"I put my son on this diet nine years ago, and there was no evidence but the evidence of my son's constant diarrhea, constant indigestion," said Estepp. "He also avoided milk. I kept thinking, what 2-year-old pushes away the milk cup?"
Estepp said her son's health improved after she put him on a restrictive diet. It didn't cure his autism, but she thinks his better health led to better behavior.
But parents who want hard scientific evidence before trying the casein-free, gluten-free diets may have to wait a while.
"For many researchers, this would be a lower priority. The problem is there's limited federal dollars, it's an amazing amount of work to get these [treatment] studies done," said Volkmar.
The vast spectrum of behaviors and medical issues within autism only compounds the problem. Autism ranges from severe behavioral and communication problems to cases of highly-functioning but somewhat socially challenged people. Since gastrointestinal issues may be more common in some areas of the autism spectrum but not in others, doctors say lumping children into one autism category to study diet may confuse the results.
"Studying them [patients] in large groups of people who are heterogeneous -- very different from one another -- is not the best way to learn what is going on with these diets," said Dr. Martha Herbert of Massachusetts General Hospital in Boston. "You will wash out any effect in a subgroup by blurring them in with others who have different problems."
The panelists contributing to Monday's report could only find one double-blind placebo controlled study on the behavioral effects of a casein-free/gluten-free diet. The study found the diet had no effect on autism symptoms, but it included only 15 children and lasted for only 12 weeks.
Still, many researchers have found links between behavioral disorders and gastrointestinal problems.
"There is a higher reported rate of GI (gastrointestinal) problems in children with autistic spectrum disorders compared with normal children and with children with other developmental disabilities," said Dr. Shlomo Shinnar, of the Montefiore Medical Center in The Bronx, N.Y. "This has been reported in multiple case control studies."
But a higher rate among children with autism doesn't mean autism caused the intestinal problems or vice versa, Shinnar said.
"At the present time, the higher rate of GI symptoms remains an interesting finding that is worthy of further research but does not provide us with either a cause or a treatment for autism," he said.
Dr. Arthur Beaudet, who contributed to the published statements, said he believes genetic studies may one day lead to better research of gastrointestinal problems in children with autism.
For instance, children with autism symptoms caused by one genetic mutation may be more likely to develop gastrointestinal issues than children who have a different genetic mutation -- or none at all.
Beaudet said current research has already pinpointed subgroups of people with autism who have a specific mutation and distinct behavior, such as the estimated 1 percent of people with autism who have the mutation known as "16p11.2"