The difference between politicians and physicians was on
full display at President Trump and HHS Secretary Robert Kennedy press
conference several days ago. It was hyped as an important announcement
about autism for a month. They announced that Tylenol (acetaminophen or APAP
hereafter) was a cause of autism. In the
associated hyperbole – Trump shouted not to take Tylenol and suggested that
pregnant women should "tough it out". That even though fever alone is a risk factor
for complications of pregnancy and there are no safer analgesics. Kennedy suggested that this was somehow
“transparency” about science at the FDA, CDC, and HHS, praised their
cooperation and suggested that past research was somehow flawed because of a
focus on genetics.
The news media had a field day with the press
conference. Twenty-four-hour news
channels were playing it every half hour. Controversy rather than accuracy is
how they get views. What better way to describe science as pushback on
the Trump Kennedy statements. Real
science is now pushing back against political rhetoric that claims to have
essentially replaced the scientific method.
It is a case study in political black and white thinking versus the
probabilistic thinking of medicine and science.
The purpose of this post is to look at the science and the
rhetoric around this conference. Acetaminophen (APAP) has been FDA approved
since 1951. The first Tylenol product
was an elixir for children marketed in 1955.
It was approved as an
over-the-counter medication in 1960 using an FDA monograph
procedure that allowed drugs to be grandfathered in if they were in
general use prior to the stricter regulations that began in the 1970s. These
monographs are updated with new information including risks in pregnancy.
The Trump-Kennedy Autism Press Conference (TKAC) suggests
that physicians may not be aware of APAP toxicity as much as they should
be. The reality is that physicians are
highly aware of that problem. As interns, most physicians are involved in
treating APAP overdoses and preventing severe hepatotoxicity and
death. APAP toxicity is the second leading cause of liver transplantation
worldwide. Only half of the overdoses are intentional with the remainder either
accidental (due to mixing APAP containing products or not following the
directions) or taking APAP with alcohol use, alcoholic liver disease, liver
disease, nutritional compromise, or herbal supplements. It is critical that
APAP toxicity is recognized as soon as possible to prevent irreversible liver
damage and death. It is the reason why
OTC bottles of APAP have the following warnings:
On the warning for pregnant or breast-feeding women, the FDA has risk categories in the approved labelling. APAP is listed as a Category C drug defined as shown in the slide below. In 2015 the FDA stopped using the category system and started using the Pregnancy and Lactation Labeling Rule (PLLR) – a more detailed narrative form. Despite a letter to physicians from the FDA on the APAP in pregnancy issue and the standard advice physicians have used for years I can find no new FDA package insert and no detailed PLLR language from that agency.
This post will discuss these issues is to look at the
history, rhetoric, and epidemiology in this post and then depending on how much
information I think is relevant to post on the genetics, pathophysiology, and
toxicology of autism in subsequent posts.
I will touch on a few of those points here to address the rhetoric.
The TKAC conference characterized autism as a “crisis” and
cited an unexplained increase in the prevalence of autism over the past 20
years. By unexplained I mean they were
taking it at face value as a real increase rather than reading the research and
what those authors had to say about the reasons. The reality of the prevalence
numbers and the design of these studies need to be examined. I have previously
posted that variation in prevalence estimates for psychiatric disorders
depends a lot on methodology. That
includes the study design, how the subjects are recruited, the assessments used
to make the diagnoses, and the data analysis.
There are also cultural effects over time on the same culture and in
comparisons of different international cultures.
In the United States there has been a marked increase in
awareness of autism. That awareness has increased significantly with the advent of the DSM 5 autism
spectrum disorder diagnosis. It is common to
hear people declare that they think that either they or someone they know “is
on the spectrum.” That includes
celebrities. Increased awareness can
increase early identification programs that can increase the prevalence. The expansion in prevalence also reflects the
inclusion of people with less severe symptoms.
An example comes from the ADDM
CDC study that looks at autism prevalence between selected states. Cases are identified through educational and
medical records. California had the
highest 4- and 8-year-old autism rates of any state and it was thought to be
due to a program that trained hundreds of pediatricians to identify cases early
and refer them to local centers for intervention.
