ARI Publication 40 – 2013 Version

Understanding Research

We still need a lot more research on the causes of autism and how to treat it. However, we believe
that there is now enough evidence for the treatments listed in this Summary that it is reasonable to
consider them as treatment options, since they are likely to be beneficial to some individuals with
autism and generally have a low risk of adverse effects.

In this Summary we cite over 150 published research studies. In a few places we also cite some
unpublished work which we think provides additional insight, but is not as rigorous.

Placebo-controlled treatment studies

For treatment studies, the highest quality studies are randomized, double-blind, placebo-controlled
studies. This means that the researchers took a group of subjects and randomly assigned half to a
treatment and half to a placebo (no treatment), but neither the participants nor the researchers
knew who was in each group. Evaluations are done at the beginning and end of the study, and
then the code is broken to determine if the treatment group did better than the placebo group.
“Single-blind” studies are similar in design and quality, except that only the evaluators (and
not the participants) are blinded as to which group they are in.

Open Label treatment studies

An “open label” study is a lower quality study. It means that everyone receives the treatment, and
knows that they received it. Unfortunately, due to the “placebo effect,” behavioral scores usually
improve just by the hope involved in being in a study, and so it is usually unclear if improvements
from open-label studies are real or imagined. Open label studies are useful to demonstrate safety,
to determine the type of possible benefit, and sometimes they are used to gather data on changes
in medical test results (biomarkers). If an open label study has very promising results, it is then
useful to follow up with a placebo-controlled study.

ARI Survey Data

Survey data is subject to both the placebo-effect and possible biases in collecting the data.
However, an advantage of the ARI Survey of Parent Ratings of Treatment Efficacy and Safety is
that it includes responses from over 27,000 families, and it involves comparisons of many different
treatments (which is rarely done in traditional studies). So, although there is some “placebo effect”
such that benefits may be somewhat over-estimated, it can provide useful insight into potential
relative benefits of one type of study vs. another.

Individuality

It is important to remember that autism is a heterogeneous condition, and recent research suggests
there are probably several subtypes of autism. There are likely different genetic and environmental
factors associated with each subgroup. So, it is unlikely that any one treatment will apply to
everyone with autism. However, most of the treatments here will probably help many children and
adults with autism. In some cases medical testing or nutritional assessment can be very helpful as
a guide as to whether or not a treatment is likely to be helpful.

Pediatric Reference Ranges

In interpreting personal lab results for children, it is important that the lab use pediatric reference
ranges, not adult reference ranges. Otherwise, the test results are usually invalid and misleading,
and it is better to not do the test. It is very important to check this, as many labs do NOT have
pediatric reference ranges. Ideally, the lab should have multiple pediatric reference ranges, as a 3-
year-old is very different from a 16-year-old for most medical tests.

7