urinary metabolite profile of rats that were moved to a new colony in a separate room of the same facility changed very quicklyto fit the new colony; when those same rats were moved into isolation, the new microbiota and urinary metabolite profile persisted (ie the new microbiota had successfully colonized).

      Interesting study for you; give it a quick read please.

  20. best parkour shoe


To create a better user experience for everyone, comments are now moderated. Please allow up to one business day for your comment to post. In order to avoid the appearance of spam, please avoid posting links, especially to commercial destinations, and using all-caps.