The FDA Received 990 Reports About Compounded Semaglutide—But Social Media Is Reading the Number Wrong
The concise answer
As of May 31, 2026, the U.S. Food and Drug Administration says it had received 990 adverse-event reports associated with compounded semaglutide and more than 730 associated with compounded tirzepatide. Those are real FDA figures. But they do not mean the agency proved that these products caused 990 or 730 injuries.
An adverse-event report records a medical occurrence associated in time with a product. It is a safety signal for investigation—not a confirmed finding of cause and effect, not a count of unique proven injuries, and not an adverse-event rate. The FDA explicitly says it is not always possible to determine whether the drug directly caused the reported event or whether other factors contributed.
What did the FDA actually report?
On its page addressing unapproved GLP-1 drugs used for weight loss, FDA states that, through May 31, 2026, it had received:
- 990 reports of adverse events associated with compounded semaglutide; and
- more than 730 reports associated with compounded tirzepatide.
The word “associated” matters. The same FDA section immediately adds that direct causation cannot always be determined and that other factors may have contributed.
FDA also provides two pieces of context that can seem contradictory at first. Adverse events from compounded products are probably underreported because many state-licensed pharmacies are not federally required to submit them. At the same time, the reports that do reach FDA do not automatically establish that the compounded product caused the event.
Both statements can be true. A surveillance system can miss events while still containing reports that are incomplete, duplicated, unverified or affected by factors other than the named product.
What is an adverse event?
FDA describes an adverse event as an unexpected medical occurrence in a person who received a medicine. It does not necessarily have a causal relationship with that treatment.
That is different from an adverse drug reaction. An adverse drug reaction is a harmful, unintended response determined to have been caused by the drug. Social-media posts often collapse these two categories, turning “reported after or during use” into “proven to have been caused by.”
The distinction is not semantic fine print. It is fundamental to pharmacovigilance.
Why does the FDA collect unconfirmed reports?
Waiting for proof before accepting a report would make a rapid-warning system much less useful. FDA encourages patients and health professionals to report suspected problems even when they are unsure that the product was responsible.
Individual reports can help regulators identify unusual patterns, medication errors, product-quality problems or events that were too uncommon to emerge in premarket trials. FDA can then investigate a signal using case review, medical records, laboratory information, inspections, epidemiologic studies, controlled trials or other evidence.
As FDA explains in its MedWatch materials, reports alone rarely confirm causation, but they can provide an early signal that a problem may be occurring.
Why 990 reports cannot tell us the event rate
An event rate requires a meaningful denominator: how many people used the relevant product, how much exposure occurred and over what period. The FDA figure supplies a report count, not a verified exposure denominator.
It also does not answer several other necessary questions:
- Were multiple submissions filed for the same case?
- Was the compounded product correctly identified?
- What other medicines or medical conditions were present?
- Was the event consistent with known effects of FDA-approved GLP-1 drugs?
- Was there a dosing, labeling, storage or measurement error?
- Did a product-quality defect contribute?
- How complete was the clinical information?
FDA says adverse-event reporting data cannot, by themselves, establish incidence, compare the safety of products or show that a product caused an event. Report counts should not be converted into percentages, “odds” or comparative rankings without valid exposure and case-validation data.
What the reports do tell us
Avoiding overstatement does not mean ignoring the reports. The figures show that FDA has received a substantial body of postmarket safety information involving compounded semaglutide and tirzepatide. That information is relevant enough for the agency to monitor, evaluate and discuss publicly.
FDA identifies several specific concerns elsewhere on the same page:
- dosing errors involving compounded injectable semaglutide, including errors that led some patients to seek medical attention or hospitalization;
- prescriptions for doses beyond the approved product label, including larger or more frequent doses and faster titration;
- compounded products arriving warm or with inadequate refrigeration;
- fraudulent labels and products attributed to pharmacies that did not make them; and
- semaglutide salt forms that differ from the active ingredient used in FDA-approved products.
These are documented areas of concern. They still should not be presented as though every one of the 990 reports belongs to one category or has a confirmed common cause. FDA does not make that claim.
Are compounded semaglutide and tirzepatide FDA approved?
No. Compounded drugs are not FDA approved, and the agency does not review them for safety, effectiveness and quality before marketing in the same way it reviews approved products.
