autism intervention can seem like a maze to parents


Picture this scenario. You walk out of the paediatrician’s rooms in a daze, after hearing that your child has an Autism Spectrum Disorder. You go home, log onto the Internet and look up the ‘official’ websites, which all say something like this: ‘Autism is a life-long disability. There is currently no known cure’. But then you start searching further and discover seemingly dozens of websites offering treatments which they claim can ‘cure’ or ‘reverse’ autism. What’s going on? Are these claims to be believed? Who’s right?


Unfortunately the official view is much closer to the truth. There is no ‘miracle cure’ for Autism. The very fact that there are so many and varied therapies is testimony to that. As Joe’s paediatrician succinctly put it, ‘if there was a cure for autism, there would only be one cure.’ This quote probably sums up the current situation best:


When you have many medical therapies, often contradictory, two possibilities exist. One is that the correct therapy for that disease has yet to be identified. The history of medicine is replete with incidents in which multiple, unsatisfactory therapies are offered until the single effective treatment for a single disease is discovered. The second possibility is that there is no such thing as one medical treatment because Autism is not one disease after all, but a syndrome, a final common pathway of many, many different diseases, each of which may require its own medical therapy. In such cases accurate diagnosis must precede therapy.1


But let’s not be too pessimistic. Researchers throughout the world are strenuously working to improve our understanding of Autism and to search for more effective treatments for this perplexing condition. The future of a child diagnosed today is much more promising than that of a child diagnosed only 10–15 years ago. And while there may be no quick fix there is certainly a lot you can do to help your child right now.

Remember that Autism Spectrum Disorder is a spectrum — a spectrum which ranges from people with a significant disability (low functioning Autism) to very high functioning people who can cope pretty well in the world with little or no support. There are many examples of able people with Autism Spectrum Disorder going on to lead successful adult lives and, whilst Autism may remain a life-long condition, for these people at least, it is not much of a ‘disability’.


Current interventions aim to reduce your child’s symptoms and equip them with new abilities, at very least to reduce the level of their disability and at best to give them the opportunity to live a near normal life. This is not something that can be achieved overnight. It will probably be a long and difficult journey which may require a multi-pronged approach of early intervention, speech and/or occupational therapy, medication and, for some people dietary changes and complementary medicines.


With this sort of approach a very few lucky children will improve so much that their symptoms will all but disappear. These children are sometimes referred to as ‘recovered.’ Maybe your child will be one of the fortunate few, but the majority will not. Whatever the case, please don’t lose sight of how much they do improve and learn, become ‘more able than disabled’, as you pursue the ‘Holy Grail’ of recovery.


The main way we can help our children is through early intervention. In this chapter we will talk about how medicines can be used to control some of the more disabling symptoms of Autism Spectrum Disorder and related conditions, and then delve into the more controversial area of complementary and alternative medicines, the so-called biomedical approach. But first, evidence-based treatments, what are they and how you can use them to maximise the effectiveness of your child’s intervention program?


What are evidence-based treatments?

Imagine another scenario. You speak to a well-respected paediatrician, an expert in their field. When you ask about a complementary therapy for Autism Spectrum Disorder they say, ‘There is no evidence to support its use.’ You visit a biomedical practitioner anyway and they say there is ‘lots of research’ to support this therapy. Who’s right this time? Well, it might interest you to know that, technically, they both could be. It’s how you interpret the word evidence.


There is mind-boggling amount of research out there about Autism Spectrum Disorder but not all of it is of equal value when it comes to evidence of treatment effectiveness. It’s all about trying to establish cause and effect. The hierarchy of evidence runs from the weakest (the anecdote) to the strongest (the randomised, double-blind, placebo-controlled clinical trial).


The problem with Autism Spectrum Disorders is that the most emotionally powerful stories for parents often are the anecdotal. Anecdotes are great because they can personalise the stories of the various interventions. But they are not considered good evidence.


For example, say you read a story of a mother whose child starts talking after going on a gluten-free casein-free diet. You’d be tempted to rush off to the health food store there and then in the hope that the same thing happens with your own child. But just because B (child starts talking) follows A (starting of GFCF diet) doesn’t definitely mean that A caused B. It might suggest that possibility, but it doesn’t prove it, because we haven’t controlled for all the variables. It could be that the child was just ready to talk (many children with Autism Spectrum Disorder do learn to speak without intervention), or it could be a result of the child’s other educational interventions and/or speech therapy finally bearing fruit. We really can’t be sure.


So when a specialist says there is ‘no evidence’ what they are really saying is that there is no evidence of the quality that can really establish cause and effect. So let’s look at the different components of a randomised, double-blind, placebo-controlled, clinical trial to see why these trials are the most powerful research tools we have.


