Practiceby Lab 10 Get started ›
Crossfire

Should you avoid Tylenol in pregnancy?

What the experts say

Aviva RommMD · family physician (NPPES taxonomy 207Q00000X, MA licence 253326 and NY licence 9040902; Yale MD awarded 1 Sep 2009, verified in Yale's own thesis repository; stated board certification not confirmed against ABFM) · midwife, apprenticeship-trained, approx. 25 years, current standing not established · herbalist, past president of the American Herbalists Guild and former Medical Director of the American Herbal Pharmacopoeia, both confirmed by the organisations themselves

Use it sparingly and only when needed. Do not avoid it at all costs, and do not blame yourself for having used it.

What it rests onShe walks the literature in order and reports what each design can and cannot show: 2018 cohort data raising the alarm, a 2021 Nature Reviews Endocrinology consensus statement supported by 91 researchers recommending avoidance unless prescribed, a 2024 JAMA sibling analysis of nearly 2.5 million Swedish children finding no association and no dose-response once siblings were controlled, and an August 2025 BMC Environmental Health Navigation Guide review of 46 studies in which 27 found associations and the higher-quality studies were more likely to. Against that she sets the risk of the untreated condition, citing an association between untreated high maternal fever and neurodevelopmental disorders, and notes ibuprofen is not recommended in pregnancy. She quotes the FDA Commissioner that the choice belongs with parents and that use remains reasonable in certain scenarios.

Source: avivaromm.com ↗
Paul Gregory SaladinoMD · physician and animal-based nutrition creator

Acetaminophen should be avoided in pregnancy, labour and the first three years, and doing so could cut autism rates.

What it rests onA metabolic argument plus a reading of the same evidence. His mechanism: acetaminophen clears safely by glucuronidation and sulfation, newborn livers have virtually no glucuronidation capacity in the first weeks, so exposure is shunted to the cytochrome P450 route that produces NAPQI, depletes glutathione and damages mitochondria, including in the developing brain. His reading of the sibling study is that its design excludes the vulnerable: 'all this proves is that among families with NO genetic susceptibility to acetaminophen toxicity, the drug doesn't cause autism.' He cites the same 2025 46-study review Romm cites, for the proposition that stronger designs found stronger associations. He labels his conclusion a hypothesis and calls it very controversial.

Source: paulsaladinomd.com ↗

Overview

Romm is addressing a pregnant woman deciding about a headache or a fever now, and her frame includes the risk of leaving that fever untreated. Saladino's window is wider than pregnancy: pregnancy, labour and delivery, and the first three years of life, with the newborn period as the point of maximum vulnerability in his account.

So they are not addressing the same exposure, and treating this as one question about one drug in one population flattens the disagreement.

Where they agree

Both agree the evidence is unsettled and that exposure should be limited. Romm's own sentence is 'when in doubt, limit or avoid it if you can, until we have more information.'

Both cite the same 2025 Navigation Guide review and both accept its finding that higher-quality studies more often found associations. Neither claims a randomised trial exists. The disagreement is about what the sibling analysis proves and about whether avoidance should be total around birth.

What would settle it

Nothing available will settle it: a randomised trial of acetaminophen exposure in pregnancy will not be run. What would move it is a sibling analysis stratified by the metabolic vulnerabilities Saladino names, which would test his objection directly, or a cohort with measured biomarkers of glucuronidation capacity rather than self-reported intake, which is the weakness Romm flags in the existing cohort data.

Practice does not pick a winner here

You do, and your call is dated, kept, and yours to change. Follow the people you already trust and what they publish turns into a short daily practice.

Get started

Free to sign up and use. We do not ask for a card.