Penn AI Study Flags GLP-1 Side Effects in 410,000 Reddit Posts
Updated
Updated · ScienceAlert · Jun 2
Penn AI Study Flags GLP-1 Side Effects in 410,000 Reddit Posts
3 articles · Updated · ScienceAlert · Jun 2
More than 410,000 Reddit posts over six years led University of Pennsylvania researchers to flag possible GLP-1 side effects beyond established complaints, notably menstrual irregularities and temperature-related symptoms such as chills and hot flashes.
GPT models scanned posts mentioning semaglutide, tirzepatide and brand names including Ozempic, Wegovy, Mounjaro and Zepbound, aiming to surface patient-reported issues faster than clinical trials can.
Researchers said the method also picked up known effects like nausea, suggesting a real signal, but stressed the findings do not prove the drugs caused the newly highlighted symptoms.
The team said Reddit discussions may capture concerns patients do not raise with doctors, though the platform skews younger, male and US-based, limiting how representative the data is.
Published in Nature Health, the study adds to widening scrutiny of GLP-1 drugs, whose benefits in weight loss and diabetes are being weighed against a growing list of possible risks and other effects.
AI found hidden Ozempic side effects on Reddit. Which other popular drugs have risks that only patients are discussing online?
Genetics can predict Ozempic's side effects. Can AI and DNA tests soon tell you if a new drug is safe for you?
With AI scanning social media and the FDA running real-time trials, is the old drug safety system now obsolete?
Real-World GLP-1 Side Effects Revealed: Penn AI Study Mines 410,000 Reddit Posts
Overview
A groundbreaking study by the University of Pennsylvania, published in April 2026, used advanced AI to analyze Reddit posts about GLP-1 medications. By examining real-world discussions, researchers identified previously underreported side effects, such as reproductive symptoms, temperature changes, and fatigue. This innovative approach marks a major step forward in drug safety monitoring, as it detects subtle or emerging issues that traditional clinical trials might miss. Leveraging social media data with AI not only uncovers new safety concerns earlier but also highlights the value of patient-shared experiences in improving pharmacovigilance and healthcare outcomes.