Every page about GLP-1 prices eventually answers the cash question: what the pen costs, what the self-pay programmes charge, what a telehealth bundle includes. Our own cost hub does exactly that. None of it answers the question people actually ask first, which is whether somebody else will pay.
That answer turns out to be unusually clean, because it is not really an insurance question. It is a statutory one, and the statute is short enough to read.
One sentence, written in 1990
Federal law sets out a list of drug categories that public drug programmes are permitted to exclude. It is at 42 U.S.C. §1396r-8(d)(2), and its heading covers "The following drugs or classes of drugs, or their medical uses".
The first item on the list:
(A) Agents when used for anorexia, weight loss, or weight gain.
Medicare Part D adopted that list wholesale. Under 42 U.S.C. §1395w-102(e)(2)(A), a "covered part D drug" does not include drugs, classes of drugs, or their medical uses excludable under §1396r-8(d)(2) — with explicit carve-outs for smoking cessation agents, barbiturates used in epilepsy, cancer or chronic mental health disorders, and benzodiazepines. Weight-loss agents were not carved out.
This was written in 1990. Every GLP-1 discussed here was approved decades later. The exclusion was not a response to their price, their popularity or their effectiveness; they arrived into a category that had already been closed.
The four words that decide everything
The operative phrase is not "weight-loss drugs". It is "when used for".
The exclusion attaches to the use, not to the molecule. The statute says so twice — once in the heading, which covers medical uses as well as drugs, and once in the wording of the subparagraph itself. Nothing in it identifies a product. It identifies a purpose.
That single grammatical choice is why the entire coverage landscape looks the way it does, and it is what most explanations of GLP-1 coverage leave out. The same drug, the same dose, the same patient, the same prescriber can sit inside or outside the exclusion depending only on the indication the prescription is written under.
Which is why the labels have been growing
Read the approved indications for the two leading products with that sentence in mind, and they look less like clinical expansion and more like a map.
Wegovy is indicated, per its FDA label effective 2024-04-23, to reduce the risk of major adverse cardiovascular events — cardiovascular death, non-fatal myocardial infarction or non-fatal stroke — in adults with established cardiovascular disease and either obesity or overweight. Reducing heart-attack risk is not anorexia, weight loss or weight gain. The exclusion does not reach it.
Zepbound is indicated, per its label effective 2026-08-28, to treat moderate to severe obstructive sleep apnea in adults with obesity. Treating sleep apnea is not a weight-loss use either.
Both products retain their weight-management indications, which remain squarely inside the exclusion. Each simply now holds a second door.
Congress has done this before, on purpose
It would be easy to treat the manoeuvre as a loophole somebody noticed. It is not. The identical structure appears a few lines further down the same Part D provision, written deliberately for a different drug class:
Such term also does not include a drug when used for the treatment of sexual or erectile dysfunction, unless such drug were used to treat a condition, other than sexual or erectile dysfunction, for which the drug has been approved by the Food and Drug Administration.
That is sildenafil: excluded when prescribed for erectile dysfunction, payable when prescribed for pulmonary arterial hypertension. One molecule, two indications, two coverage outcomes, spelled out in statute.
The GLP-1 situation is the same architecture, arrived at without needing its own sentence. Congress has already decided, in writing, that an excluded use and an approved non-excluded use can live in the same product and be treated differently. That is not an oversight being exploited; it is the design.
What follows from it
Three consequences are worth stating plainly, because they explain most of what looks arbitrary about GLP-1 coverage.
Coverage is decided by diagnosis, not by drug. The question a plan is really asking is not "is this medicine covered" but "which indication is this prescription written under, and is the supporting diagnosis documented". Two people on identical prescriptions can receive opposite answers, and neither answer is a mistake.
The variation between states is built in. The statute says the listed categories may be excluded. That is permission, not instruction. Each state Medicaid programme decides for itself, so identical prescriptions produce different outcomes across a state line — which is why any national summary, this page included, is the wrong tool for checking an individual position. The current drug list for the specific programme is the only reliable source.
Off-label uses can hit a second exclusion. Subparagraph (B) of the same list covers "Agents when used to promote fertility". That is independent of the weight-loss exclusion and catches a use these drugs are increasingly discussed for — the territory our PCOS page covers, where no product holds an approved indication at all.
What this page cannot tell you
It cannot tell you whether you are covered. Plan terms, formularies, prior-authorisation criteria and state programme rules change continually, and commercial plans are not bound by this statute in the first place. What the statute does explain is the shape of the answer: why the denial letter cites a category rather than a price, why a cardiology or sleep-clinic diagnosis changes the outcome, and why the same pen costs one person nothing and another several hundred dollars a month.
For what that second person actually pays, the cost hub carries the current self-pay and list figures with the date each was checked, and our editorial standards set out why every price on this site is dated. Nothing here is advice about a treatment or a benefits decision; both belong with a prescriber and a plan administrator.