Tuesday, September 7, 2010

Diet and Autism 1

Let's begin with the Harvard Mental Health Letter (April 2010 page 4). HMHL just loves to rain on my "diet and mental health" parade:

"Researchers have long disagreed about whether gastrointestinal problems may underlie some symptoms of autism spectrum disorders. This has not stopped some researchers and celebrities from promoting theories and special "autism diets" with no scientific support. Yet these unfounded recommendations might appeal to grieving and vulnerable parents who are heartbroken about a child's sudden developmental regression."

Wow. That's harsh. I hope to use the next several posts to explore some of the theories behind what might be causing austim and to see just how wacky and dangerous these diets really are.

Autism spectrum disorders (autistic disorder, Asperger's disorder, Rett's disorder, childhood disintegrative disorder, and pervasive developmental disorder not otherwise specified, together known as "ASDs") seem to impair a child's ability to communicate and interact with other people. ASDs are also characterized by ritualistic or repetitive behaviors, such as tapping fingers or head banging, or picking. Autistic children in general are more picky than children without ASDs, and temper tantrums around unsatisfactory meal times can be extreme. (A decent point is that restricting a picky child's menu by going gluten and casein free could be a nutritional issue.) Constipation is also more common among kids with autism, and it seems that kids with ASDs frequently experience abdominal pain, stool leakage, bloating, and reflux (1).

The Harvard Mental Health Letter is blunt, at this point: "There is not enough evidence to support special "autism diets" that eliminate casein... or gluten." Previous studies were not properly done, or too small, or didn't even use standard measures to diagnose autism in the first place. There is only one double-blind, crossover trial of casein and gluten free vs. regular diets (not sure how you really double-blind that but okay), of 15 children with ASDs, and after 12 weeks the diet seemed to have no effect (2). Some of the parents felt the intervention trial was more successful than the independent reviewers.

All right, all right, we get the picture, Harvard Mental Health Letter. But let's step back a bit. What is autism, and what causes it? Well, there is no universal theory, no consistent genetic findings (though autism does seem to run in families sometimes), and no agreed upon biological mechanism for what causes autism (3). Some researchers feel it is a brain development problem. Others wonder if autism is a set of disorders similar to phenylketonuria (where an inability to metabolize a certain amino acid can lead to progressive mental retardation, brain damage, and seizures, but can be averted by not eating that amino acid, phenylalanine).

The whole idea that gluten-free casein-free diets might be useful for autism is based on theories that exorphins (protein fragments that act as opiates) from gluten and casein might make it through the intestinal barrier and go on to act on the brain, causing harm. A similar theory is at work in gluten-free diets and schizophrenia. One very interesting tidbit of information: kids with autism and their family members seem to have leakier guts than families without autism (4). (Here's a finding which makes the Harvard Mental Health Letter's statement that researchers are promoting "autism diets" with no scientific support, seem, well, cranky. To be fair, the largest and best gluten-free casein-free diet trial was published in April of 2010 too, and the leaky gut study I'm looking at is epublished ahead of print, meaning it doesn't seem to even be officially published yet.)

Wait a minute. Back up. How does one measure a leaky gut? Well, to test the gut integrity, one can drink a solution of metabolically inert sugars, like lactulose and mannitol, and in a person with a leaky gut, those sugars will end up in the urine (this is called an IPT test). In addition, a leaky gut tends to be an inflamed gut, and one can skip intestinal biopsies and check fecal calprotectin (FC), a protein produced by intestinal granulocytes. Lots of FC in your poop apparently means you probably have an inflamed bowel. Of course celiac disease is associated with inflamed, leaky guts, so the researchers in the study took 90 kids with autism spectrum disorders and 146 of their first-degree relatives, and also 64 children and 146 adult controls. Everyone was given an IPT test, some were checked for FC, and all the kids with autism were screened for celiac (using anti-tTG antibodies, IgG anti-gliadin antibodies, and IgA anti-gliadin antibodies, anti-endomysium antibodies, and the genetic testing for HLA DQ2 and 8). Relatives and controls with abnormal IPT and FC tests were similarly screened for celiac.

Sounds interesting! What are the results?

