You are here:
-A A +A

How do you know if a treatment will work?

Marina Sarris
Interactive Autism Network at Kennedy Krieger Institute
June 20, 2016

Photo of Josh the horse, courtesy of Elisabeth Mekosh/Tara Katherine Photography LLCWill this work? Many parents wonder that before investing in a new autism therapy. A range of activities have been promoted as treatments for autism, from behavioral interventions to swimming with dolphins. Some have years of scientific research showing that they have helped a significant number of people. Others have little or no research to support their claims.

So Dr. Robin L. Gabriels' recent study of therapeutic horseback riding for autism comes as welcome news to people with questions about riding. Her team conducted a large randomized controlled trial, a type of study considered to be scientifically rigorous. In this study, children who took part in therapeutic riding showed more improvements in behavior and speech than children who didn't ride.

Of course, one study does not mean that a treatment is necessarily proven. Science requires that studies be repeated, preferably with larger numbers of people, to make sure that the results were not caused by chance or by something else. Dr. Gabriels is continuing to study the topic and looking for better ways of rating changes in behavior. Her previous study relied on behavior reports from parents, who knew whether their children were in the riding or control group. That knowledge could have influenced their ratings, the study said. (The speech ratings came from a therapist who did not know whether the children rode.)

How do you decide if a treatment has scientific evidence to show it works? One objective place to look is the website of the U.S.-based National Autism Center, a nonprofit organization that publishes the reports of the National Standards Project. People "are constantly bombarded by a massive amount of confusing and often conflicting information" about treatments, the website explains.1 The National Standards Project examines the research and classifies interventions as established, emerging or unestablished for children and adults.

An intervention may move from unestablished to emerging or established over time, as scientists conduct more or better research on it. Examples of established (proven) treatments for children and young adults include behavioral interventions, activity schedules, Applied Behavior Analysis, Pivotal Response Treatment, social skills intervention, and social stories, according to the National Standards Project.2 It classified the larger category of "animal-assisted therapy" as "unestablished" its 2015 report. Therapeutic riding was not evaluated by itself. Another organization that rates treatments, the charity Research Autism in England, was unable to rate animal therapies.3

Examining the research can help parents choose therapies and treatments. But even established or proven interventions do not work for everyone with autism spectrum disorder, according to the National Standards Project. "We wish it were that easy," its report concludes.3

To learn about Dr. Gabriels' research, see our article, Something About a Horse: Finding Benefits in Therapeutic Riding.

The Cochrane Collaboration in London provides reports on some autism treatments, including medications and early interventions, at

Have you or your child tried therapeutic riding? Please give us your opinion by taking our nonscientific poll.

  1. National Autism Center. National standards project. Retrieved from
  2. National Autism Center. (2015). Findings and conclusions: National standards project, phase 2.Randolph, MA: National Autism Center.
  3. Research Autism. Our evaluations of autism interventions, treatments and therapies. Retrieved from

Photo credit: Elisabeth Mekosh/Tara Katherine Photography LLC