Compounded GLP-1s and the Adverse Event Reports

A pharmacovigilance study of FDA adverse event reports from 2018 to 2024 found that compounded GLP-1 drugs were reported more often than the approved versions for abdominal pain, nausea, diarrhea, gallbladder inflammation and suicidality, with more hospitalizations and more errors.^[1] Before that means anything, you need to know what a reporting rate is — because it is not an event rate.

What a FAERS study can and cannot do

The FDA Adverse Event Reporting System collects reports that people volunteer. Nobody is obliged to file one, and nobody counts how many people took the drug in the first place. That second point is the fatal one for interpreting comparisons: without a denominator, you cannot convert reports into a rate.

Now consider what happened to compounded GLP-1s across exactly this window. They were the subject of sustained news coverage, FDA statements, state pharmacy board actions and a national conversation about shortages. Attention drives reporting — clinicians and patients file reports about the things they have been told to worry about.

More reports about the thing everyone was told to report about is not, by itself, evidence of more harm.

The authors state the limit plainly: the study was not designed to show a cause-and-effect relationship.^[1] That is not boilerplate here, it is the governing fact.

The finding that survives the critique

One result is far harder to explain away, and it is the one worth carrying: prescribing and preparation errors were reported more frequently for compounded products.^[1]

That is not a subjective symptom whose reporting could be inflated by expectation. It is a discrete category of mistake, and there is an obvious mechanism specific to compounding: a compounded product usually arrives as a vial at a concentration the pharmacy chose, and the dose has to be produced with a syringe. A pen meters the dose; a vial asks a person to.

It also matches something entirely independent — the FDA’s own reports of dosing errors with compounded semaglutide, some requiring hospitalization, from patients and clinicians miscalculating. We covered that in is the vial you received what the label says. Two different data sources pointing at the same failure mode is a stronger signal than either alone.

About the suicidality item

Suicidality appears in the list of more frequently reported effects, and it needs the same reading as everything else here — more so, not less. Spontaneous reporting is the weakest form of evidence on any psychiatric outcome, and it is the most susceptible to amplification by news coverage, which this question received a great deal of.

The randomized and regulatory picture is covered in what happened with the suicidality signal, and a large Swedish cohort published this year points the opposite way on mental health outcomes. A FAERS signal does not overturn either, and it is not nothing either — it is a reason for the monitoring the authors call for.

What this does not say

• Not that compounded GLP-1s cause more harm. No rate was measured and no causal design was used.
• Not that compounded products are equivalent either. Absence of proof runs in both directions, and the quality and dosing variability the authors describe is real.
• Not a comparison of specific pharmacies. “Compounded” covers everything from a well-run 503B outsourcing facility to the fraudulent labels FDA has documented.
• Not a reason to stop a treatment you are on without talking to whoever prescribed it.