For a Coordinated Action on the Relationship Between Nutrition and Brain Function

Concerned about autism for the past fifteen years, I have been interested in this subject for nearly a decade. In my profession, I often observe the links between diet and various pathologies, but I was still surprised that such a link could exist between diet and brain function—and that this link could sometimes be so determining in the school integration of an autistic child. After the initial surprise, I found that a great deal of curiosity, a significant investment of time, the collection of numerous publications, and direct contacts with their authors were necessary to fill, at least partially, the gaps in training and information.

In fact, the relationship between nutrition and brain pathology is an idea that dates back even to twentieth-century medicine. For some time now, a few teams have explored the effect of gluten-free and casein-free diets (GF/CF) on autistic children. While clinical and fundamental knowledge in this very complex field of the relationship between nutrition and brain function tends to accumulate, enormous gaps still persist.

Regarding the GF/CF diet in autism, the pathogenic hypotheses underpinning its positive effect mainly involve opioid-active peptide fragments contained in casein (casomorphins) and gluten (gliadinomorphins). These fragments would reach the central nervous system in sufficient quantities to disrupt brain function. An abnormal permeability at the intestinal level associated with enzymatic deficits (notably peptidases) would also be part of this theoretically relatively simple but probably very complex pathogenic framework. We cannot exclude the possibility that other peptide sequences play a role in this pathogenesis.

The effects of GF/CF diets were primarily studied by Karl Reichelt in Norway as early as 1979, and Paul Shattock in Great Britain, as well as by the Italian team led by Loredana Lucarelli from an allergy perspective. In Reichelt's and Shattock's studies, serum antibody assays (anti-gliadin, anti-casein) and urine chromatography (HPLC) were performed before the diet was instituted. Then, throughout the diet, new urine analyses were conducted at regular intervals. Anti-gliadin and anti-casein antibodies are found more frequently in autistic individuals than in the normal population, which would only mean that fragments of gliadin and casein have passed through the intestinal barrier more frequently in autistic individuals. Urine analyses reveal characteristic urinary excretion profiles. Peptide fragments are found in the urine of some autistic children and correspond to opioid-active molecules, and their good predictive value could determine the type of diet to try.

Many parents, like Reichelt and Shattock, have observed that signs of improvement, when they occur, are faster when the diet is instituted for a young subject and is properly monitored: these improvements are typically seen in less than a month in children around 7 to 8 years old, and the digestive disorders that are often present improve first. This effect is still perceptible even after a year on the diet, and the urinary excretion profiles continue to normalize beyond a year (and even after four years according to one study).

Is there not an ethical issue in not mentioning this possibility during diagnosis, even if it is not fully substantiated, knowing that a GF/CF diet with the guidance of a nutritionist poses no danger? It is interesting to note, and this has been documented elsewhere, that family members suffering from ankylosing spondylitis observe remission of these pathologies with a GF/CF elimination diet. The slow improvement and the maintenance of this improvement with the cessation of anti-inflammatory medications, as long as there are no dietary transgressions, does not favor any placebo effect.

Sometimes, we observe in certain autistic children an apparent behavioral deterioration often related to the speed of change in the child compared to a relational and educational environment that has been somewhat fixed due to months or years of stagnation or slow evolution. The child may thus suddenly become disruptive for those close to them, as they require a completely different type of care. However, this disruption should also be related to the frequency of addictive behaviors towards the foods responsible for some of the observed autistic disorders. These "cravings" seem to support the "opioid hypothesis"—even if addiction is not synonymous with opioid.

The improvements sometimes observed by parents, who are the first concerned, during these diets should encourage researchers to look more closely at the question. If the studies are not robust enough, then let new ones be conducted, with solid, indisputable protocols, as well as sufficient cohorts and durations. It is not so complicated to implement, and many families would be interested in participating in such trials.

On the other hand, all these dietary interventions are considered marginal by various specialties concerned with autism, including child psychiatrists, geneticists, neurologists, and imaging specialists. However, they represent a subject of study in its own right, framed within the general context of "the action of nutrients on brain function." Other stakeholders, such as immunologists or gastroenterologists, could also be involved. But to develop high-quality working hypotheses commensurate with the stakes, I believe it is necessary to organize a coordinated reflection among all these specialists gathered together. Given the complexity of the subject, it might be useful to employ some proven methods for managing complex human systems.