Reddit deep dives can involve anything from TV show fan theories to DIY advice.
A new AI-assisted analysis now suggests that the 'front page of the internet' could help researchers spot potential side effects of GLP-1drugs used to manage weight and diabetes, such as Ozempic, Wegovy, Mounjaro, and Zepbound.
A team from the University of Pennsylvania analyzed more than 410,000 Reddit posts across a six-year span, looking for mentions of two active ingredients in widely used GLP-1-based drugs: "semaglutide" or "tirzepatide", or their brand names.
" Clinical trials are the gold standard, but by design, they are slow," says computer and information scientist Sharath Chandra Guntuku.
"This is not a replacement for trials, but it can move much faster, and that speed matters when a drug goes from niche to mainstream almost overnight."
When it came to potential side effects that doctors may not know about, two groups of complaints stood out: reproductive health issues (including irregular menstrual cycles) and temperature-related problems (such as chills and hot flashes).
The way the body's metabolism burns energy is known to impact temperature balance, so there is a real mechanism that potentially links the side effect to the drug in a causative way. But there is less research on how these drugs affect the menstrual cycle.
It's worth noting that these previously "unrecognized potential effects" were far from the most commonly reported on Reddit, and the researchers emphasize that they don't see their new analysis as a replacement for trials or clinical assessments.
However, the findings suggest that online boards and peer-to-peer conversations may be places where people feel able to discuss certain symptoms, some of which they might not mention to a doctor.
"Some of the side effects we found, like nausea, are well known, and that shows that the method is picking up a real signal," says Guntuku.
"The underreported symptoms are leads that came from patients themselves, unprompted, and clinicians could potentially pay attention to them."
This kind of large-scale data capture and interrogation is made possible by the latest AI models. Here, the researchers used GPT models from OpenAI to crunch through Reddit posts and find patterns.
That's no easy task, given the volume of text and the variety of ways people might talk about GLP-1 drugs and their side effects. These scans can be completed rapidly, identifying potential sources for investigation in future clinical trials.
With little else known about the Reddit posters involved, the research can't be definitive about whether drugs like Ozempic or Mounjaro are actually causing these symptoms.
However, these self-reported side effects are worth further investigation.
"These drugs are thought to work by engaging part of the brain called the hypothalamus, which helps regulate a wide variety of hormones," says psychologist Jena Shaw Tronieri.
"That doesn't mean the medications are necessarily causing these symptoms, but it could suggest that reports of menstrual changes and body temperature fluctuations are worth studying more systematically."
GLP-1 drugs are named after glucagon-like peptide-1, a natural hormone that the medications mimic. Specifically, the drugs limit appetite, slow down digestion, and trigger the release of insulin from the pancreas in response to high blood sugar levels.
While these treatments are associated with significant benefits in weight loss and diabetes management, research is ongoing into other potential consequences of GLP-1 use.
Those consequences potentially include protection against Alzheimer's, better cardiovascular health, and a greater risk of acute or chronic pancreatitis – so, quite the mix. We also know that regaining most of the weight after GLP-1 treatments is common.
The kind of online analysis done here may highlight problems that might otherwise be missed.
"The whole point of this kind of approach is that it can move quickly, and that's exactly when it's most valuable," says Guntuku.
Reddit tends to skew towards younger, male, US adults – but that doesn't mean it can't be useful for flagging problems that researchers need to know about.
Related: Ozempic-Like Drugs Can Help You Lose Weight, But There's a Catch
"Clinical trials generally identify the most dangerous side effects of drugs, but they can fail to find what symptoms patients are most concerned about," says computer and information scientist Lyle Ungar.
"Even though social media is not necessarily representative, a large collection of posts may reflect additional concerns."
The research has been published in Nature Health.