MONDAY, April 4 (HealthDay News) -- Little evidence supports the use of most medications for treating autism in children, with the exception of the anti-psychotic drugs risperidone and aripiprazole that have shown success in controlling severely disruptive, hyperactive and repetitive behaviors, a new study shows.

However, anti-psychotics can have significant side effects, including substantial weight gain, sedation and tremors or rigidity, and should only be used when absolutely necessary, the researchers added.

"The medications are efficacious, but they have significant side effects that should limit their use to patients with severe impairment or at risk of injury," said study author Melissa McPheeters, co-director of the Vanderbilt Evidence-Based Practice Center in Nashville.

Other medications sometimes prescribed for children with autism -- including selective serotonin reuptake inhibitors (SSRIs) to treat repetitive behaviors and stimulants used to treat hyperactivity and impulsive behavior -- have scant evidence supporting their use, according to the study. It was published online April 4 in the journal Pediatrics.

A randomized, controlled trial on SSRIs showed no benefit in controlling repetitive behaviors, according to the study. And while one study did show that stimulants may help with hyperactivity, far more research is needed.

A second study in the same issue of the journal analyzed seven randomized controlled trials on secretin, which is used to treat peptic ulcers, and the study found no benefit in treating autism symptoms and that it wasn't even worth further study. (Animals studies have showed secretin affects the central nervous system, which sparked the interest in using the drug for children with autism, but those hopes have not been borne out, the researchers said.)

A third study in the journal looked at 34 studies on behavioral interventions in preschoolers. The research generally shows such interventions help children improve their language and social skills, but most studies were poorly designed, and more research is needed, the authors of the new study said.

"We see a real variability in response, with some subgroups having dramatically positive effects and some children remaining very impaired," McPheeters said.

An estimated one in 110 U.S. children have an autism spectrum disorder, which is often marked by problems with language, social and communication skills, and repetitive behaviors.

But the condition can vary widely among individuals, with some people never developing the ability to speak and others going on to college.

"This trio of review articles is important in that they remind parents and professionals how difficult it is to effectively treat children with autism, how little research there is to support some of our medical and developmental interventions and how daunting it is to conduct well-designed clinical research effectively," said Dr. Andrew Adesman, chief of developmental and behavioral Pediatrics at Steven and Alexandra Cohen Children's Medical Center of New York.

Children with autism also often have other conditions, such as aggressive or disruptive behaviors, seizure disorders and hyperactivity. No medications target the communication issues or problems with social functioning that define autism itself, experts said. Instead, physicians prescribe drugs to try to ease those other conditions.

"The drugs are used to treat co-morbid symptoms, not the core autism symptoms," McPheeters explained.

In the review, the researchers looked at studies conducted between 2000 and 2010 on the use of various drugs in children aged 12 and under with an autism spectrum disorder. The study on medications included nine studies on anti-psychotics, five on SSRIs and four on stimulants.

While the evidence for anti-psychotics is strong, far more research is needed on newer anti-psychotic drugs that may have fewer side effects, McPheeters said.

As for the early behavioral interventions, "we have a small number of studies that suggest positive outcomes, but we are in critical need of replication, extension and controlled studies," McPheeters said.

She added: "Even though the paper says there is low and insufficient strength of evidence, that should not be interpreted as the interventions are not effective, but as encouragement for additional research."

The three studies were part of an 18-month long project done in conjunction with the Agency for Healthcare Research and Policy to review all of the evidence on treatments for autism spectrum disorders for children 12 and under published after 2000, said McPheeters, the senior author of the report that will be published soon on the AHRQ Web site.

That report started with more than 4,000 abstracts and was narrowed down to 183 studies in all

For the medication study, studies had to include at least 30 people, while for the behavioral study, studies had to include at least 10 people.

"The report includes studies that are very rigorous and also studies that we rated to have poor quality," McPheeters said. "We felt like it was important to make sure we covered the waterfront so that people could see what all was out there."

Another challenge for physicians is determining which children will respond best to which treatment, said Dr. Eric Hollander, director of the Compulsive, Impulsive and Autism Spectrum Disorders Program at Montefiore Medical Center in New York City.

"There is no single treatment that works for all patients," Hollander said.

More information

To learn more, visit Autism Speaks.