Abnormal leaky gut - the IPT test (a higher percentage of lactulose and mannitol in the urine)
Adult controls = 4.8%
Child controls = 0%
ASD patients = 36.7%
Relatives = 21.2%

Pathological fecal calprotectin (FC - measure of gut inflammation)
ASD patients = 24.6%
Relatives = 11.7%
(Due to budget constraints, FC was only measured in controls who had an abnormal IPT test - none of them had FC values above the normal range)

ASD patients and celiac testing
Genetic predisposition for celiac (+HLA DQ2 and/or DQ8): 32% of kids with "leaky" IPT test
Genetic predisposition for celiac (+HLA DQ2 and/or DQ8): 35.9% of kids with normal IPT test
Gastrointestinal symptoms: 45.5% of kids with "leaky" IPT test
Gastrointestinal symptoms: 47.4% of kids with normal IPT test.
Positive AGA IgA: 1.6% +/- 2.5%
Positive AGA IgG: 13.8% +/- 24.2%
Positive tTG: 1.04% +/- 0.91%
EMA: all negative

Relatives and celiac testing
Positive AGA IgA: 0.8% +/- 0.4%
Positive AGA IgG:8.4% +/- 5.9%
Positive tTG: 3.0% +/- 3.7%
EMA: all negative

Well! That's a mixed bag. The overall findings - there seems to be a subgroup of kids with autism and their close relatives who have leaky guts. The celiac findings are a little more all over the place. GI symptoms seemed to have no correlation with gut leakiness, meaning the standard recommendation to investigate for celiac or intestinal barrier problems only in autistic kids with GI symptoms seems to fly in the face of scientific findings.

Perhaps more importantly, "gluten itself augments IPT" (in other words, makes gut leakiness worse in cellular models (5)(6)(7)). "We can hypothesize that subjects with ASD are gluten-sensitive.... and hence their intestinal barrier function [abnormalities] will ameliorate with with a gluten-free diet. The well-recognized intestinal mucosal effects of gliadin - the major component of gluten - would justify a treatment with gluten-free diet in ASD."

There is a lot more to discuss. But for now, I'll leave the above information to simmer for a little while.

The bottom line from this post:  The clinical trials evidence for gluten-free, casein free diets is poor, so far (though I'll go over what I consider to be the best study later this week).  BUT, there is newer evidence of a rather large subset of kids with ASDs who have an especially leaky gut, and the susceptibility to the leakiness seems to be genetic.  You are not going to find these kids by looking at patients with celiac markers or by looking at patients with GI symptoms.


  1. You spurred me to fact-check the autism advocacy ads running on TV about a year ago, stating that autism affects one in 150 people.

    According to UpToDate.com, the prevalence in the U.S. of autism spectrum disorder is between one in 111 and one in 500. One in 150 is close enough. And way too many.

  2. Even in 2003 JAMA was reporting a 3-4 fold increase in cases over the previous 30 years. That was alarming enough, but hard to sort, as the diagnostic categories had changed over those years. But even between the early '00 studies and the late studies of that decade the prevalence increased again. I'll go over the exact numbers in a post, but in boys the number is particularly scary - I've seen numbers reported that autism now affects between 1 in 77 to 1 in 92 boys in the US. As always I'm interested in why it would be increasing, especially over the last 30-40 years like that.

    I think overall it is difficult to study as there is likely a mix of diagnoses with different pathological causes going on.

    One thought, though - if diet is causing some of the change, and the exorphin mechanism is similar to what theoretically might be causing some schizophrenia, then why isn't schizophrenia increasing? Hmm.

  3. Great post! What are your thoughts about the Specific Carbohydrate Diet and the success families are having treating ASD kids with the diet?

    Can't wait for the rest of the series.

  4. Thanks for the post. I've been waiting for you to write about Autism, as it's something that we hear about so much these days.

    I see so many problems with these numbers. It's really quite frustrating. First of all, I don't know what baseline they are using for "normal" and "abnormal" when it comes to leaky gut, but unless they're coming from hunter-gatherers with perfect health these guts are probably still quite permeable.

    Let's assume for a second that the baseline for gut health is actually a healthy gut. What this now says is that gut permeability isn't a direct cause of autism, due to the fact that not every ASD patient exhibits it. But that doesn't necessarily mean that gluten or casein intake isn't the cause. When two things are correlated in the way that ASD and gut permeability are, yet there instances in which one exists without the other, it is pretty safe to say that they likely share a causal factor, in this case likely dietary lectins like gluten.

    All this said I do suspect that the gut permeability has a causal role, and the reason it's not detected in all of these subjects is because of baseline issues. Additionally, our tests for gut permeability are probably highly incomplete which would allow for other gut issues to go undetected.

    It irritates me to no end when someone recommends a "gluten free" or "casein free" diet to combat something like autism. Look, gluten may have caused the syndrome in the first place, but to reverse them, you're probably going to have to do a lot more than eliminate gluten. My sister has celiac, and eats gluten free as a result. But she doesn't cut out grains from her diet, she just avoids gluten grains and uses gluten free substitutes for these foods. So pasta made of rice or quinoa, corn starch instead of flower, etc. Needless to say since being diagnosed with celiac she has gone from thin to chubby.

