I am beginning to feel a real sympathy for Sisyphus, who was punished for his sins by the ancient Greek gods and condemned to roll a stone uphill for all eternity. Sisyphus’ sin was to imagine he could use trickery to outsmart the gods and cheat death. The punishment for his hubris was an eternal life of toil. My sin was to imagine that once you have *refuted* a stupid argument about autism you can move on and forget about it. Alas, no. The stupid, it burns. And when it’s not burning it smoulders. So, in the months since my last post, all manner of myths about autism have refused to die. Even the most unlikely embers can burst into flames. Then we have to have the argument all over again. We will never convince the true believers. But new people turn to the internet for information all the time and false information has to be challenged.

This happened with a recent story on the BBC website about the building of a new school for autistic children. The piece described how the building was being made autism friendly by accommodating the sensory needs of autistic children. It also discussed the reasons for creating fresh specialist provision at a time when the emphasis is on inclusion and  educating all children in the mainstream and briefly referred to possible reasons why more boys are diagnosed than girls, especially at the more able end of the spectrum. It was a very good article. It faced up to the disabling effects of autism without demonizing or scaremongering. At the same time it showed how society can lessen the impact of autism by making adjustments to meet people’s needs. This was achieved by designing an autism friendly school. Architects take note. Getting rid of harsh lighting and bright, shiny surfaces and replacing sharp corners with curves is not just autism friendly. It is people friendly.

When I turned to the comments section it had been closed. But not before a supportive and intelligent discussion had been hijacked by those who insist that vaccines cause autism. I am not going to tackle that particular canard right now. It will never go away. At least we have succeeded in limiting its scope and influence so it no longer attracts mainstream support. But tagged on to the end of the discussion another ember was rekindled, namely that there is an association between autism and acetaminophen (more commonly known as Paracetamol in the UK or Tylenol in the USA). This is a spin off from the vaccine myth that has its origins in the use of Paracetamol to treat localized irritation or fever, a common side effect of the MMR vaccine. So instead of the vaccine causing autism it is the medication taken after the vaccine!

I first encountered this hypothesis in an article by Schultz et al. in 2008. It was published in a respectable journal, Autism, that is supported by the National Autistic Society. But it is a weak article that drew this critical response. Schultz recruited parents from an online autism discussion group and asked those parents to recruit the controls from amongst their acquaintances. This was a recipe for massive selection bias. In addition the sample size was small. No child was examined. No diagnosis or case history was independently verified. Everything was done by parental questionnaire. I checked PubMed for any follow up. Schultz published two further articles in obscure journals. In one,  a five page discussion of the literature he offered the following.

The role of acetaminophen (paracetamol) in an increased risk for asthma is described and a possible similar link to an increased risk for autism is suggested.

Another five page discussion of the literature ends with this hypothesis.

The purpose of this report was to explore a possible correlation between acetaminophen and autism which acts through activation of the cannabinoid system. If this hypothesis is correct, it opens new avenues of investigation for possible autism treatment including
agonists and antagonists of the CB1 and CB2 receptors.

And finally, when he did join a serious study, using genetically modified, socially impaired mice, they found that acetaminophen actually improved sociability which is often impaired in autistic individuals who have difficulties with social understanding.

Hence, it appears that other indirect actions of acetaminophen, including 5-HT receptor agonism, may underlie its sociability promoting properties outweighing any CB(1) mediated suppression by locally-elevated endocannabinoids in these mice.

So there we have it. A very tenuous link between paracetamol and autism was suggested five years ago. Two years later a possible biological pathway via activation of the cannabinoid system was proposed. But when this hypothesis was tested last year the benefits of paracetamol were deemed to outweigh the remote possibility that it might be connected to autism.

However, it was the original study that made the news. It had power in the public mind because it was about parents telling their story. It was easy to understand and easy to believe. So, five years later, it is still making the rounds, even though there is no science to support it and what evidence there is actually contradicts it.

* I use the word “refute” in the old sense of providing compelling evidence against an argument and not the weaker modern sense where “refute” is used as a synonym for “deny”.*