On Monday 22 September, US President Donald Trump made a widely-publicised announcement that paracetamol (acetaminophen/Tylenol) during pregnancy was confirmed as causing autism spectrum disorder (ASD). The claim – backed by a single, rather dodgy study – brings to a head long-standing concerns about the apparent, well-documented increase in ASD rates. QuickNews dives into the controversy to find out if there is any validity to the claims.
President Trump said, “With Tylenol, don’t take it. Don’t take it. And if you can’t live, if your fever is so bad, you have to take one because there’s no alternative to that, sadly,” adding that other medicines such as aspirin were also “proven bad”.
The announcement had been expected for some time and doctors, scientists and medical organisations were quick to respond. The president of the American College of Obstetricians and Gynecologists stated that the paracetamol–ASD claim “is not backed by the full body of scientific evidence and dangerously simplifies the many and complex causes of neurologic challenges in children”.
At the very least, such an announcement will causing pregnant women to second-guess their taking one of the few over-the-counter pain medications widely regarded as safe during pregnancy. In 2017, the X account for Tylenol stated “We actually don’t recommend using any of our products while pregnant.” But a major pharmaceutical manufacturer would want to protect itself from liability as broadly as possible. The politically-charged nature of the announcement has also seen pregnant women making TikTok videos of themselves apparently taking paracetamol (often with no reason to).
It is generally accepted that ASD is caused by a combination of genetics and environmental factors. About 1000 genes are believed to be related to ASD. And there is a very long list of possible risk factors, with an uncertain risk contribution: “Non-genetic factors mediating ASD risk could include parental age, maternal nutritional and metabolic status, infection during pregnancy, prenatal stress, and exposure to certain toxins, heavy metals, or drugs.”
Study validity questioned
To date, paracetamol during pregnancy had generally been linked by a number of poorly powered studies to a wide variety of outcomes: in addition to ASD, asthma, lower performance intelligence quotient (IQ), shorter male infant anogenital distance (predicting poor male reproductive potential), neurodevelopmental problems (gross motor development, communication), attention-deficit/hyperactivity disorder, poorer attention and executive function, and behavioural problems in childhood. A study of nearly 2.5 million children, the largest and most comprehensive do date, found a slight link for paracetamol exposure and autism – which vanished when controlled for sibling exposure (representing shared environment).
Before President Trump’s announcement, news releases on a review making the link were published some weeks before – QuickNews even covered it. The review, published in Environmental Health, selected certain related studies using the ‘Navigation Guide’ methodology – a non-quantitative methodology for the narrow use of inferring health impacts from environmental toxins, but was nevertheless used for pharmaco-epidemiology and teratology. According to Nathan A. Schacthman, legal counsel for scientific matters, the study has serious conflicts of interest: for example, last author Andrea A. Baccarelli is an environmental epidemiologist who has been involved with a lawsuit against manufacturers and sellers of paracetamol. In that lawsuit, his claims were thrown out on the basis of not having sufficient validity, including cherry-picked data and over-generalisation to distinct disorders (such as grouping attention-deficit hyperactivity disorder [ADHD] and other neurodevelopmental disorders). In addition, the study also made misleading claimed about being funded by the National Institutes for Health (NIH). Finally, although this is not mentioned by Schacthman, Robert F. Kennedy Jr. is also an environmental lawyer.
What’s this about a ‘cure’ for autism?
The bombshell announcement by President Trump came with an another bombshell announcement that there was ‘cure’ for autism. The cure is allegedly leucovorin – which sounds very impressive to the non-medical public. But leucovorin is merely folinic acid, which is a vitamer of plain old folic acid – aka vitamin B9. Folinic acid is on the World Health Organization’s essential medicines list. On the same Monday, the US Food and Drug Administration approved it for the treatment of ASD in children – bypassing the normal review process.
Can a simple medication – or rather, supplement – really ‘cure’ ASD, an extremely complex neurological disorder? The NIH funds about $300 million in ASD research annually, nearly double the amount in 2011. If anything, this has echoes of President Trump’s touting of hydroxychloroquine as a now-discredited cure for COVID in the height of the pandemic. It might indeed be beneficial if a patient had a bout of malaria and COVID at the same time, but was rapidly discredited.
The largest controlled study for the use of folinic acid supplementation plus usual care found only a modest ~1 point increase in the Childhood Autism Rating Scale (CARS) compared to usual care plus placebo.
There are also concerns about potential conflicts of interest. One of the manufacturers of folinic acid, iHerb, had the celebrity heart doctor Mehmet Oz as an investor, and is now the administrator for the Centers for Medicare & Medicaid Services (CMS), served until recently. CMS has however denied that Dr Oz will receive any financial reward from this.
But why are autism rates on the rise?
There are a few good explanations about why the rate of autism diagnoses is increasing, which do not depend on the addition of some new environmental variable. The first and most obvious is that there is increased awareness of this, and more referrals for assessment. A second reason is that the guidelines for diagnosis have become a lot less stringent. The definition of autism diagnoses, unlike schizophrenia, has drifted over time, with more “normal” people being likely to be diagnosed. The DSM-III of 1980 had more stringent criteria, for example, an individual needed to exhibit “a pervasive lack of responsiveness to other people” [emphasis added].
Introduced from 1994, the DSM-IV had broader criteria, and folded Asperger disorder into ASD. With the introduction of the DSM-V, new diagnoses were curbed – for a time. This is because about 20% of the children diagnosed with ASD under DSM-IV-TR would not have received one under the DSM-V. The DSM-V relaxed the criteria for language delay, co-occurrence, and IQ, making it easier for borderline cases to qualify for a diagnosis.
Another underappreciated element is that of social contagion. If parents know a family with a child with ASD, they may be more likely to seek a diagnosis. One study from California showed that ASD diagnoses were more likely the more other ASD diagnoses under one kilometre away.
Back to square one
Considering the weakness of the cited studies, the difficulty of explaining ASD, the underlying social phenomenon of shifting diagnostic thresholds and increased awareness, it seems as though these announcements are mostly without substance. In amidst the headline-grabbing news, the NIH quietly announced the launch of a $50 million initiative into the causes of ASD.
According to the news release, the NIH will fund 13 projects “that draw on genomic, epigenomic, metabolomic, proteomic, clinical, behavioral and autism services data. These projects will integrate, aggregate and analyze existing data resources, generate targeted new data and validate findings through independent replication hubs.”
With this in mind, it really doesn’t look like paracetamol is the singular, mysterious controllable risk factor for ASD rates that President Trump and Robert F. Kennedy Jr have made it out to be. Maybe paracetamol use simply reflects infection, or some other related factor.