The diagnostic criteria for autism have also evolved as shown in the diagram below. From very few criteria applied to more disabled populations (Kanner, Rutter) to more elaborate criteria that went from a syndrome (DSM-IV) to a spectrum (DSM 5) encompassing milder forms of the disorder (7,8). That expectedly increases the prevalence of the disorder. The smaller graphic illustrates that 3 DSM-IV syndromes were collapsed into a DSM 5 autism spectrum disorder. I am on record that the term spectrum makes no biological sense to me. It is merely a convenient way that humans have to deal with very complicated biological processes. In this case nosological convenience has blurred the boundary between people with mild forms of the disorder and no disorder. The DSM deals with that like it does with all disorders by including a necessary significant impairment in functioning term.
Very few prevalence studies look at cross sections of all the patients with that diagnosis in the community (12). In acute care psychiatry it is common to see 50- to 70-year-old adults in crisis situations because the parents they were living with have been hospitalized or died. These same people will not be in a medical or educational database with the studied diagnoses and will not be counted in those prevalence estimates. I have been able to locate only one study (12) showing that using the same criteria at the same point in time - the prevalence of autism in the older population is the same as it is in the younger population.
Another consideration of prevalence is that is the diagnosis
of autism is not an easy one. It assumes
the clinician has expertise in making the diagnosis and has adequate time to
gather and consider all the necessary information. A paper by Fusar-Poli et al (13)
highlights typical errors of misdiagnosis, the lag between first presentation
and the accurate diagnosis, and reasons behind those misdiagnoses in a large sample
of people presenting to specialty clinics for a diagnosis of autism spectrum
disorder (ASD). That same paper begins with a vignette of a middle-aged man
living in the community with some assistance to illustrate how autism can
present in the older undiagnosed population.
An interesting footnote about criteria. Like all psychiatric
disorders at one point in time only psychological causes were considered as
etiologies for the disease. In the case
of autism it was the refrigerator mother hypothesis. Cold, distant mothers were considered the
cause of autism. Folstein and Rutter’s 1977
genetic study of autism helped to reverse that line of thinking and
bring the likely cause back to genetics and biology.
That biological cause was a focus of criticism in the TKAC
conference. Secretary Kennedy went on
record stating that genetic research is unproductive and produces no
“actionable” information. Throughout
most of my career the same was said about Huntington’s Disease. When the genetic tests for Huntington’s came
on the scene, we used them like everyone else but there was still not much
optimism about an effective treatment that addressed the pathophysiology of the
disorder. All of that may have changed a few days ago when a report about a
therapy for Huntington’s (10) that may slow progression was made public. (9)
There are currently several papers about the potential for using gene therapy
for autism and other developmental disorders(11).
With all the criticism of current research at this news
conference a couple of major actionable research discoveries were not
covered. The first are studies that show
paternal and maternal age are risk factors for autism in offspring (14-21). Increasing age of the father and mother are
both risk factors for offspring with autism.
Paternal age greater than 50 years old doubles the risk of a child with
autism compared with 20-29 yr old fathers. (20). Spontaneous mutations in DNA are a likely
mechanism but several others are hypothesized.
DNA effects would also suggest that environmental factors
leading to mutations may be important.
The work done at the NIH (22-25) on this issue was not mentioned at all.
One of the researchers in this area announced that her lab
was terminated by the Trump administration.
She was working on the effect of environmental toxins on parental DNA
and her research showed an effect for maternal solvent exposure, pesticide
exposure and low fatty acid intake, and occupational exposures to phenols,
ethylene oxide and pharmaceuticals. All
these exposures are actionable by a government interested in protecting
people from environmental and occupational toxic exposures.
