Compounded GLP-1 is not categorically dangerous, but the bar for 'safe enough' is higher than most patients know

Published Jul 14, 2025Updated Jul 14, 2025Last reviewed Jul 14, 2025

The access problem that started this argument

Roughly 40% of adults in the United States meet BMI criteria for obesity pharmacotherapy. Branded semaglutide and tirzepatide carry list prices between $900 and $1,400 per month, and most commercial insurance coverage remains patchy at best. The shortage period from 2022 through late 2024 made access worse: FDA placed both semaglutide injections and tirzepatide injections on the Drug Shortage List, a designation that under Section 503A of the Federal Food, Drug, and Cosmetic Act explicitly permits licensed compounding pharmacies to compound copies of shortage-listed drugs for individual patients.

That statutory permission is not a loophole. It is the designed safety valve in the federal compounding framework, and dismissing every compounded GLP-1 formulation as categorically unsafe ignores the law's own structure.

The shortage designations changed significantly between October 2024 and early 2026. FDA removed tirzepatide injections from the shortage list in late 2024, and semaglutide injections followed in early 2025, triggering a formal wind-down period for compounders. That timeline matters for everything that follows, because the legal justification for compounding under shortage authority has now narrowed considerably.

The evidence base: what we actually know about compounded GLP-1 risk

The FDA's concern with compounded GLP-1 is not theoretical. Between 2023 and 2025, the agency issued multiple warning letters and public safety communications directed at specific compounders, citing products marketed with unsubstantiated claims, incorrect dosing instructions, and in some cases, ingredient substitution. The agency's documented findings included:

Compounded products labeled as "semaglutide" that contained semaglutide sodium or semaglutide acetate rather than the base peptide used in Ozempic and Wegovy, raising bioavailability and safety questions that have not been resolved in published trials.
Products sold without a valid patient-specific prescription, a direct violation of 503A requirements.
Marketing language implying therapeutic equivalence to FDA-approved products, which compounded drugs are explicitly prohibited from claiming.

The FDA stated in its 2023 communication that it had "received reports of adverse events" associated with these products, though it has not published a systematic adverse event count specific to compounded GLP-1 as of this writing. That absence of a clean denominator is itself a data quality problem we should be honest about.

On the branded drug side, the evidence base is deep. The SUSTAIN-6 trial showed semaglutide 0.5 mg and 1.0 mg reduced MACE by 26% versus placebo in high-cardiovascular-risk patients with type 2 diabetes (HR 0.74 to 95% CI 0.58, 0.95). The SURMOUNT-1 trial showed tirzepatide 15 mg produced mean weight reduction of 20.9% versus 3.1% for placebo over 72 weeks. These outcomes rest entirely on the pharmacokinetic profile of the specific molecular form studied. There are no published RCT data on semaglutide sodium or semaglutide acetate in humans.

Dr. Peter Marks, director of FDA's Center for Biologics Evaluation, said in a 2024 public statement that "the agency cannot assure the safety, effectiveness, or quality of compounded GLP-1 products," a position that is technically accurate but applies with full force only to the non-compliant products FDA was actively sanctioning.

The 503A statutory framework sets specific conditions under which a licensed pharmacy may compound: the drug must be compounded by a licensed pharmacist or physician, pursuant to a valid prescription for an individually identified patient, using bulk drug substances that comply with USP monograph standards or appear on FDA's approved bulk drug substance list. A pharmacy operating within those conditions occupies a different risk category than the direct-to-consumer online dispensaries that dominated FDA's warning letter activity.

Where the consensus falls short

We find two failure modes in the current public discourse, and they run in opposite directions.

The overcorrection from professional societies. Several obesity medicine and endocrinology societies have issued statements that, read plainly, advise against all compounded GLP-1 use. The Obesity Medicine Association's 2024 guidance called compounded products "unregulated" and recommended patients pursue branded drugs exclusively. That framing is imprecise. A 503A-compliant pharmacy is regulated, just not through the same pre-approval pathway as a new drug application. Calling it unregulated conflates "not FDA-approved" with "operating outside legal oversight," which is not the same thing. The guidance also does not acknowledge the price reality: for a patient paying $1,200 per month out of pocket versus $150 for a compounded formulation through a compliant pharmacy, telling them to "pursue branded" is functionally telling them to go without treatment.