  5. Hi Geoff. My main issue is that obviously we have been eating gluten and casein for thousands of years, and autism has only increased in the last 3-4 decades, so we are talking about other factors. With wheat at least we can say, hey, wheat has changed rapidly - quick-rise yeast, industrial "superlectin" genetically modified wheat could possibly change how we absorb it or tolerate it.

    The IPT test for intestinal permeability is actually rather neat - let me explain it in more detail. Patients are given two types of sugar, mannitol and lactulose. They have different molecular sizes and absorption routes. 2 grams of mannitol and 5 grams of lactulose are given to fasting subjects, and then urine samples are collected for the next five hours. The urine is measured with "high performance" anion exchange chromatography and there is an expected ratio between the two sugars in a "normal" gut. The normal value should be less than 0.030 LA/MA. Since they are using a ratio of sizes of molecules seeping through to determine the leakiness, they do get a reasonable physical measure of actual "leakiness" which is kind of a cool design. In any even the autistic kids had LA/MA ratios twice as high as the child or adult controls.

  6. The biggest mistake, IMO, is assuming that gluten/casein is the be-all-and-end-all of food intolerances. I fit most or all of the criteria for asperger syndrome, and I am gluten intolerant, but my gluten exposure symptoms have nothing to do with my AS symptoms.

    HOWEVER, I also have a real problem with salicylates, and they have a huge, noticeable, and very direct relationship to my AS symptoms. I do a great job of passing for merely geeky in my regular life, but feed me a couple of aspirin and a bowl of tomato sauce, and you'll quickly see me become truly impaired in some classic autistic ways. I become clumsier, my auditory and visual processing go all to hell, I become totally mute, and my usual not-so-great eye contact becomes nonexistent. Also, I start drawing little tiny repetitive patterns on whatever surface is available. I've experimented and seen that change from normal-me to impaired-autistic-me happen in half an hour, given a sufficiently large serving of curry.

    Gluten probably has something to do with it, in some cases, but it's not the whole picture. I am dying to see more research on autism and impaired Phenolsulphotransferase pathway.

  7. Gluten opens the gate and other compounds march through 10-deep? So let's say the like of salicylates, glutamates, etc, are perhaps the culprits. In a normally functioning and sealed gut, these compounds may only get through in small amounts. But increase the environment that can lead to a leaky gut (in which gluten is no doubt a player), and these compounds seem to get an access all areas pass to your neurological system and go nuts in the mosh pit.

    So treatment? You do what you can to make your gate unleaky, and this should include going gluten and perhaps/probably casesin free. But there are other factors. And perhaps this explains some of the plateau effect that Emily noted in her most recent post.

    Emily - will flick you some stuff off air on ASD/vitamin D/gluten links... will see if I can find the reference for gluten leading to a rapid turnover of vitamin D stores...

  8. Jamie, there may be something to that, but I'm not totally convinced. From personal experience: I've been off gluten for over a year, but an accidental run-in with salicylates still provokes a reaction. From my reading: poor sulfation is extremely common in autistic people, and proper sulfation may be directly involved with the integrity of the gut lining, as well as things like detoxing salicylates. I'm still clobbering my way through the reading on that one, and ultimately it may require an understanding of biochemistry that I lack. But... I still think that zeroing in on gluten gives autism reasearch a bad case of tunnel vision sometimes. What if we start out, from birth, with guts compromised by inefficient sulfation, and this is what makes us more susceptible to the ravages of gluten?

  9. Could be... the mistake to make (and this get made often enough in nutrition) is to aim for reductionism... that if we can narrow something down to one cause, then we can fix it. Research suffers heavily from that. If research isn't convinced that gluten fixes the entire ASD picture then it tends to conclude that it isn't part of the problem, carry on with your bread.

    I see this time and again in clinical practice... people want THE ONE thing that they can do to fix . I'm a proponent of Paleo nutrition and people will ask me if it is the low carb OR the zero gluten OR the no/low dairy that I benefit from... because if they can do that ONE thing, they won't have to go the whole hog. Telling people that the whole is greater than the sum of the parts is a hard sell.

    And research programmes generally don't want too many factors involved in the aeitiology of a disease/disorder... because that makes it all the more harder to come up with an all encompassing pharmaceutical treatment.

  10. Simple Perspective - SCD is interesting but I would like to see the research implicating "bad " gut fungus and bacteria as the inflammatory problem.

  11. What would be a sensible update interval for this blog post?


Tired of receiving spam comments! Sorry, no new comments on the blog

Note: Only a member of this blog may post a comment.