Eric Feigl-Ding @DrEricDing·5h
⚠️TYLENOL & AUTISM—RFK Jr and Trump are wrong—the largest & best study in the world in 2.5 MILLION KIDS—found no increased autism risk with acetaminophen (aka paracetamol, Tylenol) use by the mother during pregnancy. A crude unadjusted analysis found only a preliminary 5% risk, but once you adjust for family by matching using sibling controls (who didn’t get autism), the even tiny 5% risk vaporizes to 0% 📉. (Fun fact: I used to do drug safety epidemiology and have been whistleblower against big pharma when their drugs were dangerous—so I know a few things about drug safety data). Thread

X (Twitter): Eric Feigl-Ding on X: "⚠️TYLENOL & AUTISM—RFK Jr and Trump are wrong—the largest & best study in the world in 2.5 MILLION KIDS—found no increased autism

2) “To address unobserved confounding, matched full sibling pairs were also analyzed. Sibling control analyses found no evidence that acetaminophen use during pregnancy was associated with autism…”

3) “Conclusions and Relevance Acetaminophen use during pregnancy was not associated with children’s risk of autism, ADHD, or intellectual disability in sibling control analysis. This suggests that associations observed in other models may have been attributable to familial confounding.”

4) Interestingly, there is also no dose-response (higher dose, higher risk) relationship either. A higher dose of acetaminophen did not find a higher risk than a lower dose or medium dose pattern. This also suggests a lack of causality usually. While non linear results can occasionally happen, there usually is a reason, and so a funky unexplained up and down thing with no significant trend is a tell tale sign in epidemiology that Tylenol autism thing ain’t real.

5) The above study was in Sweden. Meanwhile, did another study find the same phenomenon? Yes, in a Norwegian study, where once you adjust the acetaminophen-autism analysis by family sibling analysis, then the association goes away to a very non-significant null 📉. ➡️Also, short term acetaminophen use under 29 days had no increase risk in either type of analysis.

6) I will post a more in depth analysis later tonight/tomorrow. Take a moment to subscribe to my Journal. I will publish a debunk on my site shortly. Thanks for your support.