Diet Free of Gluten and Casein Has No Effect on Autism Symptoms

Daniel M. Keller, PhD

May 24, 2010

May 24, 2010 (Philadelphia, Pennsylvania) — A gluten-free, casein-free (GFCF) diet or challenges with these food substances did not alter sleep or activity patterns in preschool children with autism spectrum disorder (ASD) who were also receiving intense behavioral therapy, suggests the first study to control for nutritional sufficiency and other interventions.

Slight differences in social language, approach, and play that were seen at 2 hours after gluten or casein exposure were not apparent at 24 hours, lead author Susan Hyman, MD, chief of the Division of Neurodevelopmental and Behavioral Pediatrics and associate professor of pediatrics at the University of Rochester in New York, reported here at the 9th Annual International Meeting for Autism Research.

Although dietary interventions are often used with children with ASD, have a popular image among the public, and result in anecdotal reports of improvement, prior trials have not borne out such positive outcomes. Dr. Hyman explained that she and her colleagues therefore designed a study to test whether a commonly used dietary intervention was safe and effective.

Study Population Stable at Baseline

Researchers recruited 22 children (age, 30 - 54 months) who were very consistent in their clinical presentations (positive on the Autism Diagnostic Interview and the Autism Diagnostic Observation Schedule), their medical conditions, and the therapies they were receiving, which was an early intensive behavioral intervention program. "This is important because if you're changing other parameters, you want to have other effective treatments stable," Dr. Hyman said. Children were excluded from the study if they had celiac disease, food allergies, or deficient iron stores.

The investigators formulated and monitored a nutritionally sound, strict GFCF diet, which they maintained children on for a minimum of 4 weeks. A staff of dieticians worked with the families to identify a food that their child would eat and that could be formulated to be indistinguishable with or without the test ingredients.

Fourteen of the children were able to maintain the diet and allow data collection. They remained on the diet and were observed and then challenged with the food substances (20 g wheat flour, 20 g evaporated milk, both, or placebo) only if they were at their behavioral baselines. Challenges were administered in a randomized, double-blind fashion. Each child received a food challenge on 3 separate occasions over 12 weeks.

To ensure nutritional adequacy, laboratory monitoring, body mass index, weight, and growth recording occurred at baseline, 6, 8, and 30 weeks. The researchers also collected behavioral data at these times, as well as the day before and 2 and 24 hours after each food challenge.

No Difference in Activity Levels After Dietary Challenge

Dr. Hyman reported that there was no difference in the length of sleep recorded by parents over the course of the study before and after challenges and compared with baseline. There were also no changes in the number of night wakings or in the number or consistency of stools.

Compared with placebo challenges, no significant differences occurred in length of sleep or waking with gluten (P = .21 and P = .10, respectively), casein (P = .48 and P = .15, respectively), or both (P = .99 and P = .18, respectively). Similarly, there were no differences in stool consistency compared with placebo.

Children's activity levels recorded by parents, researchers, or applied behavior analysis program teachers did not differ after placebo, gluten, casein, or gluten/casein challenges. These observations were consistent with recordings from actigraphs — watch-like devices that measure activity.

Dr. Hyman noted that these measures are not specific to autism. Thus, the play-based Ritvo-Freeman Real Life Rating Scale for autism was used to gauge sensory motor behaviors, social approach, and language. "With correction for multiple comparisons, there was no difference with the challenges compared to placebo, and there was no difference with introduction of the diet," she said.

To see whether any individual responses were obscured by group statistics, the researchers examined the single subject data but did not identify any child with significant effects after dietary challenges or who had improvements in core features of autism during the trial.

In summary, Dr. Hyman said, "The data that we have do not demonstrate effect of the GFCF diet on the behaviors we measured." However, she said that study limitations include the study's small size and that all the included children were in an effective early intervention program (≥10 hours/week), were of similar age, and were all stabilized on a monitored diet. Furthermore, none of the children was iron- or vitamin D-deficient.

Dr. Hyman said a question remains whether any autistic children could respond to the diet used in the study. For example, children with celiac disease or bad gastrointestinal symptoms were not included. "So could it be that children who have more significant [gastrointestinal] symptoms are the ones that drive the anecdotal reports?" she asked. Another possibility is that foods designed to exclude gluten could also then lack food preservatives or dyes, which is another open question.

Dr. Hyman concluded, "The data that we have do not offer support for the [GFCF] diet in young children who carry a diagnosis of autism and who are receiving other effective behavioral and educational interventions." She cautioned that these data should not be extrapolated to any child with food allergies or intolerances or other gastrointestinal problems, and that "any child who is on the diet needs to be monitored from a nutritional standpoint to make certain that all of the things that we know about typical child development are monitored for."

Jonathan Green, MD, professor of child and adolescent psychiatry at the University of Manchester, United Kingdom, commented that "studies of dietary interventions like this are extremely difficult to do." He calls himself "an interventionist" and leads the Medical Research Council preschool autism communication trial, currently the largest intervention trial internationally in this subject area.

"The [University of Rochester] study is of significance even though sample size is really small, but they really took a lot of trouble to blind the dietary intervention, and that's the really difficult thing to do," he said. He also commended Dr. Hyman's rigor in recording even what she called "oops events," where the child got a bit of food that was not planned, such as a cookie from grandma.

Dr. Green said that although there are hundreds of foods and ingredients that could be tested, he thought that Dr. Hyman addressed well 2 of parents' concerns by testing gluten and casein. "She's done the right test. She's used the right kind of methodology, which is really difficult on a small group of kids, and her results are pretty clear," he said.

Addressing the possibility that an autistic child with a preexisting gut problem would feel better on a gluten-free diet, he warned, "That, however, does not mean it's having an effect on the autism itself, and that's the point of what Dr. Hyman did.... What she's suggesting is that the diet in itself doesn't have a specific effect on autism as such." He said this kind of information should reach parents, who should see that autism researchers take their concerns seriously, and who thus need to believe the science.

In Dr. Hyman's opinion, "The real future of autism treatment is going to be informed by science. It's going to be informed by what we really do know about the brain and the designer interventions," she said. "What we have now in terms of intervention is empiric observation."

Dr. Hyman and Dr. Green have disclosed no relevant financial relationships.

9th Annual International Meeting for Autism Research: Abstract 140.007. Presented May 22, 2010.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.