Research & Regulatory News

Which Drug Signals Where, and What That Is Worth

UKPWritten and published by UK Peptides4 min readUpdated 3 cited sources

The short answer

Shokr and colleagues analysed FAERS reports for three weight-management agents across gastrointestinal, renal and pancreatic outcomes. Semaglutide showed the strongest gastrointestinal signals, liraglutide showed the strongest renal and pancreatic ones, and tirzepatide showed weaker or absent signals across most of the outcomes examined.

Key facts

Publication
Pharmaceuticals (Basel) 2026;19(1):136 (PMID 41599734)
Method
Disproportionality analysis of FAERS, PRR and ROR with 95% CIs
Window
From each drug's approval for weight management
Strongest GI signal
Semaglutide, PRR 3.97, ROR 14.21 (n=12,321)
Liraglutide GI
PRR 2.76, ROR 5.01 (n=5,972)
Tirzepatide GI
PRR 1.64, ROR 1.90 (n=4,056)
Acute pancreatitis, liraglutide
PRR 18.9, ROR 19.4
Diabetic ketoacidosis, semaglutide
PRR 5.86, ROR 5.9
Declared conflicts
None

What was compared

The analysis looked at gastrointestinal, renal and pancreatic adverse events for GLP-1 and dual receptor agonists, using FAERS data from each drug's approval for weight management, with signals identified by proportional reporting ratio and reporting odds ratio with 95% confidence intervals. The comparator is the rest of the database rather than another drug, which matters for how the numbers should be read: each figure describes how a drug-event pair stands against the overall background pattern, not against another agent in the class.

The gastrointestinal result and its internal oddity

Semaglutide showed the strongest gastrointestinal signals, with 12,321 reports representing 1.65% of its total and a PRR of 3.97 alongside an ROR of 14.21. Liraglutide followed at 5,972 reports, 0.45%, PRR 2.76, ROR 5.01. Tirzepatide was lowest on the ratios at 4,056 reports, PRR 1.64, ROR 1.90, though its 3.48% proportion is the highest of the three, which is a useful reminder that the proportion and the disproportionality measure answer different questions. The gap between semaglutide's PRR of 3.97 and its ROR of 14.21 is unusually wide. PRR and ROR converge when the event is a small share of the drug's reports and diverge when it is a large one, so a spread that size is itself information about the composition of the reports rather than about risk.

Research material referenced

Retatrutide 10mg, third-party HPLC tested

Buy Retatrutide · £59.99

Renal and pancreatic: liraglutide leads, which is the counterintuitive part

For renal and pancreatic outcomes the ordering inverts. Liraglutide showed the highest overall signal (PRR 4.91, ROR 5.35), driven by acute pancreatitis (PRR 18.9, ROR 19.4) and pancreatic carcinoma (PRR 18.6, ROR 19.5). Semaglutide showed stronger associations with diabetic ketoacidosis (PRR 5.86, ROR 5.9) and acute kidney injury (PRR 1.25, ROR 1.25). Tirzepatide showed weaker or absent signals across most outcomes. The liraglutide pancreatic figures are the largest in the analysis by a wide margin, and they belong to the oldest and least-used of the three agents. That is the configuration in which reporting artefacts are most likely, because a drug marketed since 2010 has accumulated years of literature associating GLP-1 agonism with pancreatitis, and notoriety bias increases reporting of an association once it is expected.

Reading tirzepatide's weaker signals correctly

Tirzepatide comes out cleanest across most of this analysis, and the temptation is to read that as a safety advantage. It may be one. It is also what a drug approved in 2022 looks like relative to drugs approved in 2010 and 2017, in a system where reporting accumulates with time on market, with literature attention, and with the number of clinicians who have learned to associate a drug with an event. A signal that is absent may be absent because the association does not exist, or because not enough time has passed for it to surface. Disproportionality cannot distinguish the two.

Where this fits against the drug-versus-drug analyses

Compare this with Gandhi and colleagues, who set tirzepatide directly against semaglutide for acute kidney injury and found a reporting odds ratio of 0.44 (95% CI 0.38 to 0.50). Both analyses agree that semaglutide's kidney-injury reporting exceeds tirzepatide's. They differ enormously in magnitude (1.25 against the whole database, 0.44 in the reciprocal direction against semaglutide specifically) because they are measuring against different reference groups. Neither is wrong and neither is a risk estimate.

What the authors themselves conclude

The paper's own conclusion is appropriately narrow: that semaglutide was most associated with gastrointestinal events and liraglutide with renal and pancreatic signals, that novel associations with ketoacidosis and acute kidney injury warrant clinical vigilance, and that the findings should be cautiously interpreted given surveillance limitations, with large-scale real-world studies needed to confirm safety profiles. The authors declare no conflicts of interest. That framing, in which a signal is a prompt for a properly designed study and not a result, is the one that should survive into any summary of this work, and it is usually the first thing to be dropped.

Frequently asked questions

Does this mean liraglutide causes pancreatic cancer?

No. A pancreatic carcinoma PRR of 18.6 in a spontaneous reporting database means reports pairing the two appear far more often than the background pattern predicts. Where an association has been discussed in the literature for over a decade, clinicians who see the event are more likely to attribute it to the drug and report it, which produces this pattern in the absence of any causal effect. Cohort studies with denominators are what settle the question.

Why does tirzepatide look safest here?

Partly because its signals are weaker in this data, and partly because it is the newest of the three. Time on market drives reporting independently of risk, so a comparison across drugs of different vintages is systematically biased in favour of the newer one.

Are these drugs kidney-protective or kidney-harmful?

Randomised outcome trials in this class have shown kidney benefits in defined populations, and spontaneous reports have raised acute kidney injury as a signal. Both can be true: a chronic benefit and an acute risk in particular circumstances (dehydration from vomiting or diarrhoea being the obvious mechanism) are not contradictory.

Should any of this change what someone takes?

That is a conversation for a prescribing clinician, not a reference article. These are licensed medicines; nothing supplied on this site is an alternative to any of them, and no article here should be used to inform a decision about a prescription.

Continue researching

Related guides and reference material.

Primary sources & clinical trials

Published papers, trial registrations, regulator pages and company announcements. Links open the original.

UKP

Written and published by UK Peptides, the trading name of AURYXRESEARCH LTD (company 17388727, registered in England and Wales).

Questions or a correction? Email hello@ukpeptides.org or call +44 7426 987848.

Corrections log

More Research & Regulatory News articles

Research use only. The information above is provided for scientific and educational reference. Compounds referenced are not approved for human use and are supplied for in vitro research or reference-material purposes only. No efficacy, safety, or therapeutic claims are made.