Researchers used AI to analyze 400k Reddit posts, revealing GLP-1 side effects (www.seas.upenn.edu)

Researchers at the University of Pennsylvania have utilized AI to analyze over 400,000 Reddit posts, uncovering patient-reported side effects associated with GLP-1 drugs, like semaglutide and tirzepatide, that are often missing from clinical trials. The study, published in Nature Health, revealed underreported symptoms such as reproductive issues (notably menstrual irregularities) and temperature-related complaints (including chills and hot flashes) from a cohort of nearly 70,000 Reddit users. This innovative application of AI highlights how social media can provide real-time insights into patient experiences that traditional clinical methods may overlook. The significance of this research lies in its potential to reshape how clinicians and researchers approach drug side effects. While the findings do not establish causation, they suggest that further investigation is warranted given that almost 4% of users reported menstrual changes. The researchers emphasize that AI-empowered analysis, utilizing large language models, enables faster and more comprehensive data processing, allowing for timely identification of possible adverse effects as medications transition from niche to widespread use. This method could prove crucial in monitoring emerging drugs and health trends, offering a proactive approach to understanding patient experiences beyond conventional means.