Coming back to the rhetoric of the TKAC conference the
overall goals seem clear – to persuade the American people that there is a
crisis, that politicians rather than scientists are best equipped to solve that
crisis, and that politicians can give you medical advice but at the same time
you should consult with your physician. This
is typical authoritarian rhetoric and if you really believe it – there is no
longer any need for science or medicine. The “crisis” in terms of increased
prevalence is explainable by broadened diagnostic criteria, inclusion of less
severely disabled individuals, and increased awareness. The statement about the
toxicity of acetaminophen is also exaggerated since in the end – despite the
President declaring that nobody should use acetaminophen – both he and the HHS
Secretary walked those statements back to the current recommendations to “consult
your physician.” The criticism about the
lack of actionable research suggests a lack of awareness of what has
been done – including work by government scientists who were fired by this
administration. I have illustrated this
with a small fraction of the autism research that is currently out there.
As a final preliminary comment – politics and rhetoric occur both inside and outside of medicine. I have seen similar statements made by researchers over the years that in the end did not pan out. They did not pan out because those hypotheses were exhaustively investigated and disproven by other researchers attempting to replicate that research. There is no similar political process. In medicine especially in some epidemiological research - a clear answer at the margins is often not possible. That is why medical treatment does not guarantee a result and involves a detailed informed consent discussion of potential risks and benefits. In politics - all it takes for a new hypothesis is somebody winning an election. And when that happens it is more likely to be a declaration than a hypothesis.
Do the
American people really want to make health care decisions based on who won an
election?
George Dawson, MD, DFAPA
Supplementary: An interview with the study author was released today and it can be accessed on the JAMA web site: https://jamanetwork.com/journals/jama/fullarticle/2839562
The reference is: Schweitzer K. Acetaminophen Use in Pregnancy—Study Author Explains the Data. JAMA. Published online September 29, 2025. doi:10.1001/jama.2025.19345
References:
1: Rosen NE, Lord C, Volkmar FR. The Diagnosis of Autism:
From Kanner to DSM-III to DSM-5 and Beyond. J Autism Dev Disord. 2021
Dec;51(12):4253-4270. doi: 10.1007/s10803-021-04904-1. Epub 2021 Feb 24. PMID:
33624215; PMCID: PMC8531066.
2: Brugha TS, McManus S, Bankart J, Scott F, Purdon S, Smith
J, Bebbington P, Jenkins R, Meltzer H. Epidemiology of autism spectrum
disorders in adults in the community in England. Arch Gen Psychiatry. 2011
May;68(5):459-65. doi: 10.1001/archgenpsychiatry.2011.38. PMID: 21536975.
3: Kanner L. Autistic disturbances of affective contact.
Nervous child. 1943 Apr;2(3):217-50.
4: Folstein S, Rutter M. Infantile autism: a genetic study
of 21 twin pairs. J Child Psychol Psychiatry. 1977 Sep;18(4):297-321. doi:
10.1111/j.1469-7610.1977.tb00443.x. PMID: 562353.
5: Murphy D, Glaser K, Hayward H, et al. Crossing the
divide: a longitudinal study of effective treatments for people with autism and
attention deficit hyperactivity disorder across the lifespan. Southampton (UK):
NIHR Journals Library; 2018 Jun. (Programme Grants for Applied Research, No.
6.2.) Chapter 17, Improving outcomes through better diagnosis: the effects of
changes in DSM-V on clinical diagnosis. Available from: https://www.ncbi.nlm.nih.gov/books/NBK518676/
6: Shaw KA, Williams S, Patrick ME, et al. Prevalence and Early Identification of Autism Spectrum Disorder Among Children Aged 4 and 8 Years — Autism and Developmental Disabilities Monitoring Network, 16 Sites, United States, 2022. MMWR Surveill Summ 2025;74(No. SS-2):1–22. DOI: http://dx.doi.org/10.15585/mmwr.ss7402a1.
7: Arvidsson O, Gillberg C, Lichtenstein P, Lundström S.
Secular changes in the symptom level of clinically diagnosed autism. J Child
Psychol Psychiatry. 2018 Jul;59(7):744-751. doi: 10.1111/jcpp.12864. Epub 2018
Jan 29. PMID: 29377119.
