- An artificial intelligence (AI) analysis of over 400,000 Reddit posts found that people taking GLP-1 medications frequently report side effects not fully captured in clinical trials or drug labels.
- Two notable categories of underreported symptoms include reproductive issues, such as irregular menstrual cycles, and temperature-related complaints, including chills and hot flashes.
- Additionally, fatigue was commonly discussed online, despite being less prominent in clinical trial data, while known side effects like nausea were also widely reported.
- Researchers stress the findings do not prove causation but suggest that AI analysis of social media could help identify patient concerns earlier and guide future clinical research.
Common side effects of GLP-1 drugs are those of a gastrointestinal nature, such as nausea, vomiting, diarrhea, and constipation. Research suggests these adverse events may occur in 40 to 85% of people. Health experts advise that people can make dietary adjustments to
Although many of these side effects are mild to moderate in severity and generally resolve shortly, adverse events remain a common cause of discontinuing the drug.
While these side effects are well-documented, many people also often report anecdotal adverse events while using GLP-1 drugs.
Now, a new study published in
Led by researchers from the University of Pennsylvania, the study examined more than 400,000 Reddit posts from nearly 70,000 users discussing GLP-1 receptor agonists, including semaglutide and tirzepatide.
The analysis identified two main symptom categories that may warrant further investigation. These included reproductive symptoms, including irregular menstrual cycles and unexpected bleeding, and temperature-related issues, such as chills, hot flashes, and feeling unusually cold.
Additionally, fatigue emerged as a frequently discussed symptom, despite being less prominent in clinical trial data.
Commonly recognized side effects, such as gastrointestinal symptoms, were also widely reported, lending credibility to the approach.
“Some of the side effects we found, like nausea, are well known, and that shows the method is picking up a real signal,” said senior author Sharath Chandra Guntuku in a press release. “The underreported symptoms are leads that came from patients themselves, unprompted, and clinicians could potentially pay attention to them.”
The study highlights a growing field, sometimes referred to as computational social listening, where AI tools analyze large volumes of online content to identify health trends.
Historically, this approach has been limited by the difficulty of interpreting how individuals describe symptoms in their own words.
Mapping these descriptions to standardized medical terminology can be time consuming. However, recent advances in large language models make it possible to process such data at scale, enabling researchers to analyze patterns across hundreds of thousands of posts more efficiently and consistently.
Surprising GLP-1 side effects
Commenting on which newly identified symptoms were the most surprising, Neil Sehgal, ME, MS, the study’s first author and a doctoral student advised by Guntuku and Lyle Ungar, PhD, told Medical News Today:
“There were a few interesting findings. The reproductive symptoms are one of the most interesting signals. Nearly 4% of users who reported side effects mentioned menstrual changes, things like irregular cycles, heavy bleeding, or intermenstrual bleeding, and these symptoms aren’t prominently featured in the current prescribing information.”
“When you consider that we’re looking at a mixed-gender sample on Reddit, and that Reddit skews male, the true rate among women taking these medications could potentially be higher,” Sehgal said.
“Fatigue was also the second most commonly reported symptom overall, but has met relatively few reporting thresholds in existing trials. This gap between what patients are self-reporting online and what gets captured in trials is really what motivated this whole line of work,” he added.
Clinical trials remain the gold standard for identifying drug safety issues. However, they may not always capture the full range of patient experiences.
Typically, trials are designed to detect the most serious or common adverse effects. As such, they may overlook symptoms that individuals find bothersome but less dangerous, or that occur less frequently.
“Clinical trials generally identify the most dangerous side effects of drugs,” adds co-author Ungar. “But they can fail to find what symptoms patients are most concerned about; even though social media is not necessarily representative, a large collection of posts may reflect additional concerns.”
Still, the researchers emphasize that these findings do not prove causation. They highlight that the study only shows that users reported these symptoms while taking the medications, not that the GLP-1 drugs directly caused them.
The research team suggests there may be plausible biological reasons behind some of the reported symptoms.
GLP-1 drugs are thought to act on the hypothalamus, a region of the brain that regulates hormones, body temperature, and energy balance. Therefore, this could potentially explain reports of menstrual problems and temperature fluctuations.
However, more rigorous, controlled studies are necessary to determine whether GLP-1 medications are truly responsible.
The study authors caution that Reddit users are not representative of the general population. The platform tends to skew younger, more male, and more U.S.-based than the broader population of people taking GLP-1 medications. As a result, the findings may not reflect the experiences of all individuals.
“The most important challenge is understanding what we’ve actually found. There are a few things that make that tricky,” Sehgal told MNT.
“First, Reddit users are not a representative sample of everyone taking these medications. If you had a good experience, you’re less likely to be writing about it online. So we’re almost certainly capturing a skewed slice of the full picture.”
— Neil Sehgal, ME, MS
“Second, we can’t say anything about causality. When someone reports a menstrual irregularity while taking one of these drugs, we don’t know if the drug caused it,” Sehgal continued.
“The way I’d summarize all of this is that our findings are best thought of as signals worth investigating further, not conclusions. We’re pointing in directions that we think deserve more rigorous follow-up, not closing the case on anything,” he concluded.
Additionally, one of the study authors reports receiving an investigator-initiated grant, on behalf of the University of Pennsylvania, from Novo Nordisk and receiving consulting fees from Currax Pharmaceuticals, LLC.
The research team hopes their findings will encourage clinicians and regulators to pay closer attention to patient-reported experiences shared online. They also hope to expand future studies beyond Reddit to include other platforms and more diverse populations, as well as non-English-language data.
The researchers conclude that AI-driven analysis of social media could serve as an early warning system, especially for medications or health products that gain widespread use quickly.
“Prior research had already established that social media can be a useful tool for drug safety monitoring,” Sehgal adds.
“Ungar actually did a study back in 2011 looking at how to identify adverse events from breast cancer drugs on online message boards, so in some ways this work builds on a tradition that goes back a long time.”
— Neil Sehgal, ME, MS
“What’s changed is the technology. Large language models have made it possible to do this kind of analysis much faster with a level of standardization that could be difficult to achieve before,” Sehgal said.
For now, it is advisable for individuals taking GLP-1 medications, such as semaglutide or tirzepatide, to discuss any new or unusual symptoms with a healthcare professional.
