NEW YORK (Reuters Health) - While antidepressants are commonly given to people with autism, there is no evidence from clinical trials that the drugs are helpful for children with the disorder, and only limited evidence that they benefit adults, a new research review finds.
The analysis, reported in the Cochrane Database of Systematic Reviews, adds to doubts about the use of antidepressants known as selective serotonin reuptake inhibitors (SSRIs) in autism.
Last year, a U.S.-government-funded study found that the SSRI citalopram (Celexa) was no better than a placebo at improving repetitive behaviors in children with autism. At the time, experts expressed surprise at the lack of benefit and said the results illustrated the need to test antidepressants against placebos in people with autism.
For the new review, researchers evaluated the findings of the Celexa study, along with those of six other -- much smaller -- clinical trials in the medical literature.
Overall, they found no evidence that SSRIs were better than placebos at improving repetitive behaviors or other symptoms in children with autism. And there was only limited evidence from two small clinical trials that certain SSRIs might improve anxiety, depression and other symptoms in autistic adults.
On the whole, there is no basis for recommending the routine use of SSRIs in treating autism, according to the researchers, led by Dr. Katrina Williams, a pediatrician at the University of New South Wales and Sydney Children's Hospital in Australia.
However, the researchers are not recommending that people with autism who are already on an SSRI and doing well stop taking their medication.
As it stands, no medications are specifically approved for treating autism spectrum disorders (ASDs), a group of developmental disorders that hinder people's ability to communicate and build relationships. The conditions range from severe cases of "classic" autism to the relatively mild Asperger's syndrome.
Behavioral and educational therapies that target the social, developmental and communication problems are the mainstay of autism treatment. But SSRIs are often prescribed to aid with certain symptoms; by one estimate, up to 40 percent of children with autism have been treated with an antidepressant.
In the U.S., three SSRIs - sertraline (Zoloft), fluoxetine (Prozac) and fluvoxamine (Luvox) -- are FDA approved for children older than seven.
Part of the rationale for SSRI use in ASDs is that the drugs can be effective for anxiety and obsessive-compulsive disorder, conditions whose features are similar to some behaviors seen in autism. For example, repetitive behaviors -- such as repeating specific words or actions, or obsessively following a routine or schedule -- are a main feature of autism.
In addition, SSRIs enhance levels of the brain chemical serotonin, and serotonin is thought to influence sleep, mood, aggression and other brain processes that are often altered in autism, Williams told Reuters Health in an email.
But few clinical trials have been done to test the drugs' effectiveness in improving the symptoms of children or adults with autism.
Williams and her colleagues were able to find only seven small, short-term trials where people with autism were randomly assigned to take an SSRI or a placebo for comparison.
The Celexa study, by far the largest, included 149 children with ASDs who were given either the SSRI or a placebo for three months. Roughly one-third of the children in each group showed improvements in repetitive behaviors during the study period, with no advantage from the antidepressant.
All of the other studies Williams and her colleagues found were quite small, with the largest including 39 children. None went beyond three months.
Overall, the five studies that focused on children and teenagers showed no benefits of SSRI treatment, according to the researchers; the trials tested the drugs fluoxetine, fluvoxamine, and, in the two oldest studies, fenfluramine -- a medication that has since been pulled from the U.S. market.
Two studies included adults, with one testing fluoxetine and the other fluvoxamine. The trials found improvements in SSRI users' obsessive behaviors, anxiety, depression and aggression versus placebo users. However, the studies were very small -- one included six participants, the other 30 -- and treatment lasted eight to 12 weeks.
Moreover, SSRIs can have side effects, and concerns about adverse effects are greater with children and teens. In the citalopram study, one child given the drug developed seizures that required hospitalization, and continued to have repeat seizures after being taken off the drug. Children on the drug were also more likely than placebo users to show impulsive behavior, sleep problems and difficulty concentrating.
No increased risk of side effects was seen in children given Prozac; the study that looked at Luvox provided little information on side effects, according to Williams' team.
Given the lack of effectiveness and potential for side effects, SSRIs cannot be recommended for children with autism, the researchers say.
For adults, Williams told Reuters Health in an email, there is "preliminary information that suggests effectiveness" for alleviating depression, anxiety, obsessive-compulsive behavior and aggression. Decisions on whether an adult with autism should try an SSRI should be made on a case-by-case basis, according to Williams.
That said, some people with autism currently on an SSRI may be doing well.
"If children or adults are on an SSRI or other antidepressant and it has improved the problem that it was prescribed for and is not causing side effects, they should continue on the medication," Williams said.
Larger, well-conducted trials of SSRIs in the treatment of autism are still needed, according to Williams. That includes studies of other SSRIs that have yet to be put to the test in clinical trials but are being prescribed to people with autism -- such as sertraline and paroxetine (Paxil).
Larger studies, Williams and her colleagues point out, might allow researchers to find out whether certain subgroups of people with autism respond better to SSRIs than others.
SOURCE: link.reuters.com/kat93n Cochrane Database of Systematic Reviews, August 8, 2010.