Gluten-free diet eases autism, say parents

PENN STATE (US) — A gluten-free, casein-free diet may lead to behavior and physiological improvement in children with autism spectrum disorders.

A recent study by researchers at Penn State is the first to use survey data from parents to document the effectiveness of a gluten-free, casein-free diet on children diagnosed with an autism spectrum disorders (ASDs).

“Research has shown that children with ASD commonly have GI [gastrointestinal] symptoms,” says Christine Pennesi, medical student at Penn State College of Medicine. She and co-author Laura Cousino Klein, associate professor of biobehavioral health and human development and family studies, report their findings in the journal Nutritional Neuroscience.

“Notably, a greater proportion of our study population reported GI and allergy symptoms than what is seen in the general pediatric population. Some experts have suggested that gluten- and casein-derived peptides cause an immune response in children with ASD, and others have proposed that the peptides could trigger GI symptoms and behavioral problems.”

The team asked 387 parents or primary caregivers of children with ASD to complete a 90-item online survey about their children’s GI symptoms, food allergy diagnoses, and suspected food sensitivities, as well as their children’s degree of adherence to a gluten-free, casein-free diet.

Pennesi, Klein, and their team found that a gluten-free, casein-free diet was more effective in improving ASD behaviors, physiological symptoms, and social behaviors for those children with GI symptoms and with allergy symptoms compared to those without these symptoms.

Specifically, parents noted improved GI symptoms in their children as well as increases in their children’s social behaviors, such as language production, eye contact, engagement, attention span, requesting behavior, and social responsiveness, when they strictly followed a gluten-free, casein-free diet.

According to Klein, autism may be more than a neurological disease—it may involve the GI tract and the immune system.

“There are strong connections between the immune system and the brain, which are mediated through multiple physiological symptoms,” Klein says. “A majority of the pain receptors in the body are located in the gut, so by adhering to a gluten-free, casein-free diet, you’re reducing inflammation and discomfort that may alter brain processing, making the body more receptive to ASD therapies.”

The team found that parents who eliminated all gluten and casein from their children’s diets reported that a greater number of their children’s ASD behaviors, physiological symptoms, and social behaviors improved after starting the diet compared to children whose parents did not eliminate all gluten and casein.

The team also found that parents who implemented the diet for six months or less reported that the diet was less effective in reducing their child’s ASD behaviors.

According to the researchers, some of the parents who filled out the surveys had eliminated only gluten or only casein from their children’s diets, but survey results suggested that parents who completely eliminated both gluten and casein from their child’s diet reported the most benefit.

“While more rigorous research is needed, our findings suggest that a gluten-free, casein-free diet might be beneficial for some children on the autism spectrum,” Pennesi says. “It is also possible that there are other proteins, such as soy, that are problematic for these children.”

The reason Klein and Pennesi examined gluten and casein is because they are two of the most common “diet offenders.”

“Gluten and casein seem to be the most immunoreactive,” Klein says. “A child’s skin and blood tests for gluten and casein allergies can be negative, but the child still can have a localized immune response in the gut that can lead to behavioral and psychological symptoms. When you add that in with autism you can get an exacerbation of effects.

“If parents are going to try a gluten-free, casein-free diet with their children, they really need to stick to it in order to receive the possible benefits,” she says. “It might give parents an opportunity to talk with their physicians about starting a gluten-free, casein-free diet with their children with ASD.”

More news from Penn State:

chat20 Comments


  1. Don McArthur

    It’s like science, except that it isn’t.

  2. MJ

    @Don –

    If “science” would ever design and execute a proper study of the GFCF diet then there would be no need for studies based on surveys. But there has never been a good quality study that looked at how the diet works in children with autism who have GI issues.

  3. cd

    Survey data can be very useful or very misleading. Surveying a sample of parents that are already hell-bent on finding any clues to explain autism or reduce symptoms can be unreliable and offer very biased results. Certainly some sort of clinical trial would show whether or not this diet is helpful in reducing GI issues, and it could be done objectively.

