When political figures issue medical warnings, the weight of evidence deserves careful examination. Recent claims linking paracetamol use in pregnancy to autism have circulated widely, yet the most methodologically rigorous research—particularly a Swedish study of 2.5 million children using sibling comparisons to control for genetics—finds no such causal connection. The apparent associations in earlier studies appear to reflect the shadow of confounding factors rather than the drug itself. In the meantime, the documented risks of untreated fever in early pregnancy remain a more pressing concer
Major study finds no autism link to pregnancy paracetamol use
The link vanished when researchers compared siblings with different exposures
Mark
So the headline says there's no link, but the Swedish study initially found a marginally increased risk. What changed?
Mimi
The initial analysis looked at all children exposed to paracetamol versus those not exposed. But that comparison mixes together a lot of variables—maybe mothers who took paracetamol were sicker, or had different socioeconomic circumstances. When they compared siblings instead, controlling for shared genetics and home environment, the link disappeared.
Luke
But we should be clear: that sibling analysis is more methodologically sound, but it's still one study. The Harvard review found 27 studies showing associations. Are we saying all 27 had confounding problems?
Mimi
Not necessarily all of them, but the February review found most had biases in participant selection and didn't account for confounding factors. When confounding was accounted for, the associations weakened substantially.
Mark
What are these confounding factors we keep mentioning?
Mimi
Maternal illness, medication use, body mass index, smoking, alcohol, pregnancy complications, parental age, genetics, socioeconomic status. Any of these could influence both whether a mother takes paracetamol and whether her child develops autism.
Luke
So the real problem is that we can't easily separate cause from correlation in observational studies?
Mimi
Exactly. The sibling design is clever because it holds those shared factors constant. But it requires detailed data on individual families, which most studies don't have.
Mark
What about the actual risk? If untreated fever is dangerous, shouldn't pregnant women take paracetamol when they have a fever?
Mimi
Yes. Untreated high fever in early pregnancy is linked to miscarriage, neural tube defects, cleft lip and palate, and heart defects. Infections themselves are linked to higher autism risk. So avoiding paracetamol entirely could create worse outcomes.
Luke
The category A classification—does that mean it's proven safe, or just that it's been used widely without obvious problems?
Mimi
It means it's been widely used without evidence of increased birth defects. That's not the same as proven safe in every circumstance, but it's the standard for pregnancy medications.
O Pulso
- A high-profile political warning about paracetamol and autism has alarmed pregnant women, despite the claim outpacing the science behind it.
- Decades of observational studies produced conflicting signals—27 showing possible links, nine showing none, four suggesting lower risk—leaving the public with a confusing and incomplete picture.
- The critical breakthrough came from sibling analysis: when researchers compared children in the same family with different prenatal exposures, the apparent autism risk disappeared entirely.
- Earlier studies are now understood to have been distorted by unmeasured variables—maternal illness, genetics, socioeconomic status, and lifestyle factors—that mimicked a paracetamol effect.
- The real clinical danger runs in the opposite direction: untreated fever in early pregnancy is linked to miscarriage, neural tube defects, and higher autism risk from infection.
- Medical authorities maintain that paracetamol remains a Category A safe medication in pregnancy, recommended at the lowest effective dose for the shortest necessary duration.
When political figures issue medical warnings, the weight of evidence deserves careful examination. Recent claims linking paracetamol use in pregnancy to autism have circulated widely, yet the most methodologically rigorous research—particularly a Swedish study of 2.5 million children using sibling comparisons to control for genetics—finds no such causal connection. The apparent associations in earlier studies appear to reflect the shadow of confounding factors rather than the drug itself. In the meantime, the documented risks of untreated fever in early pregnancy remain a more pressing concern for expectant mothers and their physicians.
Donald Trump's recent advice that pregnant women avoid paracetamol—except in cases of severe fever—has prompted a closer examination of what the research actually shows. The picture is considerably more complicated than a simple warning allows.
Paracetamol, known as acetaminophen or Tylenol, is one of the most widely used medications during pregnancy. Australia's Therapeutic Goods Administration classifies it as Category A—meaning extensive use has shown no evidence of increased birth defects. The medical concern, in fact, runs the other way: untreated high fever in early pregnancy carries documented risks including miscarriage, neural tube defects, cleft lip and palate, heart defects, and higher autism risk from infection itself.