A clinical trial is a research study conducted with patients which tests out a medicine or other intervention to assess its effectiveness and safety.2
This is an important distinction. A lot of the research published about Autism Spectrum Disorder is actually not in the form of clinical trials but descriptive (or observational) research. Examples might include a parent survey looking at the frequency of gastrointestinal symptoms in children with Autism Spectrum Disorder, or a study which measures the blood levels of the neurotransmitter serotonin in Autism Spectrum Disorder children at a single point in time. This sort of research is absolutely vital but does not provide any evidence of cause and effect; it is a means of generating theories to be tested in future clinical and other research.


Unfortunately this sort of research is often reported uncritically by the media and afforded much more significance than it might actually merit (think about all the ‘major cancer breakthroughs’ you’ve heard about but yet we still haven’t eliminated cancer). Each of these individual studies is a tiny piece in an enormous puzzle, a very small step towards improving our understanding of Autism Spectrum Disorder, but by no means the ‘answer’.


A controlled clinical trial is a study testing a specific medicine or other treatment involving two (or more) groups of patients with the same disease or condition. One (the experimental group) receives the active treatment (that which is being tested) and the other (the control group) receives an alternative treatment, a placebo, or no treatment.


Control groups are needed to determine whether the intervention actually caused the observed change and to rule out other possible explanations. For example, if a group receiving intervention X improves you might conclude that intervention X is effective. But if the control group (who did not receive intervention X) also improves to the same degree, something other than X was responsible. This was the problem highlighted with the TEACCH research, discussed earlier. There was no control group so we cannot say definitely that the IQ gains reported were as a result of the TEACCH program or just a consequence of natural progression.

Randomisation is a method similar to tossing a coin to assign patients to treatment groups; the ‘active treatment’ is given if the coin lands heads and the control is given if the coin lands tails (in reality this is usually done by a computer program). In randomised controlled trials each person has an equal chance of receiving the active treatment, which avoids any possibility of selection bias in a trial. Selection bias is where the people delivering the treatment might be (subconsciously) tempted to give the real treatment to the people they think most likely to benefit (or perhaps in the case of children with Autism Spectrum Disorder, those whose parents look the most likely to carefully follow the treatment regimen).

Also, if the number of participants in the study is large enough, randomisation should ensure that the active and control groups are well matched for any other factors which could affect the outcome, such as age, IQ etc.

A placebo is a dummy or inactive treatment which is given to the control group and must be indistinguishable from the treatment being tested. Giving a placebo to the controls ensures that any changes we observe in the treatment group are not simply the result of their beliefs and expectations or the extra attention involved in taking part in the study, or other unexplained variables.

The best quality trials are double-blind, meaning both patients (or, in the case of children with Autism Spectrum Disorder, their families) and researchers are unaware of which treatment group a patient has been assigned to until all the research results have been collected and analysed. (In a single-blind trial only one of these groups — usually the patients — is blinded to which intervention they are receiving). Blinding is important even if you have an objective outcome measure (such as a reduced cholesterol blood level with a cholesterol-lowering medicine) but in the case of Autism Spectrum Disorder, where you are going to be looking at subjective outcomes, such as changes in behaviour, blinding is even more essential.

Why is blinding so important? Because when people participate in research their beliefs and expectations that the treatment will work can influence the results of the research quite significantly. This is commonly referred to as the placebo-effect. Even the researchers conducting the trial can fall victim to the placebo effect, although in their case the correct term for it is observer bias. This is not meant to imply that they set out to be deliberately biased when they report their results. It just makes sense that if you have a lot financially and emotionally invested in the research you are more likely to believe that you can see a treatment effect.
So, in a randomised, double-blind, placebo-controlled trial (RCT) the only difference between the two groups is the type of treatment they receive. If the outcome between the two groups turns out be significantly different it is reasonable to conclude that this difference is caused by the treatment — cause and effect.

Non-blinded (or open label) studies have actually been found to overestimate the effects of a treatment by about 17%. The effect of randomisation is even greater; analyses have shown that non-randomised studies may overestimate treatment effects by up to 40%!3 So, in theory, an unrandomised, open-label study could overestimate the effectiveness of a treatment by over 50%. That’s enough to make an ineffective intervention appear effective.