The cardiovascular and metabolic benefits of GLP-1 agonists are not modest. A 2023 SELECT trial analysis showed semaglutide 2.4 mg reduced major cardiovascular events by 20% in non-diabetic patients with obesity and established cardiovascular disease (HR 0.80 to 95% CI 0.72, 0.90). Denying access to this class on the grounds of compounding concerns carries its own mortality risk calculus that the society statements do not address.

The undercorrection from compounder marketing. The opposite failure is the direct-to-consumer telehealth and compounding market, where brands routinely imply therapeutic equivalence, post before-and-after content without disclosing that outcomes derive from branded trial data, and obscure the molecular form of the active ingredient. When a patient asks "is this the same as Ozempic," the honest answer requires knowing whether the product contains semaglutide base, and whether the pharmacy has third-party certificate-of-analysis documentation to support that claim. Most patients never ask. Most telehealth intake flows do not prompt the question.

The actual risk in the compounded GLP-1 space is not evenly distributed. It concentrates in three places: wrong molecular form (semaglutide salts versus base peptide), incorrect reconstitution of lyophilized powder products, and absence of physician oversight for dose titration. A patient self-titrating compounded tirzepatide from a vial with unclear concentration labeling, without a clinician tracking nausea, vomiting, or dehydration, is genuinely at elevated risk compared to a patient on Zepbound with monthly telehealth follow-up. A patient receiving a properly compounded semaglutide base formulation from a 503A pharmacy, titrated by a physician who knows their baseline GI history and renal function, is at a risk level that we believe is defensible to offer.

The shortage-status shift changes the legal picture, not the access problem

With both semaglutide and tirzepatide now removed from FDA's shortage list, the primary statutory basis for 503A compounding of these drugs has narrowed. FDA has stated that compounders must wind down shortage-based production following these removals. 503B outsourcing facilities face a harder stop than 503A pharmacies, which retain some capacity to compound for individual patients based on documented clinical need, including documented allergy to excipients in branded formulations or documented inability to access branded products at a price the patient can sustain.

That last category, economic inaccessibility, is not explicitly enumerated in FDA's 503A shortage authority. We are extending judgment here: our position is that a physician documenting a patient's inability to access branded GLP-1 therapy at any covered price, combined with a clinical indication, provides a reasonable basis for a 503A pharmacy to compound. This is not settled law. FDA has not confirmed it. But the 503A text does not prohibit it, and the clinical case for continuity of care for an established patient is stronger than for initiating new therapy.

Our position

The HealthRX Medical Team holds the following position, stated plainly:

Compounded semaglutide and tirzepatide from a compliant 503A pharmacy, prescribed by a physician with clinical oversight of titration and monitoring, is a reasonable option for patients who cannot access or afford branded GLP-1 therapy. This position rests on the 503A statutory framework, the documented benefits of the drug class, and the real mortality risk of untreated obesity and type 2 diabetes.

The position is conditional. Our minimum bar for "compliant" means:

The pharmacy holds a current state board of pharmacy license and has not appeared on FDA's list of warning letter recipients.
The pharmacy provides a certificate of analysis from an independent third-party laboratory confirming that the active ingredient is semaglutide free base (or tirzepatide) at labeled concentration, not a salt form, not a peptide fragment.
The patient has a valid, patient-specific prescription, not a standing order or protocol fill.
A licensed prescriber reviews the patient's renal function, GI history, and personal or family history of medullary thyroid carcinoma or MEN2 before initiating.
Dose titration follows a schedule with at least one clinical touchpoint per titration step.

Most patients who ask us about compounded GLP-1 cannot confirm even the first criterion. That is the actual risk, and it is a patient-education and system-design failure, not an inherent property of compounded drugs.

We do not think the professional society position, as currently written, serves patients who cannot pay $1,200 per month. We also do not think the telehealth compounder position, which implies equivalent safety to branded products without disclosing molecular form or analytical testing, is honest.

What would change our mind

Two categories of evidence would update this position.

First, published pharmacokinetic data showing that compounded semaglutide base from a 503A pharmacy produces meaningfully different plasma exposure curves than branded semaglutide at equivalent labeled doses. There are theoretical reasons to expect this could vary by formulation and reconstitution method, but we have no human PK data to evaluate. If such data showed clinically significant under-exposure or erratic absorption, the risk-benefit calculation would shift.

Second, a systematic FDA adverse event analysis specific to 503A-compliant compounded GLP-1, separated from the documented adverse event reports associated with clearly non-compliant products. The current aggregate framing obscures whether harm is concentrated in the bad-actor segment or distributed across the compliant segment. That distinction is the one we need most and currently lack.

Until that data exists, we are making a judgment call. We are saying so explicitly.