Key Takeaways
- In a survey of 30,388 US adults, 26.6% said childhood vaccinations can increase autism risk and 23.1% said the same of acetaminophen use in pregnancy — claims the study classifies
- Genetics, the best-supported factor, was endorsed by 66.7%, while being male — an established risk factor — was endorsed by only 16.5%, the least of any item.
- Endorsement of the unsupported claims peaked among 35- to 44-year-olds and was lowest among adults over 65; trusting scientists was strongly associated with not endorsing them.
More Americans Endorsed a False Autism Claim Than a Real Risk Factor
Ask 30,388 American adults what can increase a child's risk of autism, and one result stands out: an unsupported claim about vaccines received more agreement than an established risk factor.
In a survey published in JAMA Health Forum, 26.6% of respondents said childhood vaccinations can increase autism risk. Another 23.1% said acetaminophen (Tylenol) use during pregnancy can.
Researchers classified both beliefs as non–evidence-based.
The percentages represent people who answered "probably true" or "definitely true." For the vaccine question, 30,288 participants provided an answer.
Now consider the comparison.
Genetic factors, the best-supported item on the list, were endorsed by 66.7%. But being a boy, an established risk factor because autism is diagnosed far more often in males, was endorsed by just 16.5% — the lowest of all twelve factors researchers asked about.
That means the unsupported vaccine claim attracted 10.1 percentage points more endorsement than an actual risk factor.
Who Was More Likely to Believe the Vaccine Claim?
The differences weren't random.
Endorsement of the vaccine claim peaked among adults aged 35 to 44 and was lowest among those over 65.
Political affiliation also mattered, as did trust in science. Respondents who said they trusted scientists "a lot" had about one-third the odds of endorsing the vaccine claim compared with those who said they trusted scientists "not at all."
Then came an unexpected finding: more frequent social media use was associated with slightly lower endorsement of the vaccine claim.
That doesn't mean social media reduces misinformation. The survey identified an association, not a cause-and-effect relationship.
What the Survey Can't Tell Us
The researchers acknowledge several important limitations.
The survey used an opt-in online panel rather than a probability sample. Results were statistically weighted, but weighting cannot fully eliminate the effects of self-selection.
It was also a snapshot, not a measure of how public beliefs have changed over time.
And answering "probably true" doesn't necessarily mean someone firmly believes a claim.
For context, the authors noted that their 26.6% vaccine figure was close to a fall-2025 YouGov result of 27%.
Health note: This article summarizes a survey of public beliefs and is not medical advice. Questions about vaccines or medication during pregnancy are best put to your own doctor, who knows your history.
The Twikup Take
The most striking comparison isn't between the two unsupported claims. It's between the vaccine figure and the figure for an actual risk factor.
An unsupported autism claim received more endorsement than an evidence-supported risk factor. That's a serious gap in public understanding.
Image: AI-generated illustration.