The recent alarm traces partly to a 2021 statement from an international expert panel suggesting paracetamol may alter fetal development. Harvard researchers subsequently reviewed 46 studies on the topic, finding deeply mixed results. The most rigorous among them—a 2024 Swedish study of nearly 2.5 million children—initially appeared to show a marginally increased autism and ADHD risk. But when researchers compared sibling pairs with different levels of prenatal paracetamol exposure, the link vanished completely. No increased risk of autism, ADHD, or intellectual disability remained.
This sibling method matters because it controls for the genetic and environmental factors families share. Siblings of autistic children already face a 20 percent chance of autism themselves, and household conditions can shift risk independently of any medication. By isolating prenatal exposure as the only variable, researchers could see that paracetamol itself was not the culprit—shared family characteristics were creating the illusion of a link in earlier, less rigorous studies.
A separate February review confirmed that most prior studies were compromised by selection bias and failure to account for confounding factors such as maternal illness, body mass index, smoking, genetics, and socioeconomic status. When those factors were properly controlled, apparent associations weakened substantially. The medical consensus holds: there is no clear evidence paracetamol harms an unborn baby. Pregnant women with fever are still advised to treat it—and paracetamol remains an appropriate, medically supported option for doing so.
Donald Trump has advised pregnant women to avoid paracetamol except in cases of severe fever, citing concerns about a possible link to autism. The claim has prompted a closer look at what the actual research shows—and the picture is considerably more complicated than a simple warning allows.
Paracetamol, sold as acetaminophen or Tylenol, is one of the most commonly used pain relievers and fever reducers during pregnancy. Australia's Therapeutic Goods Administration has reaffirmed that it remains safe for pregnant women, classified as a Category A medication—meaning it has been used widely by pregnant women and women of childbearing age without evidence of increased birth defects. The real medical concern cuts the other direction: untreated high fever in early pregnancy carries documented risks including miscarriage, neural tube defects, cleft lip and palate, and heart defects. Infections during pregnancy are also linked to higher autism risk.
The source of recent concern traces to a 2021 statement from an international panel of experts who reviewed evidence from human and animal studies and warned that paracetamol use during pregnancy may alter fetal development. Last month, researchers from Harvard University conducted a systematic review of existing research on paracetamol and neurodevelopmental disorders, including autism and ADHD. They identified 46 studies. Of these, 27 reported links between paracetamol use in pregnancy and neurodevelopmental disorders in offspring, nine showed no significant link, and four suggested paracetamol was associated with lower risk.
The most rigorous study in their review examined nearly 2.5 million children born in Sweden between 1995 and 2019, published in 2024. When researchers first analyzed the data, they found a marginally increased risk of autism and ADHD associated with paracetamol exposure during pregnancy. But then they performed a different analysis: they looked at sibling pairs where one child was exposed to paracetamol in the womb and the other was not, or where siblings had different levels of exposure. This approach accounts for genetic and environmental factors the siblings share. When the researchers applied this method, the apparent link vanished. No increased risk of autism, ADHD, or intellectual disability emerged.
The reason this matters is that siblings of autistic children have a 20 percent chance of also being autistic, and environmental factors in a home can shift autism risk. By comparing siblings who grew up in the same household with the same parents and genes but had different prenatal exposures, researchers can isolate whether paracetamol itself is the culprit or whether something else—shared family factors—explains the pattern. The Swedish study's authors concluded that associations found in other studies likely stem from "confounding" factors: unmeasured influences that distort research findings.
A separate review published in February examined the strengths and weaknesses of the literature on this topic. It found that most studies were difficult to interpret because they contained biases in how they selected participants and failed to account for confounding factors. When researchers did account for shared genetic and environmental influences among siblings, any apparent associations weakened substantially. This suggests that shared family characteristics—a mother's medication use, illnesses, body mass index, alcohol consumption or smoking, pregnancy complications, parental ages, the newborn's Apgar scores, breastfeeding patterns, genetics, and socioeconomic status—may have created the illusion of a paracetamol link in earlier studies.
The medical consensus remains that there is no clear evidence paracetamol harms an unborn baby. Like any medication taken during pregnancy, it should be used only at the lowest effective dose and for the shortest possible duration. If a pregnant woman develops a fever, treating it is medically important—and paracetamol remains an appropriate option. If the standard dose doesn't control symptoms or pain persists, the guidance is to contact a doctor, midwife, or maternity hospital for further advice.
Citações Notáveis
Associations found in other studies may be attributable to confounding factors—influences that can distort research findings— Authors of the 2024 Swedish study
Paracetamol should be used only at the lowest effective dose and for the shortest possible duration, but treating fever in pregnancy is medically important— Medical guidance cited in the review
Quer a matéria completa?
Leia o original em The Tribune ↗
Politics & Governance