The NY Post recently published an article The miracle that cured my son’s autism was in our kitchen. It is a horrible piece of science journalism published in the “Living” section of the paper by a writer, Mackenzie Dawson, who writes about reality TV and how to organize your closet when she is not mangling science. The piece stumbles from trope to trope as it follows a tired formula.

First open up with a heart-warming anecdote of a child “cured” by a plucky parent who would just not give up. Let a bunch of cranks explain how it all works, and sneak in one paragraph of token skepticism that you then immediately contradict. Then finish off with flourish of how amazing your star anecdote is doing. Dawson gives us an added bonus, a cherry-picked study to create the illusion she did some actual journalism.

The story is an old one – dietary restrictions to treat autism. Dawson writes:

“I remember thinking, ‘Oh my God. What are we going to do?’ ” Levin recalls. “Everyone knew autism was a lifelong disorder and couldn’t be cured.”

Except that in Ben’s case, it could be. And it was.

The treatment was a gluten-free and casein free diet. Claims that gluten (a wheat protein) and casein (a dairy protein) cause or worsen autism have been around for decades. It’s interesting to see that “organic” has been added to the list, a tribute to the effectiveness of pro-organic propaganda. Here comes Dawson’s extensive science research:

While the scientific verdict is still out on diet as a cure, statistics point to a definite link between gastrointestinal issues and autism.

A 2012 study published by the Journal of Abnormal Child Psychology found a direct link between GI issues and behavior.

The first part of that is absolutely incorrect – the jury is not out. The decision was handed down years ago. A 2013 systematic review of gluten and autism found:

Currently, there is insufficient evidence to support instituting a gluten-free diet as a treatment for autism.

A 2008 Cochrane review of gluten and casein-free diets in autism found:

Current evidence for efficacy of these diets is poor.

And a 2014 review of the same concluded:

We observed that the evidence on this topic is currently limited and weak. We recommend that it should be only used after the diagnosis of an intolerance or allergy to foods containing the allergens excluded in gluten-free, casein-free diets.

There does appear to be room for larger rigorous trials, but over 40 years of research has failed to find any signal in this data. It is also possible, as some of the authors note, that children with autism might also incidentally have food allergies or insensitivities, and removing this food will have a positive effect on their behavior simply because it is removing an irritant. There is no evidence to suggest, however, that gluten or casein is playing any causative or significant role in the etiology of autism itself. There is no diet that treats autism, let alone is a cure.

If you’re interested, Paul Offit reviews diet and other fake cures for autism in his book, Autism’s False Prophets. He goes into more detail about the history of the gluten and casein-free diets for autism.

This information is easily accessible, and I’m sure if Dawson spoke to an autism expert they would have quickly filled her in. Maybe they did, but it just didn’t fit her narrative. She does quote Andrew Baumann, president and CEO of New York Families for Autistic Children, as her token skeptic, but he appears to be an advocate, not a scientist or physician. If my experience is any guide, Baumann may have been interviewed for a long time, providing all sorts of useful information that was dutifully ignored in the final article.

Regarding GI symptoms and autism, this link is currently speculative. There is evidence to suggest that a subgroup of children with autism do have GI symptoms, but its cause and relation to autism is unclear. Some scientists think that mitochondrial dysfunction may be the link. Others believe that children with autism simply react more to the background of GI symptoms and so it is more noticeable.

For the subgroup that do have GI symptoms, identifying any food sensitivities and easing their symptoms can certainly improve their behavior, but this does not mean there is any effect on the underlying autism.

At present there are some interesting theories concerning potential connections between GI symptoms and the pathophysiology of autism, such as GI flora and immune regulation, but these are all speculative. Part of the complexity is that autism is not one disorder but a category including many disorders that look similar clinically. This is an area for further research, but at present not the basis for any interventions.

What we do know is that there are strong connections between autism and a variety of genetic mutations, most of which involve the function of neurons, suggesting a direct effect on brain anatomy and function. This is further consistent with other lines of evidence about the development and function of the autistic brain. Our best current evidence suggests that children are born with autism, which has its roots in neural development while they are still in the womb. This does not leave much room for the role of diet, except as a modulator of symptoms in a subgroup of children with autism.

Why, then, do so many parents feel they have “cured” their children with diet. Many of the clues to an answer are given in Dawson’s article, but she fails to connect the dots. She reports:

Levin doesn’t call this particular cure a silver bullet for autism: There is no silver bullet, no one-size-fits-all approach. Rather, she credits his transformation to a number of things, including a home-based educational and behavioral program called the Son-Rise Program.

But one of the biggest factors was what was on his plate.

Children with autism and other developmental disorders do often improve just as a matter of time They are still growing and maturing, just on their own curve. Therefore any intervention may anecdotally seem to work. Further, families often try multiple interventions. Whatever intervention they are giving when their child crosses a developmental milestone, and starts engaging in a behavior they previously didn’t, will be credited with a “cure.” The situation is a setup for false correlations.

Also note that Levin’s child was receiving an educational program, something which is known to help. Why is she so sure, then, that diet was the key? Because people like to focus on the novel or surprising elements, rather than the mundane elements. It is far more likely that the educational interventions, combined with tincture of time, had the real benefit, but the magical diet gets the credit.

Conclusion

Dawson and other journalists writing for the “Living” section of their newspapers should really just not write about science. In this case she failed to consult or listen to any experts on the topic, she relied upon misleading anecdotes and the pronouncements of gurus. She failed to properly reflect that state of the research, or mention the common caveats in interpreting these stories. In short, anyone reading her article would be significantly misled about the science.

This is also a serious issue that directly affects peoples lives. Parents of children with autism face a daunting amount of misinformation. In this case they are being told that they need to put their child on a draconian and expensive diet, and will be made to feel guilty if they don’t. This is just magnifying the burden that raising a child with a developmental challenge can pose.  This type of misinformation causes direct harm.