What 37,000 Tirzepatide Side Effect Reports to the FDA Reveal — Including Sex-Based Differences

Pharmacovigilance study mining the FDA Adverse Event Reporting System (FAERS) database from Q2 2022 through Q1 2024. Disproportionality analyses used four validated algorithms (ROR, PRR, BCPNN, MGPS) to identify statistically significant adverse event signals. Only events flagged by all four algorithms were considered significant. Sex-specific subgroup analyses were performed.

With millions of people now using tirzepatide (Mounjaro/Zepbound), real-world safety data from spontaneous reporting systems like FAERS captures adverse events that clinical trials — which are controlled and time-limited — may miss. The identification of starvation ketoacidosis as an unexpected signal is clinically important, and the sex-based differences in side effect patterns could inform personalized monitoring strategies.

As GLP-1 drugs become some of the most prescribed medications worldwide, post-market safety surveillance becomes critical. Clinical trials tested tirzepatide in thousands of carefully selected patients, but real-world use involves millions of diverse individuals — including those with conditions that would have excluded them from trials. The starvation ketoacidosis signal is particularly relevant because it may occur in patients who dramatically reduce food intake on the drug, a known behavioral pattern with GLP-1 agonists.

FAERS is a spontaneous reporting system subject to reporting bias — serious events are overreported while mild events are underreported. Reports don't prove causation; many patients have comorbidities and take other medications. Off-label use as a 'top 5 event' reflects reporting patterns rather than a true adverse event. The database cannot calculate true incidence rates (no denominator of total users). Duplicate reports may inflate numbers.