In the fall of 2025, a familiar worry took on an unfamiliar urgency. Acetaminophen—the pain reliever and fever reducer found in Tylenol and countless generic products—had long been considered one of the safest options during pregnancy. But on September 22, 2025, when the U.S. Food and Drug Administration announced a label change for acetaminophen products, a wave of confusion rippled through the public. Pregnant people, already navigating endless decisions about what to put into their bodies, suddenly had new reason to pause. Many did what people do when they need answers fast: they went online. And for millions of younger adults, “online” means TikTok. There, amid dancing videos, recipe clips, and comedy sketches, they found something else: a flood of content linking prenatal acetaminophen use to autism in children. The only problem, according to a new study presented at the American Academy of Pediatrics 2026 National Conference & Exhibition in San Diego, is that much of what they encountered was not based in medical evidence. The research, presented October 4 at the San Diego Convention Center, reveals how a tiny fraction of misleading videos can dominate a massive public conversation, and it sounds a clear alarm for pediatricians, public health officials, and parents alike.
To understand exactly what people were seeing, researchers from Rutgers School of Medicine analyzed 53 TikTok videos posted between September and November 2025, using search terms such as “Tylenol autism,” “acetaminophen autism,” and “paracetamol autism.” These were not random videos; they were the videos most likely to surface for anyone searching for guidance after the FDA announcement. Two independent reviewers watched each video, categorized its content, rated its educational quality using a validated tool called the Global Quality Scale, and tracked engagement metrics like views and shares. Their findings were stark. Just over half of the videos—52.8 percent—contained evidence-consistent information, meaning they reflected the scientific consensus that there is no proven causal relationship between taking acetaminophen during pregnancy and developing autism. About 28 percent contained outright misinformation. The remaining 18.9 percent were classified as mixed or uncertain. On the surface, the numbers might suggest that the platform was mostly delivering responsible content. But the real story was not in the labels; it was in the reach. The misinformation videos almost entirely outdistanced the accurate ones, raising serious questions about how social media algorithms shape health beliefs in a vulnerable audience.
The central finding of the study, and the one that most alarmed the researchers, was the glaring mismatch between the amount of misinformation and the number of views it accumulated. Altogether, the 53 videos analyzed accounted for just over 28 million views. Misinformation videos represented only 28.3 percent of all videos, yet they generated 52.4 percent of all views. In other words, a small minority of videos were responsible for the majority of exposure. Even more striking: a single misinformation video reached 10.3 million views, a number that represented 36.6 percent of all views across the entire sample. One video. Ten million people. One-third of everything watched. It is a stark illustration of how TikTok’s recommendation engine can seize on a compelling, emotionally charged narrative and amplify it far beyond the reach of careful, evidence-based explanations. The researchers noted that low-quality videos with conspiracy framing were the ones most likely to explode, not because they were more informative, but because they were more inflammatory. In an environment where attention is currency, fear and suspicion appear to outperform nuance and balance. This is not just a problem for the people who saw those videos; it is a problem for everyone, because misinformation spreads through families, clinics, and communities like ripples in water.
Who is creating the content that so powerfully shapes public perception? The study’s data offer a clear picture: not medical professionals. Health professionals produced exclusively evidence-consistent content; none of the doctors or clinical experts in the sample created misinformation. Instead, influencers—social media personalities without medical expertise—were responsible for 73.3 percent of the misinformation videos. That distinction matters, but the data also suggest that the label “influencer” is not the whole story. When the researchers used a specialized statistical model to identify independent predictors of misinformation, they found that lower educational quality and conspiracy framing were each independently associated with misinformation, while creator type alone was not statistically significant after adjustment. In plain language: it is not simply that the wrong people are making videos; it is that the content itself is designed to mislead. Misinformation videos scored far lower on the Global Quality Scale—an average of 1.73 out of 5, compared to 4.18 for evidence-consistent videos. They were also much more likely to contain themes of conspiracy (46.7 percent) and deep distrust of the pharmaceutical industry (53.3 percent). These are not calm, balanced presentations of a scientific question. They are emotional, adversarial, and designed to provoke a reaction. And in the attention economy, that is exactly why they perform so well.
The study hits home for clinicians because the confusion it documents has already walked through their exam room doors. As co-author Shivani Srivastava, a medical student at Rutgers School of Medicine, explained, the research began after pregnant patients in their own clinics started asking whether it was safe to take Tylenol. The researchers wanted to know what those patients had been exposed to online. They found an environment where conspiracy-tinged misinformation overwhelms cautious, evidence-based medical guidance. This broader climate of uncertainty was also captured in a KFF poll conducted around the same time, which found that after President Trump’s public warning, many Americans were unsure whether Tylenol use during pregnancy causes autism, with a majority of Republicans saying it was probably or definitely true. The intersection of social media, politics, and health is a volatile one. For expectant parents, especially adolescents and young adults who rely heavily on TikTok as a primary source of health information, the stakes are enormous. Decisions about whether to take a fever reducer during pregnancy should be made with a doctor, not based on a viral video. Fever during pregnancy, left untreated, carries its own risks. Yet when people are terrified by what they see online, they may avoid medically necessary treatments, creating a dangerous situation rooted in well-intentioned but misinformed fear.
The solution, according to the researchers, is not to abandon social media but to show up on it. Pediatricians and medical institutions need to be part of the conversations happening on TikTok and other platforms, offering accurate, engaging content that meets patients where they are. The study also points to the need for platform-level changes: algorithms that reward outrage and conspiracy should not be allowed to dominate conversations about sensitive medical topics. Public health agencies, medical organizations, and social media companies all have roles to play in ensuring that when a person searches for information about a medicine they are taking during pregnancy, the first thing they see is not a conspiracy theory but a trustworthy answer from a credible source. For pediatricians, the takeaway is more immediate: ask patients and families about what they are seeing online, address misinformation directly and without judgment, and offer clear, compassionate guidance based on the best available evidence. The study, while sobering, also offers hope. It demonstrates that rigorous research can capture the shape of the problem and point toward practical solutions. In a world where a single misleading video can reach ten million eyes, the presence of knowledgeable voices—encouraged by the American Academy of Pediatrics and researchers like those at Rutgers—has never been more important. The goal is not necessarily to win the internet, but to ensure that parents, especially the young and vulnerable, have what they need to make informed decisions for themselves and their children.