UK MHRA Yellow Card Biobank data count 153 adverse reaction reports linked to fatalities for liraglutide, semaglutide and tirzepatide, part of a broader tally of more than 150 suspected deaths. Gastrointestinal, nervous system and psychiatric reports run into the tens of thousands, yet remain a…
The UK medicines regulator's spontaneous reporting database contains 153 adverse reaction reports linked to fatalities in which liraglutide, semaglutide or tirzepatide were suspected, according to data from the Medicines and Healthcare Products Regulatory Agency MHRA Yellow Card Biobank. The tally underpins a broader count of more than 150 deaths to which GLP-1 weight-loss drugs may be linked, a figure that aggregates suspected reactions across the incretin-based therapies most used for obesity in the UK.
The Yellow Card Biobank records suspected adverse reactions submitted by clinicians, patients and manufacturers. It holds thousands of adverse reaction reports for GLP-1 drugs, spanning gastrointestinal, nervous system and psychiatric disorders. The MHRA stresses that a report is not proof that a drug caused the outcome; the deaths could have unrelated causes, and spontaneous reports cannot by themselves establish causation. The numbers should not be read as a list of established side effects.
Despite that caveat, the volume of reports, and their concentration in the drug classes most widely prescribed for weight loss, makes the dataset a useful early-warning signal. It arrives as use of the drugs expands rapidly in the UK, with nearly 7% of the population currently using or having used GLP-1 drugs, according to The Food Foundation.
The fatality-linked reports break down unevenly across the three drugs. Semaglutide accounts for 37 suspected deaths, liraglutide for 18, and tirzepatide for 98. The database also records 28 deaths after exenatide use, 3 for dulaglutide, and 2 for lixisenatide, though these older GLP-1 products are not commonly used for weight loss.
Adverse reaction reports for non-fatal outcomes are far larger. Gastrointestinal disorder reports total nearly 28,000 for semaglutide, more than 63,000 for tirzepatide, and 1,095 for liraglutide. Nervous system disorder reports approach nearly 5,000 for semaglutide and exceed 10,000 for tirzepatide. Psychiatric disorder reports number more than 2,200 for tirzepatide, more than 1,300 for semaglutide, and 138 for liraglutide.
Semaglutide is sold by Novo Nordisk as Ozempic and Wegovy; liraglutide is sold by Novo Nordisk as Victoza and Saxenda; tirzepatide is sold by Eli Lilly as Mounjaro. Both companies responded to the figures. A Novo Nordisk spokesperson said: "That is why medicines should be prescribed by, and used under the supervision of, a healthcare professional, who can consider the potential benefits and risks for each individual." An Eli Lilly spokesperson said: "Patient safety is Lilly’s top priority. We take reports regarding patient safety seriously and actively monitor, evaluate, and report safety information for all our medicines."
The raw counts are difficult to interpret without a denominator. A University College London estimating study calculated that 1.6 million people in England, Scotland and Wales used Wegovy or Mounjaro for weight loss in 2024 and early 2025. The Food Foundation puts the share of the UK population currently using or having used GLP-1 drugs at nearly 7%.
Against that exposure base, the reported numbers represent only a fraction of UK GLP-1 drug users, suggesting that suspected adverse reactions are very rare in relative terms, even if the absolute counts attract attention. What the ratio does not do, however, is produce a reliable incidence rate. Spontaneous reporting captures only a subset of events, and reporting behaviour varies by drug, by time on the market, and by publicity around the product.
The UCL estimate itself is an approximation. The source describes it as an estimating study with no published methodology, sample frame or confidence intervals, so the 1.6 million figure should be treated as a plausible order of magnitude rather than a precise count. It does, however, establish the essential point: exposure in the UK population is now large enough that even rare outcomes will generate hundreds of reports in a national pharmacovigilance system.
The MHRA operates the Yellow Card Scheme as the UK's system for collecting suspected adverse drug reactions. It is a passive reporting system: anyone can submit a Yellow Card, including patients, and the agency compiles submissions in its pharmacovigilance database, which now includes the Biobank resource referenced in the data. The scheme covers all medicines and vaccines used in the UK.
No regulatory action is described in connection with the figures. The MHRA's published position is that reported reactions are not proven to be related to the drug and should not be interpreted as a list of known side effects. The agency's role is to monitor signals and, when warranted, update product information or refer questions to the Commission on Human Medicines.
The legal basis of the scheme is the duty on healthcare professionals and marketing authorisation holders to report suspected adverse reactions; manufacturers carry additional obligations for expedited reporting and periodic safety update reports. Spontaneous reporting systems are strongest at generating hypotheses about rare or unexpected harms and weakest at quantifying risk, because they lack a denominator and are subject to selective reporting. That distinction is central to reading the 153 figure correctly: it is a signal count, not an effect estimate.
The drugs in question are peptide therapeutics that mimic the action of endogenous glucagon-like peptide-1. Semaglutide is a GLP-1 receptor agonist; tirzepatide is a dual agonist of the GLP-1 and glucose-dependent insulinotropic polypeptide GIP receptors; liraglutide is a shorter-acting GLP-1 receptor agonist. By activating GLP-1 receptors in the pancreas, gut and brain, they potentiate glucose-dependent insulin secretion, slow gastric emptying, and reduce appetite through central pathways.