  4. Jenny

    Unfortunately, there are no lobby groups raising money to fund studies that will ELIMINATE the consumption of things like wheat and dairy. Food has changed well beyond our capacity to evolve–we have to depend on grassroots initiatives to figure out alternatives to the non-answers science has provided us thus far. I applaud these scientists for considering the possibility that the answer isn’t always in the form of a pill.

  5. cd

    Jenny, who says we need to eliminate things like wheat and dairy? That’s a totally extremist idea; vegetarianism and veganism are not cures for all that ails man. I’m all for leaving pharma out of medicine, but natural remedies should be sought only as long as science can show some evidence of efficacy.

  6. Jenny

    Firstly, I mentioned wheat and dairy because they are the primary sources of gluten and casein in western diets. Secondly, I make no claim that vegetarianism, nor veganism are cures for anything at all. I am neither.

    My point was that it is very expensive to run long term science experiments. Private organizations typically fund research based on a hypothesis that encourages the consumption or purchase of a cure. Be it pharmaceutical or naturopathic, There is no money to be made in telling people to STOP consuming things.

    You call my idea extremist, but the survey documents that families are experiencing major improvements in their children’s development through changes in diet. That “science” hasn’t run a peer-reviewed double blind study on that hypothesis doesn’t diminish it’s validity. They cannot all be crazy.

  7. Kiwismommy

    It is about time that the experiences of parents are listened to by anyone in the scientific world. We are so sick of hearing the word “anectdotal” instead of eye-witness testimony. Pediatricians are taught to listen to parents as 24/7 caregivers for those that can’t speak for themselves or say where it hurts. Doctors have always known that parents are almost always the ones with the subtle clues that give doctors answers in tough cases. Until Autism, now parents are written off as lunatics looking for something to blame. The diet works miracles for some children yet parents are attacked and accused of “experimenting on their children” while mainstream medicine ignores their pleas for help, and science ignores the important studies. We simply cannot wait for those with conflicted interests to throw a bone once in a while that helps no one, and refuses to look at any answers other than a lifetime of drugging our kids into a chemical straightjacket stupor. I’m always amazed at those that seem offended by a diet, even when miraculous things happen. I am amazed that I am a “bad mom” for trying a diet that hasn’t undergone a double-blind placebo controlled study but if I put my kid on Ritalin, Prozac, or Adderall I am eligible for Mother of the Year and no one would question it. The diet gave me my child back, she withdrew as if I had taken her off of Heroin, and within 10 days she was speaking, pointing, and showed empathy. It wasn’t a “do you notice any difference”, it wasn’t “do you think we are seeing things that aren’t there because we did the diet and want to see changes”….no this was profound withdrawal and dramatic change from non-functioning to functioning child and we have never turned back. We have challenged the diet, and behaviors return so we stick with it. To do otherwise would be the experiment and abuse in my book. I don’t care what any study says, the diet works for my child and many others I know. The scientists can quit feigning interest or concern, as parents we are forced to find the answers ourselves or like those pretending to care, simply write our children off as lost causes…another sacrifice for the good of the herd. No thank you!

  8. cd


    You fail to see my first comment that surveying parents, especially parents of sick children, can yield extremely biased results. No real conclusion can be made from this to show that reducing casein and gluten has any affect on these children’s behavior or GI issues. While gluten and casein diets free diets have been clinically shown to have some positive effects among IBS and other GI symptomatic patients, a proxy survey provides little real evidence. While it may be expensive to run clinical trials, self report survey results must not be blown out of proportion. Doing some chi square tests and logistic regression on survey data is pretty weak when trying to get to meaningful clinical outcome.

  9. Don McArthur

    As I said, it’s like science, except that it isn’t. Emotion masquerading as rational inquiry. Understandable, but of next to no use.

