It begins with a simple, anxious scroll. You’re pregnant, maybe running a fever or nursing a stubborn headache, and you reach for the medicine cabinet. Before you open the bottle, you open TikTok. Almost immediately, a video appears: a young mom, eyes wide, telling her followers that Tylenol during pregnancy caused her child’s autism. Another clip warns that the FDA knows and isn’t telling you. These videos feel urgent, personal, and terrifyingly plausible. And according to new research presented at the American Academy of Pediatrics 2026 National Conference & Exhibition in San Diego, a troubling number of these posts are exactly the kind of content the algorithm rewards. Researchers reviewed 53 TikTok videos posted between September and November 2025 and found that just over a quarter of them contained misinformation. But those misleading videos pulled in more than half of all views — about 14.7 million of the 28.1 million total views in the sample. It’s a striking example of how the most watched health content can also be the least accurate, and it leaves expecting parents trying to make a real medical decision in the middle of a very loud digital storm.
The study, reported by Contemporary Pediatrics, took a close look at what was actually circulating on TikTok during a period of intense public attention. The timeline matters: in late September 2025, the FDA announced changes to acetaminophen labeling, and suddenly the internet was buzzing with questions about whether Tylenol causes autism. Researchers searched TikTok using terms like “tylenol autism,” “acetaminophen autism,” and “paracetamol autism,” along with the #tylenol hashtag. They ended up with 53 videos, which two independent reviewers classified by accuracy, creator type, educational quality, and framing. Their agreement was strong, with a Cohen’s kappa of 0.86, meaning their judgments were highly consistent. The results split the content into three rough categories: just over half of the videos (52.8%) were consistent with available evidence, 18.9% were mixed or uncertain, and 28.3% contained outright misinformation. But the split in views tells a different story. Those misinformation videos accounted for 52.4% of all views, far more than their share of the content. What made the misleading videos so watchable? They were also more emotionally charged. Nearly half used conspiracy framing, and more than half expressed deep distrust of drugmakers. The difference in quality was stark: misinformation videos scored an average of 1.73 on the Global Quality Scale, a 1-to-5 measure of educational value, while evidence-consistent videos averaged 4.18. And the creators behind the videos were telling a distinct story. Health professionals in the sample posted only accurate, evidence-based content. Influencers, meanwhile, produced 73.3% of the misinformation. These weren’t doctors sharing medical nuance; they were personalities built on engagement, and engagement favors alarm.
Before we let the headline sink in too deeply, though, it’s important to understand what this study can and cannot tell us. This is a small, preliminary, cross-sectional analysis presented at a medical meeting. It is not a peer-reviewed paper, and it does not capture the full universe of Tylenol-related content on TikTok. The sample of 53 videos was selected through specific search terms and a public hashtag, so it may not represent every post an expecting parent would encounter. Perhaps the most significant caveat is that a single video carries a huge share of the weight. One misinformation clip drew 10.3 million views, which is 36.6% of all the exposure in the entire sample. Remove that one post, and the numbers would look very different. The study also doesn’t prove that TikTok’s recommendation algorithm intentionally amplified misinformation. The researchers suggested that the platform appears to have given low-quality content disproportionate reach, but the data demonstrate a pattern of exposure, not a direct mechanism. In other words, we know these misleading videos got watched, but we can’t say for certain why the algorithm behaved the way it did. Still, the pattern is worth taking seriously. Whether by design or by accident of engagement, the content that is least aligned with medical evidence is winning the race for eyeballs.
The underlying medical question, of course, is far more complicated than a viral video suggests. It’s easy to hear “FDA announces labeling change” and assume that regulators have confirmed a link between acetaminophen and autism. But the scientific reality is more nuanced. Observational studies have reported associations between prenatal acetaminophen use and later diagnoses of autism or ADHD. Associations are not the same as causation. When a pregnant person takes Tylenol, it’s usually because they have a fever or pain, and the underlying illness itself can influence fetal development. Sorting out whether the medicine, the illness, or something else entirely is responsible is incredibly difficult. In September 2025, federal health officials did suggest that a possible link existed and the FDA began the process of updating acetaminophen labeling, but they were careful to note that a causal relationship had not been established. Major medical organizations, including the American College of Obstetricians and Gynecologists and the American Academy of Pediatrics, continue to recommend acetaminophen as a first-line treatment for fever and pain during pregnancy. The political world has weighed in too. Dr. Heidi Overton, the nominee to lead the FDA, told senators during her confirmation process that she had not seen evidence of a causal link between Tylenol and autism. So while internet discussions often frame the science as settled — usually with either “it’s completely safe” or “they’re hiding the truth” — the actual situation is a place of careful uncertainty, not dramatic revelation.
For the people most affected by this content — pregnant people, those planning pregnancy, and the partners and parents who share videos with them — the takeaway isn’t to panic or to dismiss all medicine. It’s to pause and think about the source. Untreated high fever and significant pain during pregnancy carry their own risks. A fever can be a sign of infection, and sustained high fever in early pregnancy has been linked to certain birth defects. Severe pain can interfere with sleep, nutrition, and daily functioning in ways that are not healthy for anyone. That means choosing not to take Tylenol because of a TikTok clip is not automatically a risk-free choice. The decision is a medical one, and it deserves the context that only a clinician can provide. If a video claims that doctors are covering up the truth or that mainstream medicine refuses to listen, that’s a red flag. If the creator’s credentials are unclear or entirely absent, that matters too. The most reassuring and responsible step is to ask an obstetrician, midwife, or pharmacist directly. They can help weigh the lowest effective dose for the shortest possible duration, and they can talk through what is known, what is unknown, and what the alternatives might be. It’s also worth remembering that over-the-counter doesn’t mean risk-free. Pregnant people should never switch to another pain reliever like ibuprofen or aspirin without medical guidance, because those drugs carry their own pregnancy cautions. And there are certain symptoms that should never wait for a social media scroll to resolve: a fever of 100.4 degrees Fahrenheit or higher, a severe headache, vision changes, or reduced fetal movement all warrant prompt contact with a clinician.
Looking at the bigger picture, this study is a reminder that medical information is now being consumed in an environment where entertainment value and credibility are constantly in tension. The fact that health professionals in the sample posted only evidence-consistent content is reassuring, but their voices are being drowned out by the sheer reach of influencers. A second analysis presented at the same AAP meeting found that only about one in five health organizations reviewed had posted Instagram content addressing the acetaminophen claims, according to Contemporary Pediatrics. That’s a startling gap. While medical institutions publish papers and issue guidelines, the public is having a different conversation on platforms where attention is currency and nuance doesn’t travel well. Researchers said institutions need a stronger, more visible presence where patients are actually seeking information. That doesn’t mean doctors have to become influencers, but it does mean they need to meet people where they are — with clear, honest, and humane explanations of what science knows and what it doesn’t. Full study results may eventually be published in a journal, offering a clearer picture of how this misinformation spreads. For now, the message for any expecting parent is simple: you are not wrong to ask questions. You are not foolish for feeling confused by contradictory information. But let the people who know your medical history help you make the choice. A viral video can tell you what to fear, but it cannot tell you what is safe for you and your baby. That answer belongs to a conversation between you and a trusted clinician — not to an algorithm.