We've created this blog to help you understand why the recent Trump administration claims about autism are not correct.
On Monday, 22 September, Donald Trump, Robert F. Kennedy Jr. (RFK) and others made a statement from the White House about an autism cause and an autism treatment.
Their first claim suggested that mothers taking paracetamol in early pregnancy increase their chance of having an autistic baby. In the US, paracetamol is known as acetaminophen, or under the brand name Tylenol. This claim is not supported by strong evidence.
They also announced that the Food and Drug Administration (FDA, the agency that regulates medicines in the US) has approved Leucovorin (folinic acid, a form of folic acid) as a treatment for autism, side-stepping a usually lengthy and rigorous approval process. However, the FDA gave a supporting statement that it would only be approved for Cerebral Folate Deficiency Syndrome (CFD). This is because it is not supported by high-quality research evidence for autism.
Key messages
If you are pregnant, you should follow NHS guidance and the advice of qualified medical practitioners. Do not change your behaviour because of these headlines.
There is no rigorous research evidence supporting the idea that any medicine or supplement causes or prevents autism.
The best quality research into paracetamol and pregnancy found that using paracetamol as recommended during pregnancy does not lead to an increased likelihood of having an autistic baby, a baby with ADHD, or a learning disability.
There is no robust evidence to support claims that taking Leucovorin can treat autistic traits. Existing research on this is small-scale with mixed results.
Useful terms for understanding research
Before looking at some expert responses, you might want to familiarise yourself with two useful terms for understanding research: correlation doesn't mean causation and 'robust' research.
Correlation doesn't mean causation
You might have heard the term 'correlation doesn't mean causation'. But what does it mean?
Just because the occurrence of 'A' increases alongside 'B', it doesn't mean that 'A' causes 'B'.
For example, one can correlate the reduction in the number of pirates in the world to the increase in climate change. There are far fewer pirates these days than there were 150 years ago, but this doesn't mean climate change is caused by a 'lack of pirates'.
What makes a study 'robust?'
A robust study would have several components, including the following:
-It has a 'rigorous' design. This means a methodical process that can isolate what researchers are trying to test, without accidentally including other variables that could affect the results.
-It has a large sample size. This means the research was conducted on a large number of people, so it is more likely to reflect the population and reduce room for error. Smaller studies can help researchers decide if something is worth further research. However, we should be wary of generalisations from small studies.
-The study is replicable. You can't generalise based on the results of one study. However, when different researchers repeat similar studies and get similar results, that helps build the 'evidence base'.
Paracetamol, pregnancy and autism - a clinican's perspective
Dr Linden J Stocker, a consultant in fetomaternal medicine at University Hospitals Southampton, said:
Paracetamol has been well-studied in pregnancy, and there is no good evidence that it harms the unborn baby. It is also one of the longest-used painkillers we have used in pregnancy, and therefore the data we have on this drug is good.
Some studies suggest that paracetamol use in pregnancy might be linked to behavioural problems in children, but these studies have unreliable findings. Overall, there is no proof that paracetamol exposure in the womb affects a child’s behaviour.
Moreover, paracetamol has been used for many years to help women with pain in pregnancy. Suggesting to pregnant women that this drug is unsafe will cause much anxiety to women who already feel vulnerable and overwhelmed with information. This yet again feeds into the rhetoric that women should be deprived of a well-tested and much-needed medication in pregnancy.
Essentially, she's telling us that there has been lots of research on paracetamol during pregnancy, and plenty of good research that shows it is safe.
Linden adds that telling pregnant people not to take paracetamol could cause them unnecessary stress and could lead to them not taking pain medication when they need it.
Paracetamol, pregnancy and autism - a researcher's perspective
Next, let's look at a statement from Dr Monique Botha, an autistic researcher and associate professor in social and developmental psychology at Durham University. They said:
There are many studies which refute a link, but the most important was a Swedish study of 2.4 million births (1995–2019) published in 2024, which used actual sibling data and found no relationship between exposure to paracetamol in utero and subsequent autism, ADHD or intellectual disability. This suggests no causal effect of paracetamol in autism. This is further strengthened by there being an absence of a dose-dependent relationship.
Dr Monique Botha
Monique, and many other researchers, are referencing a 2024 study by Dr Victor Ahlqvist, and his colleagues. This is considered the most robust study into paracetamol during pregnancy and neurodivergence.
In this study, Victor and his colleagues examined health data in Sweden, which has thorough health records. They studied 2.4 million births (a huge sample size).
When Victor and his team compared data between siblings, it helped them account for genetics. For example, compare if a parent has two children and takes paracetamol during one pregnancy and not another. This sets this study apart from previous studies that implied there was a link, which didn't compare data on siblings.
We know from research evidence that genetics is the biggest factor determining whether a child is likely to be autistic. We also know that autistic people have sensory differences and can be more (or less) sensitive to their environments, or things like pain. Previous studies that link paracetamol during pregnancy and autistic babies might not account for genetic factors, ie the expecting parent might be autistic themselves, so could be more sensitive to pain and more likely to take paracetamol.
Then, Monique says that there isn't a 'dose-dependent relationship'. A dose-dependent relationship is the idea that if A causes B, taking more of A means that B is even more likely. So, if it were true that paracetamol in pregnancy increases the chance of a parent having an autistic baby, it would make sense that if a pregnant person has taken lots more paracetamol, it's even more likely their future baby is autistic. But the data does not support this.
Leucovorin - a researcher's perspective
More evidence is needed with regards to the effect of Leucovorin and core autistic traits before any inferences can be drawn in any meaningful way. Any evidence available at this point is exceptionally tentative and would not be considered to be robust. Similarly, whilst medications may help with very specific aspects, there is no medication or treatment that actively cures or erases autism, though it might adjust behaviour, or reduce co-occuring symptoms which contribute towards distress for autistic people.
Autism is a lifelong heritable disability whose primary cause is exceedingly likely to be genetic, expressed through a wide array of genes. Similarly, autistic people are exceptionally heterogeneous so any treatment or medication for specific traits is likely to work for very specific presentations of autistic traits, in very particular contexts.
Here, Dr Monique Botha states that there isn't enough evidence to use Leucovorin to 'treat' autistic traits. There are some small studies suggesting a link in some cases, but it's far too early to tell, and these early studies aren't meeting the standards of 'robust' that we mentioned earlier.
Early research can suggest all sorts of things, which can be supported or contradicted with further research, like the example above with paracetamol. It's not a good idea to approve drugs based on small-scale early research.
Monique also says that autism is 'heterogeneous', meaning it's really varied and can look completely different in two different people. Essentially, they are telling us to be wary of anyone claiming one environmental factor could reduce autistic traits.
If you are to take away one thing from this blog, it's Monique's closing remarks:
Sweeping statements about cures or treatments do not tend to be accurate, helpful, or ethical.
Final thoughts
When it comes to paracetamol, pregnancy and autism, the best research finds no link. And current research into Leucovorin is in exceedingly early stages, so it's best to remain sceptical.
Finally, we want to address every autistic person reading this: you are valid and worthy. Your autistic mind is not something that needs to be cured or prevented. You don't need to change, but the world does. And we won't stop until every autistic person can live a happy, healthy and long life.
Sadly, we don't think this is the last we'll hear on this topic from the Trump administration, so be gentle with yourself. If you feel motivated to fight back against misinformation, join the Autistica Network to take part in and influence autism-affirming research, or donate today to support our work. Together, we can fight misinformation and create a more inclusive world.