8: Avlund SH, Thomsen PH, Schendel D, Jørgensen M, Clausen
L. Time Trends in Diagnostics and Clinical Features of Young Children Referred
on Suspicion of Autism: A Population-Based Clinical Cohort Study, 2000-2010. J
Autism Dev Disord. 2021 Feb;51(2):444-458. doi: 10.1007/s10803-020-04555-8.
PMID: 32474837.
9: Tabrizi SJ, Flower MD, Ross CA, Wild EJ. Huntington
disease: new insights into molecular pathogenesis and therapeutic
opportunities. Nature Reviews Neurology. 2020 Oct;16(10):529-46.
10: Kaiser J. In a
first, a gene therapy seems to slow Huntington disease. Science, September 24,2025. doi:10.1126/science.zbkgxvm
11: Sahin M, Sur M. Genes, circuits, and precision therapies
for autism and related neurodevelopmental disorders. Science. 2015 Nov
20;350(6263):10.1126/science.aab3897 aab3897. doi: 10.1126/science.aab3897.
Epub 2015 Oct 15. PMID: 26472761; PMCID: PMC4739545.
12: Brugha TS,
McManus S, Bankart J, Scott F, Purdon S, Smith J, Bebbington P, Jenkins R,
Meltzer H. Epidemiology of autism spectrum disorders in adults in the community
in England. Arch Gen Psychiatry. 2011 May;68(5):459-65. doi:
10.1001/archgenpsychiatry.2011.38. PMID: 21536975.
13: Fusar-Poli L,
Brondino N, Politi P, Aguglia E. Missed diagnoses and misdiagnoses of adults
with autism spectrum disorder. Eur Arch Psychiatry Clin Neurosci. 2022
Mar;272(2):187-198. doi: 10.1007/s00406-020-01189-w. Epub 2020 Sep 6. PMID:
32892291; PMCID: PMC8866369.
14: Hultman, C. M., Sandin, S., Levine, S. Z., Lichtenstein,
P., & Reichenberg, A. (2011). Advancing paternal age and risk of autism:
New evidence from a population-based study and a meta-analysis of
epidemiological studies. Molecular Psychiatry, 16(12), 1203–1212. https://doi.org/10.1038/mp.2010.121
15: S. E. W. Sandin et al., "Autism risk associated
with parental age and with increasing parental age difference in a
population-based cohort of 5,766,794 children," Molecular Psychiatry, 2015
16: Wu, S., Wu, F., Ding, Y., Hou, J., Bi, J., & Zhang,
Z. (2017). Advanced parental age and autism risk in children: A systematic
review and meta-analysis. Acta Psychiatrica Scandinavica, 135(1), 29–41. https://doi.org/10.1111/acps.12666
17: Reichenberg, A., Gross, R., Weiser, M., Bresnahan, M.,
Silverman, J., Harlap, S., Rabinowitz, J., Shulman, C., Malaspina, D., Lubin,
G., Knobler, H. Y., Davidson, M., & Susser, E. (2006). Advancing paternal
age and autism. Archives of General Psychiatry, 63(9), 1026–1032. https://doi.org/10.1001/archpsyc.63.9.1026
18: Wood, K. A., & Goriely, A. (2022). The impact of
paternal age on new mutations and disease in the next generation. Fertility and
Sterility, 118(6), 1001–1012. https://doi.org/10.1016/j.fertnstert.2022.
19: McGrath, J. J., Petersen, L., Agerbo, E., Mors, O.,
Mortensen, P. B., & Pedersen, C. B. (2014). A comprehensive assessment of
parental age and psychiatric disorders. JAMA Psychiatry, 71(3), 301–309. https://doi.org/10.1001/jamapsychiatry.20
20: Sandin, S., Hultman, C. M., Kolevzon, A., Gross, R.,
MacCabe, J. H., & Reichenberg, A. (2012). Advancing maternal age is
associated with increasing risk for autism: A review and meta-analysis. Journal
of the American Academy of Child and Adolescent Psychiatry, 51(5), 477–486.e1. https://doi.org/10.1016/j.jaac.2012.02.018
21: Ye Q, Apsley AT, Hastings WJ, Etzel L, Newschaffer C,
Shalev I. Parental age at birth, telomere length, and autism spectrum disorders
in the UK Biobank cohort. Autism Res. 2024 Nov;17(11):2223-2231. doi:
10.1002/aur.3258. Epub 2024 Oct 30. PMID: 39474987.