It is still occasionally possible to get a false positive (ie showing that a treatment is effective when it isn’t) or false negative result (not showing a treatment is effective when it is) in a RCT; the latter is especially likely if there are only a small number of patients enrolled in the trial. The trouble with most clinical trials in Autism Spectrum Disorder is that they do tend to be quite small. That’s why we need the results to be replicated (or reproduced) in more than one trial, ideally including patients of different ages and severity (ie high- and low-functioning) so we can see if the treatment is effective in a larger group of people with Autism Spectrum Disorder. If a result is replicated several times over, and these studies are published in reputable peer-reviewed journals, you can be confident you are seeing a real treatment effect (or not, as in the case of secretin, discussed below).

To overcome the problem of small patient numbers in clinical trials, sometimes researchers perform systematic reviews or meta-analyses. These are the absolute top of the evidence hierarchy. In a meta-analysis data from several RCTs are pooled together and the results analysed. The larger patient numbers means you may be able to pick up a treatment effect that was not evident in the individual small trials. Only good quality RCTs should be included in meta-analyses, otherwise the results are pretty meaningless.

One of the best known producers of systematic reviews is the Cochrane Collaboration. They have performed a number of systematic reviews of Autism Spectrum Disorder treatments, which you can see yourself at The Cochrane Library. An easy way to access the library is through the Cochrane Collaboration website:
First click on the Cochrane Library link. Then it is fairly easy to find these reviews using the search facility; otherwise you can click on the link to Cochrane Reviews By Topic. The Autism Spectrum Disorder reviews are located under Developmental, Psychosocial and Learning Problems; select Developmental Problems then Autistic Spectrum Disorder. Although these are quite technical documents, each review has a plain English summary of its conclusions. We are fortunate that all residents of Australia can access The Cochrane Library for free, thanks to funding provided by the federal government.

Thus, when that specialist says there is ‘no evidence’ for a particular treatment, it doesn’t necessarily mean there is no research behind it. What it does mean is that we don’t yet have any proof that it is effective. This may be because the clinical research hasn’t been done yet, or sometimes because the research has been done and the treatment hasn’t been found to be effective.

The story of secretin — a case study in evidence-based health care

Secretin is a gastrointestinal hormone; one of the hormones that control and regulate the digestion of food. It became the focus of attention in 1998 when researchers at the University of Maryland published a report of three children with Autism Spectrum Disorder whose Autism symptoms improved after they received a secretin infusion during an endoscopy. Following the publicity thousands of children with Autism received intravenous secretin, resulting in an international secretin shortage.


People started speculating on how secretin could affect the symptoms of Autism. It was thought that secretin may act as a neuropeptide (a type of chemical signal in the brain). Secretin and its receptors have been found in the central nervous system of animals, although the exact role of secretin in the human brain is unclear.


Therefore there were both anecdotal reports that secretin was effective and animal studies suggesting a theoretical mechanism of action.


However, when researchers began conducting RCTs of secretin in children with Autism Spectrum Disorder they were unable to demonstrate any effect on Autism symptoms. In all, over a dozen studies, involving over 700 children were conducted and not one found that secretin was more effective than a placebo.4 Later in the chapter Dr Richard Couper proposes some theories to explain why this occurred.


A Cochrane systematic review concluded that secretin ‘should not currently be recommended or administered as a treatment for Autism’.5 Despite this, some of the parents involved in these trials elected to continue secretin treatment even after being told their child, according to all objective measures, had not responded, illustrating the power of the placebo effect. Secretin is still promoted as a treatment for Autism even today.


Sadly secretin did not turn out to be the ‘cure’ we all hoped. However, scientists are still continuing to investigate the role of neuropeptides in Autism Spectrum Disorder.


Why do we need evidence-based treatments

On a societal level, governments like evidence-based medical treatments because they provide them a more cost-effective way of allocating health spending. Evidence-based medicine is also behind many successful health promotion campaigns.

In Australia, the federal government makes sure that all procedures and medicines it subsidises through Medicare and the Pharmaceutical Benefits Scheme ‘are supported by evidence of their safety, clinical effectiveness and cost-effectiveness’.6 This is as it should be. I’m sure as tax payers you would prefer that your precious tax dollars were not wasted on unproven treatments for say, diabetes or cancer.

The trouble is that the treatment of Autism Spectrum Disorder is not generally known for its good evidence-base. In fact, it is the very lack of evidence which causes some of the biggest arguments. We have claim and counter-claim that a treatment is effective, or not effective, but without good evidence we cannot always be sure who’s right. Also the lack of good evidence makes it easier for governments to justify not funding many Autism Spectrum Disorder treatments. That’s why as parents we need to push for better clinical research.

On a personal level, you might like to use evidence-based principles to help you to decide what treatments to use with your child, especially if your own financial resources are limited. This doesn’t mean you have to completely disregard treatments without good evidence, but we’d recommend you give priority to those interventions with better quality research behind them.