The mechanism of action helps explain the symptom profile in the Yellow Card data. Delayed gastric emptying and altered gut motility are direct pharmacological effects of GLP-1 receptor activation, which is why gastrointestinal disorders dominate the adverse reaction counts. The nervous system and psychiatric reports are harder to attribute. GLP-1 receptors are expressed in the brain, and recent trials have investigated effects on reward, mood and cognition, but the pathways linking peripheral peptide exposure to psychiatric symptoms remain poorly defined.
Weight loss itself is mediated partly through gastrointestinal tolerability. Nausea, vomiting and diarrhoea are well-established side effects in clinical trials, and prescribers typically titrate doses slowly to reduce them. The question posed by the Yellow Card data is whether the rarer, more serious outcomes in the fatality-linked reports share the same mechanism or reflect something else: aspiration secondary to severe nausea, pancreatitis, cardiovascular events in a population with substantial underlying disease burden, or coincidental deaths unrelated to the drug.
Peptide Atlas’s trial registry gives some measure of how much clinical evidence is accumulating around the two most prominent compounds. Semaglutide has 668 registered clinical trials on file. The phase fields in the dataset record 4 Phase 2 trials, 4 Phase 4 trials and 1 Phase 3 trial, with 10 trials listed as recruiting. Tirzepatide has 251 registered trials, with 5 Phase 2, 2 Phase 4 and 1 Phase 3, and 10 recruiting.
The recruiting trials illustrate the direction of the field. Semaglutide trials include NCT07586150, the LIFETRAIN study of personalized pharmaco-lifestyle interventions for severe mental illness; NCT07430332, a Phase 2 trial in Stage 1 type 1 diabetes; NCT07614412, the SHIELD-T1D trial testing Shingrix and a GLP-1 agonist for beta-cell preservation in recent-onset type 1 diabetes; NCT07462663, the SHAPE-ENDO pilot trial of pre-surgical optimization in endometrial cancer; NCT07027969, a trial of metabolic surgery for atrial fibrillation elimination; and NCT06977438, a GLP-1 plus lifestyle study in childhood obesity. Three of these are Phase 4, meaning they track outcomes closer to real-world use.
Tirzepatide’s registered trials follow a similar pattern of expanding beyond diabetes and obesity. They include NCT06180616 in concurrent type 1 diabetes and overweight or obesity; NCT07468552 for cannabis use disorder; the shared NCT07027969 trial in atrial fibrillation; NCT07609160 combining dual agonist therapy with structured exercise in sarcopenia; NCT06732245 testing tirzepatide with the adjunct NA-931; and NCT07630454 in atrial fibrillation recurrence after catheter ablation in patients with obesity and heart failure with preserved ejection fraction.
The peer-reviewed literature indexed on Peptide Atlas is extensive for both molecules: 197 PubMed papers for semaglutide and 188 for tirzepatide. Recent publications track efficacy and safety questions alike. For semaglutide these include a meta-analysis of long-term safety and renal outcomes in non-diabetic obesity with chronic kidney disease or hypertension PMID 42340790 , a systematic review of weight-lowering drugs and female fertility PMID 42307450 , a JAMA Psychiatry trial on effort-based decision-making in major depressive disorder PMID 42054055 , the STRIDE trial in peripheral artery disease PMID 41780559 , and a scoping review of retinal vascular events PMID 42348481 . For tirzepatide, recent work includes a multicenter comparison with SGLT2 inhibitors in metabolic dysfunction-associated steatotic liver disease PMID 42397506 , a case report of starvation-type euglycemic ketoacidosis after unsupervised use PMID 42381258 , a review of molecular mechanisms PMID 42387035 , a retrospective cohort in type 1 diabetes PMID 42387290 , and a Mayo Clinic Proceedings study comparing tirzepatide and semaglutide for obesity in real-world settings PMID 42383938 .
This registry and literature picture matters for interpreting pharmacovigilance signals. The trials now recruiting will generate prospective safety data in populations, such as people with severe mental illness, type 1 diabetes and post-surgical patients, who were underrepresented in the original obesity trials. That evidence, not spontaneous reports, will be what settles whether the Yellow Card signals reflect true drug-attributable risk.
For clinicians, the practical implication is straightforward: the Yellow Card counts reinforce existing prescribing guidance. GLP-1 drugs should be prescribed by, and used under the supervision of, healthcare professionals who can weigh individual benefit and risk, as the Novo Nordisk spokesperson said. Monitoring should include gastrointestinal tolerability, review of psychiatric symptoms where appropriate, and attention to the safety surveillance obligations Eli Lilly described as central to its pharmacovigilance practice. The reports do not change the benefit-risk calculus for appropriately selected patients, but they do argue against casual or unsupervised use, a point underscored by published case reports such as the ketoacidosis case after unsupervised tirzepatide PMID 42381258 .
For the peptide research community, the episode is a reminder of the scale on which incretin-based peptides are now used. Semaglutide and tirzepatide are among the most widely used peptide drugs in history, which means rare events will surface in spontaneous reporting even if the true attributable risk is very low. The task for researchers is to convert these signals into measurable hypotheses using designs that…
Peptides referenced: Semaglutide, Tirzepatide, Liraglutide, Exenatide, Dulaglutide, Lixisenatide, Glucagon, GLP-1.
Related reading: Tirzepatide tied to 32% lower MACE risk in real-world diabetes cohort, Gedeon Richter and Adalvo Expand Partnership to Tirzepatide Therapies, Meta-Analyses Question Long-Term Value of GLP-1 Weight-Loss Drugs, Lilly Lifts 2026 Guidance as Tirzepatide Sales Surge, Peptide Rivals Split.