  10. MJ


    As one of those “extremely biased” parents, let me just say that it is pretty easy to tell the difference between a normal bowel movement and a non-normal bowel movement. It is also pretty easy to tell when your child’s belly is completely bloated and when it isn’t, when your child has a reduced amount of stims, and (to a lesser extent) when the child makes better eye contact.

    That goes double if you have a high quality ABA style program going at the same time that is actually collecting data on the occurrence of behaviors before and after you start the diet.

    That goes triple if you, like me, have identical twins with autism and stagger when you try treatments. The differences between the twins when they aren’t both on the treatment and the timing when the changes appear after the treatment has started is very telling.

    So while you are right that anecdotal evidence isn’t the best and it is easy for a parent to be mistaken, it doesn’t follow that the result has to be worthless just because it is based on a survey. If nothing else, this research suggests that there might be a connection between diet and autism and that this connection is worth investigating.

  11. Ryan

    Cd. I am unsure why you refuse to give Jenny’s comments and credence. Your argument,by your own standards, is weak. You are speculating on parents emotional states by saying they are unstable and would be irrational with zero basis to support this premise.

    Further more, you provide no supporting commentary for the other arguments that you do provide. Are you saying that logistic regression of the survey data will over estimate positive results? I am unsure how that could be since the article clearly states it warrants further study and doesn’t draw a definitive conclusion.

    All i see is that you say the study is weak without supplying us with anything but opinion.

  12. cd

    Ryan, here’s your evidence of self report survey bias:

    Parental proxy bias:

    General issues of science use of survey data:

    Being someone that has to deal with BRFSS survey data on a daily basis, I know first hand that if there’s bias in the collection, there’s bias in the results. Old motto still rings true: garbage in, garbage out.

  13. sbdh

    the placebo effect is powerful in parent-implemented therapies such as this. how do they know if they should attribute changes seen to the diet rather than 5-day a week developmental preschool their child may attend, or the two hours a week of speech therapy and occupational therapy they receive, or even to natural maturation?

  14. jb

    Whether a kids diet is GF or not, food definitely affects behavior. When I was working with kids, we really tried to choose snacks which were good and had value for kids. Now taking that approach of food nutrition to children with different challenges and modifying accordingly, makes perfect sense.

  15. gerge
  16. slk

    Could it be that a mother with a gluten sensitivity is more likely to give birth to a child on the spectrum due to lack of nutrients? If that is the case, an autistic child may improve in some way from a gluten free diet because of his or her genetic disposition of gluten sensitivity, but would probably not be “cured” from the autism that happened prior to birth. Should we be looking for what a child on the spectrum did not recieve in utero or early in life to help him/her develop?

  17. jay

    @cd, Don McArthur, I guess both of you are pediatricians who don’t believe on gfcf diet..I have a son who is undergoing the said diet and has improved a lot. I agree with Kiwismommy-to see is to believe- and not pure anecdotal.

  18. Brad

    Of course diet affects illness. For ex: It takes only 6 minutes from consumption of sugar to reach the cells of the roots of teeth, or better yet, consumed celery cells to reach the same. Watermelon is great to help illness. I believe that most illnesses are fungal. Eat veggies and fruit. That’ll clear up a lot of illnesses.

  19. Brad

    Sugar is used as an example in the above. Sugar is a bad form of diet.

  20. anon

    Look the bottom line here is obvious. If your a parent of an autistic child and think the diet will help, then try it. If it works
    then stay on it. If it helps YOUR kid, then it helps your kid.
    I have a nonverbal six year old that does heavy stimming, and used to have unreal screaming meltdowns that would last sometimes for hours. We did two things. First we properly medicated her, which helped, but didn’t solve the bowel and emotional issues. We found out she’s allergic to wheat, corn and sesame. So, we went with the GFCF diet about three weeks ago, and her moods are totally different.
    I would suggest to anyone wanting to try it with their kid to go for it and ignore the nasayers. Speak with a nutritionist on how best to do it. The doubters can rail on me all they want. I can handle it easier now that me AND my kid are getting a full nights sleep since starting the diet:)

We respect your privacy.