What are the downsides of an evidence-based approach?

Well, in the first place, we cannot always conduct clinical trials of the highest quality demanded by evidence-based practice. We can conduct placebo-controlled trials to test the effectiveness of medicines (conventional or complementary) in Autism Spectrum Disorder but for obvious reasons we cannot have a placebo-controlled trial of an educational intervention. We can have a control group, however, (children receiving an alternative intervention; it would be unethical to have a no intervention group these days) and some of the recent trials of early intervention have even recognised the importance of using randomised controls.

Also, it takes lots of time and money to conduct good quality research, especially if we replicate the research several times over. Lots of parents say, ‘but we can’t afford to wait for the evidence’ and that’s a reasonable view. So, if you decide to use a non-evidence-based treatment for your child, here are a few questions you might like to consider before you commit.


What are the potential side effects or risks?

Does the theory behind the proposed treatment make sense, given what we know about Autism?


Is it individualised, taking into consideration my own child’s behaviours and symptoms?


Is it monitored for effectiveness (based on data) and changes in dose and intensity?


What training and supervision are needed to administer the treatment?


How much does it cost?


Prescribed medicines in Autism Spectrum Disorder

We haven’t yet found a medicine which can ‘cure’ Autism but sometimes medicines can help alleviate symptoms and manage associated conditions, such as attention deficit hyperactivity disorder (ADHD), anxiety, sleep disorders or obsessive compulsive behaviour.* Some medicines which have been prescribed for people with Autism Spectrum Disorder are discussed in Table 1. These medicines appear to work by altering the level or response to important neurotransmitters (serotonin, dopamine and noradrenaline) in the brain.


We need to avoid medicines becoming the ‘easy option’ for managing some symptoms, especially in the case of antipsychotics used to manage aggression, which is more common in low functioning Autism. These medicines should really only be considered if behavioural measures and environmental supports (such as PECSÆ or other visual aids) have failed to bring the problem behaviours under control.


With this in mind, the atypical antipsychotic, risperidone (RisperdalÆ) was recently listed on the Pharmaceutical Benefits Scheme, but with some significant restrictions on its use:
Treatment under the supervision of a paediatrician or psychiatrist, in combination with non-pharmacological measures, of severe behavioural disturbances in a child or adolescent aged less than 18 years with Autism. Behaviour disturbances are defined as severe aggression and injuries to self or others where non-pharmacological methods alone have been unsuccessful.7
However, in other cases medicines may help to reduce problems such as hyperactivity, anxiety or obsessive compulsive behaviours, which prevent a child succeeding at school or slow progress of their other interventions. In these situations a carefully supervised trial of a medicine may be justified.

Joe currently takes the stimulant medicine methylphenidate (RitalinÆ) and whilst I was initially reluctant to go down this path, the benefits for him (improved attention and therefore better performance at school) do seem to outweigh the downsides (some initial insomnia and appetite suppression.

1 Coleman, Mary. Ed (2005). The Neurology of Autism, New York, Oxford University Pres.
2 Bandolier Evidence Based Medicine Glossary Bandolier – Evidence-Based Health Care (2001). Bandolier Bias Guide
3 Levy, SE, Hyman, SL. (2005). Novel treatments for autistic spectrum disorders. Mental Retardation and Developmental Disabilities Research Reviews, 11, 131–142.
4 Williams KW, Wray JJ, Wheeler DM (2005). Intravenous secretin for autism spectrum disorder. Cochrane Database of Systematic Reviews, Issue 3. Art. No.: CD003495. DOI: 10.1002/14651858.CD003495.pub2.
5 Australian Government Department of Health and Ageing. Background to the establishment of the Medical Services Advisory Committee (MSAC) .
6 Schedule of Pharmaceutical Benefits August 2007
7 Jesner OS, Aref-Adib M, Coren E. (2007). Risperidone for autism spectrum disorder. Cochrane Database of Systematic Reviews, Issue 1. Art. No.: CD005040. DOI: 10.1002/14651858.CD005040.pub2.

This button will shut down this information on ABA so you can return to the last page

Click here for the full range of Asperger's and Autism fact sheets at
Click here to read a personal story about early intervention

This information is reproduced with kind permission from a chapter of The Australian Autism Handbook published by Jane Curry Publishing and distributed by Pan Macmillan. It remains under the copyright of Jane Curry Publishing.


This information is reproduced with kind permission from a chapter of The Australian Autism Handbook published by Jane Curry Publishing and distributed by Pan Macmillan.