Compounding can be appropriate in particular circumstances. FDA says a compounded drug may be appropriate when a patient’s medical need cannot be met by an FDA-approved drug or when the approved drug is not commercially available. The agency recommends that compounded drugs be used only for patients whose needs cannot be met by an approved drug and that prescriptions be filled at state-licensed pharmacies.
That regulatory distinction is important, but it is separate from the causation question. “Not FDA approved” does not turn every reported event into a proven product-caused injury; likewise, uncertainty about individual reports does not make product-quality and dosing concerns disappear.
The four most common ways the number gets misread
1. “FDA confirmed 990 injuries”
Incorrect. FDA received 990 reports associated with compounded semaglutide. It did not state that all 990 were confirmed as caused by the product.
2. “990 reports means the product has a known X% risk”
Unsupported. The published count does not include the validated exposure denominator needed to calculate a rate.
3. “The reports prove compounded semaglutide is more dangerous than an approved product”
Not from these numbers alone. Spontaneous-report counts cannot support a fair product-to-product comparison because exposure, reporting behavior, duplicates, publicity and case quality can differ.
4. “Because causation is uncertain, the reports mean nothing”
Also incorrect. Spontaneous reports are important signal-detection evidence. They can identify patterns that justify investigation, warnings or regulatory action even though the raw reports do not prove causation individually.
A better way to describe the data
A careful summary would say:
“Through May 31, 2026, FDA received 990 adverse-event reports associated with compounded semaglutide and more than 730 associated with compounded tirzepatide. The reports are safety signals, not confirmed counts of product-caused injuries, and they cannot establish an event rate. FDA also says compounded-product events are likely underreported.”
That wording preserves both sides of the evidence. It neither minimizes the reports nor exaggerates what they prove.
What researchers and science communicators should check
Before sharing an adverse-event number, ask:
- Is the source FDA itself or a screenshot of someone else’s summary?
- Does the figure refer to reports, cases, confirmed reactions or another category?
- What is the cutoff date?
- Does the source use “associated with” or explicitly establish causation?
- Is there an exposure denominator?
- Does the source warn about underreporting, duplicates or incomplete information?
- Are different products, formulations or routes being combined?
These checks take less time than correcting a viral claim later.
Frequently asked questions
Did compounded semaglutide cause all 990 reported events?
FDA does not say that. The agency says it is not always possible to determine whether the adverse event directly resulted from the drug or whether other factors contributed.
Are adverse events from compounded GLP-1 products underreported?
Probably. FDA notes that federal law does not require many state-licensed pharmacies that are not outsourcing facilities to submit adverse events to the agency.
Can the FDA numbers be used to compare semaglutide with tirzepatide?
Not by simply comparing 990 with more than 730. Raw report totals do not control for exposure, reporting differences, duplicates, case completeness or other factors.
Does “compounded” mean FDA approved?
No. Compounded drugs are not FDA approved and do not undergo FDA’s premarket review for safety, effectiveness and quality.
The bottom line
The 990 and 730-plus figures deserve attention, but precision matters. They are counts of reports associated with compounded products—not proven causal injuries and not event rates. The responsible reading is neither “the number proves the products caused everything” nor “the reports can be ignored.” They are surveillance signals that require context, investigation and careful communication.
For related background, see JD BioWorks resources on testing methods and evidence boundaries, scientific sourcing and editorial standards and FDA’s 2026 generic peptide guidance.
Sources
1. U.S. Food and Drug Administration. FDA’s Concerns with Unapproved GLP-1 Drugs Used for Weight Loss.
2. U.S. Food and Drug Administration. Information About Reporting Adverse Events to FDA’s MedWatch Program.
3. U.S. Food and Drug Administration. Understanding CDER’s Postmarket Safety Surveillance Programs and Public Data.
4. U.S. Food and Drug Administration. FDA Adverse Event Reporting System Reporting and Review.
Medical and research-use disclaimer: This article provides general educational information about adverse-event reporting and regulatory evidence. It does not provide individualized medical advice. Patients should discuss medication questions or changes with a licensed health professional. JD BioWorks materials are intended for research use only and are not for human or veterinary use.
Related Research Materials
JD-2TR
JD-2TR research material in 10 mg, 20 mg, 40 mg, and 60 mg quantities. Select a quantity to…
View Research Material