More than 400,000 Reddit posts just did something no clinical trial could: they let real patients describe, in their own words, what weight-loss drugs like Ozempic and Mounjaro are doing to their bodies.
Quick Take
- University of Pennsylvania researchers used artificial intelligence to scan 410,198 Reddit posts made between May 2019 and June 2025.
- They found 67,008 users who said they were actually taking semaglutide or tirzepatide, the active drugs in Ozempic, Wegovy, Mounjaro and Zepbound.
- Nearly 44% of those users described at least one side effect, with stomach problems topping the list.
- Menstrual changes and temperature swings like chills and hot flashes showed up more than expected based on official drug labels.
- Researchers say the posts point to patterns worth studying further, not proof that the drugs caused these symptoms.
How Researchers Turned Reddit Chatter Into Data
The Penn team did not just skim forum posts for complaints. They built an AI system to sift through 410,198 Reddit posts and comments across nine large subreddits focused on GLP-1 drugs and weight loss. From that pile, they narrowed the group to 67,008 people who confirmed they were actually using semaglutide or tirzepatide, rather than just discussing the drugs. That confirmation step matters. It separates people speaking from experience from armchair commenters repeating headlines.
Of those confirmed users, 29,172 people, or 43.5%, described at least one side effect in their posts. That is a huge sample size compared to a typical clinical trial, which might track a few thousand patients over a set period. Reddit users, by contrast, wrote in real time about what they felt, often months or years into treatment, giving researchers a messier but much larger window into everyday drug use.
The Stomach Problems Everyone Already Knew About
Most of what the AI found lines up with what doctors have warned about for years. Nausea led the list at 36.9%, followed by fatigue at 16.7%, vomiting at 16.3%, constipation at 15.3%, and diarrhea at 12.6%. These gastrointestinal complaints match what shows up in official drug trials and package inserts. That overlap is actually reassuring. It tells researchers the Reddit data reflects real patient experience rather than noise or exaggeration, which gives more weight to the symptoms that do not match existing labels.
The Patterns Drug Labels Don’t Fully Cover
Beyond the expected stomach trouble, two categories stood out. Reproductive symptoms, mainly menstrual changes like irregular cycles, bleeding between periods, and heavier flow, showed up in the data. Researchers note this group made up close to 4% of users who reported any side effect, a small but notable slice given how little attention menstrual changes get in current drug labeling. Temperature-related complaints, including chills, feeling unusually cold, and hot flashes, also appeared often enough to catch researchers’ attention.
Why the Researchers Are Hitting the Brakes
Here is the part that deserves more attention than it usually gets. The study’s own authors say plainly that these findings do not prove the drugs caused the symptoms. Reddit posts are self-reported, unverified, and written by a self-selected group of people motivated enough to post online. That is a real limitation, not a minor footnote. Responsible reporting on this study should treat it as a lead worth chasing, not a verdict already reached.
Social Media as an Early Warning System
This is not the first time researchers have mined social platforms for drug safety clues. Pharmacovigilance experts have long argued that online listening can catch unexpected adverse events earlier than traditional reporting systems, precisely because patients describe symptoms doctors’ forms never ask about. The tradeoff is data quality. Social media posts cannot replace controlled trials or formal adverse-event databases, but they can flag where regulators and drugmakers should look harder before symptoms get dismissed or buried in fine print.
AI analyzed 410,000 Reddit posts about GLP-1 drugs—and found patterns worth investigating, not proof of new side effects.
Here’s what the study actually found:
https://t.co/jmDBeJlUXe#GLP1 #Ozempic #AIHealth #MedicalResearch— Muslance (@muslancemedia) September 18, 2026
For the millions of Americans now taking GLP-1 drugs, this study is a reminder that patient voices carry real value, even when they show up in an unlikely place like a Reddit thread. Drug companies and regulators owe patients honest, complete labeling, not marketing copy dressed up as science. If menstrual changes and temperature swings are real and common enough to trend across tens of thousands of posts, families deserve to know before they start treatment, not after.
Sources:
medicalxpress.com, sciencedaily.com, medicaldaily.com, medicalnewstoday.com, thenews.com.pk