22: McCanlies EC, Gu JK, Ma CC, Sanderson WT, Ludeña-Rodriguez YJ, Hertz-Picciotto I. The effects of parental occupational exposures on autism spectrum disorder severity and skills in cognitive and adaptive domains in children with autism spectrum disorder. Int J Hyg Environ Health. 2025 Jul;268:114613. doi: 10.1016/j.ijheh.2025.114613. Epub 2025 Jun 28. PMID: 40582232; PMCID: PMC12415903.
23: McCanlies EC, Gu JK, Kashon M, Yucesoy B, Ma CC,
Sanderson WT, Kim K, Ludeña-Rodriguez YJ, Hertz-Picciotto I. Parental
occupational exposure to solvents and autism spectrum disorder: An exploratory
look at gene-environment interactions. Environ Res. 2023 Jul 1;228:115769. doi:
10.1016/j.envres.2023.115769. Epub 2023 Mar 31. PMID: 37004853; PMCID:
PMC10273405.
24: McCanlies EC, Ma CC, Gu JK, Fekedulegn D, Sanderson WT,
Ludeña-Rodriguez YJ, Hertz-Picciotto I. The CHARGE study: an assessment of
parental occupational exposures and autism spectrum disorder. Occup Environ
Med. 2019 Sep;76(9):644-651. doi: 10.1136/oemed-2018-105395. Epub 2019 Jun 27.
PMID: 31248991.
25: Schmidt RJ, Kogan V, Shelton JF, Delwiche L, Hansen RL,
Ozonoff S, Ma CC, McCanlies EC, Bennett DH, Hertz-Picciotto I, Tancredi DJ,
Volk HE. Combined Prenatal Pesticide Exposure and Folic Acid Intake in Relation
to Autism Spectrum Disorder. Environ Health Perspect. 2017 Sep 8;125(9):097007.
doi: 10.1289/EHP604. PMID: 28934093; PMCID: PMC5915192.
26: Pernia S, DeMaagd G. The New Pregnancy and Lactation
Labeling Rule. P T. 2016 Nov;41(11):713-715. PMID: 27904304; PMCID: PMC5083079.
Reference Credit:
h/t to Tyler Black, MD @tylerblack32 for reference 12.
Graphics Credit
1: DSM-IV and DSM 5 graphics are from the respective DSMs copyrighted by the American Psychiatric Association and reproduced here only for educational purposes.
2: FDA package insert information is reproduced here and considered in the public domain.
3: The detailed ADDM graphic of ASD prevalence by state is form the Mortality and Morbidity Weekly Report (MMWR) (see reference 6 and is in the public domain).
Commentary on the Trump Kennedy Press Conference Commentaries with time points in the transcript:
1: Trump at 4:43
“Which is basically commonly known as Tylenol during
pregnancy and can be associated with a very increased risk of autism. So taking
Tylenol is not good. All right, I'll say it; it's not good. For this reason
they are strongly recommending that women limit Tylenol use during pregnancy
unless medically necessary. That's, for instance, in cases of extremely high
fever that you feel you can't tough it out; you can't do it. I guess there's
that. It's a small number of cases, I think. But if you can't tough it out, if
you can't do it, that's what you're going to have to do. You'll take a Tylenol,
but it'll be very sparingly. It can be something that's very dangerous to the
woman's health. In other words, a fever that's very, very dangerous and ideally
a doctor's decision because I think you shouldn't take it, and you”
Trump simultaneously skirts the issue of the potential dangers of acetaminophen in pregnancy while walking back that recommendation to the current package insert statement (see graphic above).
2: Trump at 36:59
“I understand it's maybe 10% of the women that are pregnant
would perhaps be forced to use it, and that would mean you just can't tough it
out. No matter what you do, you can't tough it out. So that's up to you and
your doctor.”
