The Limits of a Reporting Baseline: Sex Differences in Adverse Event Reporting for Tumour Necrosis Factor Blockers, GLP-1 Receptor Agonists and Proton Pump Inhibitors Across 1,753,642 Reports
DOI:
https://doi.org/10.64813/ejmr.2026.089Keywords:
Pharmacovigilance, Tumor Necrosis Factor, Inhibitors, Agonists, Proton Pump Inhibitors, AlopeciaAbstract
We have previously shown, across seven drug classes, that male adverse event reports are systematically more serious and that the reporting odds ratio (ROR) for death clusters near 0.63 irrespective of pharmacology. Whether this baseline is universal, and what would break it, was unknown. Case/non-case disproportionality analysis of the FDA Adverse Event Reporting System via openFDA database. All reports listing a tumour necrosis factor (TNF) blocker, GLP-1 receptor agonist or proton pump inhibitor (PPI) with a recorded patient sex were included. RORs for women versus men were calculated within each class for 34 pre-specified Preferred Terms, with Benjamini-Hochberg control of the false discovery rate. 1,753,642 reports were analysed. Male reports were more often serious in all three classes, making ten of ten. The death ROR was 0.59 for TNF blockers and 0.63 for PPIs, within the established range, but 0.38 for GLP-1 agonists, 7.4 SD below the nine-class mean of 0.632. GLP-1 reports were also far less often serious than any other class (27.3% in women). Alopecia gave the largest effects in the programme and was near-identical in two unrelated classes: ROR 8.34 (95% CI 7.74-8.98) for TNF blockers and 8.15 (7.13-9.30) for GLP-1 agonists. PPI outcomes divided between gastrointestinal symptoms in women and renal injury in men. The severity baseline is robust but not universal. It fails where the reporting population differs markedly, which bounds its generality and clarifies its cause. The near-identical alopecia signal across unrelated classes suggests differential reporting of salient events.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Nabila Ghaffar, Kinza Tahir

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.