Trump seems to confuse the analgesic effect of acetaminophen with the antipyretic effects important to prevent complications of pregnancy.
3: Trump at 40:12
“Don't take Tylenol. Don't give Tylenol to the baby after
the baby's born. Every time the baby gets a shot, the baby goes, gets a shot,
they say, "Here, take a couple of Tylenol." I've heard that for
years. Take Tylenol. Don't take Tylenol, don't have your baby take Tylenol.
Now, Tylenol is fine for people that aren't pregnant, that aren't in the
situation that we're talking about one very specific situation. If you're
pregnant, don't take Tylenol. When you have your baby, don't give your baby
Tylenol at all unless it's absolutely necessary. Don't do it.”
Trump clearly states not to take acetaminophen if you are pregnant - with no package insert qualifier. He also suggests that it is dangerous for infants.
4: Trump at 44:03
“And the other things I told you about, just… The word, tough it out. It's easy for me to say tough it out. But sometimes in life with a lot of other things, you have to tough it out also. Don't take Tylenol. Don't give Tylenol to the baby. When the baby's born, they throw it at you, "Here, give them a couple of Tylenol." They give them a shot. They give them a vaccine. And every time they give them a vaccine, they throw in Tylenol. And some of these babies they're long born, and all of a sudden they're gone. And it doesn't hurt not to do it. It doesn't hurt. There's no downside. There's no downside at all.”
Trump persists with his "tough it out" message missing the point of acetaminophen use in pregnancy. He also suggests that vaccinations lead to more acetaminophen use in infants.
5: Kennedy at 14:10
“NIH research teams are currently testing multiple
hypotheses with no area off-limits. We promise transparency as we uncover the
potential causes and treatments, and we will notify the public regularly of our
progress. Today we are announcing two important findings from our autism work
that are vital for parents to know as they make these decisions. First, HHS
will act on acetaminophen. The FDA is responding to clinical and laboratory
studies that suggest a potential association between acetaminophen used during
pregnancy and adverse neurodevelopmental outcomes, including later diagnosis
for ADHD and autism. Scientists have proposed biological mechanisms linking
prenatal acetaminophen exposure to altered brain development. We have also
evaluated the contrary studies that show no association. Today, the FDA will
issue a physician's notice about the risk of acetaminophen during pregnancy and
begin the process to initiate a safety label change. HHS will launch a
nationwide public service campaign to inform families and protect public health.”
No mention of the research program cancelled by the Trump administration as noted above. Not clear who he means when he talks about "we" evaluating studies. Does he mean him and Trump? Is there anybody left at NIH, CDC, HHS who can do those evaluations?
6: Kennedy at 15:28
“The FDA also recognized that acetaminophen is often the
only tool for fevers and pain in pregnancy, as other alternatives have
well-documented adverse effects. HHS wants, therefore, to encourage clinicians
to exercise their best judgment and use of acetaminophen for fevers and pain in
pregnancy by prescribing the lowest effective dose for the shortest necessary
duration and only when treatment is required. Furthermore, thanks also to the
politicization of science. The safety of acetaminophen against the risk of
neurodevelopmental disorders in young children has never been validated.”
7: Kennedy at 52:13
"But also it's just common sense, because you're only seeing this in people who are under 50 years of age. If it were better recognition or diagnosis, you'd see it in 70-year-old men. I've never seen this happening in people my age. I've never seen a case of full-blown autism, and that means profound autism, I want to be very careful, head banging, stimming, toe walking, nonverbal, non-toilet trained. I've never in my life seen a 70-year-old man who looks like that. You're only seeing it in kids. It's an epidemic"
Kennedy simultaneously displays his lack of knowledge about the historical development of the autism diagnosis (DSM-IV restricted age of onset to 3 years) and perpetuates a stereotype of a person with severe developmental disabilities who would typically require institutional care. There are many older individuals with ASD living in the community - some may be your neighbors. And as noted in the above post - the expansion in prevalence has occurred primarily due to milder cases that were included in new diagnostic criteria.
