When Congress created the Medicare drug benefit in 2003, it listed categories that plans would not have to cover. Fertility drugs. Cosmetic and hair-growth products. Drugs for symptomatic relief of coughs and colds.
And agents used for weight loss.
At the time this was uncontroversial, because the available weight-loss drugs were marginal, some were dangerous, and obesity was widely regarded as a lifestyle matter rather than a disease. Two decades later that clause governs whether tens of millions of older Americans can access the most effective obesity treatment ever developed.
The mechanics of the exclusion
Part D is a list of what plans may cover. The excluded categories sit outside it entirely, so this is not a formulary decision a plan could reverse — it is a matter of what the programme is permitted to buy.
Two routes exist around it, and both have been used.
A different approved indication. Semaglutide for type 2 diabetes has always been coverable, because that is not weight loss. After the SELECT trial established cardiovascular benefit in people with overweight or obesity and established cardiovascular disease, that became a second route. After the 2024 sleep apnea approval for tirzepatide, a third.
Reinterpretation. There has been a live argument that a drug treating obesity — a disease recognised as such by essentially every major medical body — is not a drug “used for weight loss” in the cosmetic sense the statute intended. CMS has considered this and has not, so far, adopted it.
The law treats obesity drugs the way it treats hair restoration. That categorisation was a judgment about seriousness, and it has outlived the evidence it was based on.
Why it has not been fixed
Not because anyone thinks the clause is good policy. Because of the number.
Roughly two in five American adults meet the criteria for obesity. A very large share of Medicare beneficiaries would qualify clinically. At current US list prices, universal coverage would be one of the largest new expenditures in the programme’s history — estimates have ranged into the tens of billions annually, and the range is wide enough that both sides of the argument can cite a supporting figure.
That produces a standoff with three positions, all internally coherent:
- Cover it. Obesity is a disease, the treatment works, and refusing to pay for effective treatment of a disease affecting older adults is indefensible.
- Cannot afford it. At these prices the arithmetic breaks the programme, and cost offsets from reduced cardiovascular and diabetes spending accrue over decades, if at all.
- Fix the price first. The problem is not coverage policy but that the same drugs cost several times more in the US than in comparable countries.
What it means on the ground
A 67-year-old with a BMI of 38, hypertension and knee osteoarthritis has no covered route. The same person with an A1c of 6.6 percent — that is, with diabetes — has one. The same person after a heart attack has one.
Clinically, the first patient is the one you would most want to treat early. The system reaches her last, if at all.
This is also the single largest driver of the compounded market. A retired person on a fixed income, facing a list price above a thousand dollars a month with no coverage, is exactly the person who ends up on a website. Every policy conversation about the safety of the gray market that does not mention this is missing the cause.
2003
Year the weight-loss exclusion was written into Part D
~40%
Approximate share of US adults meeting obesity criteria
Widely cited national survey estimates
3
Approved non-obesity indications currently used as coverage routes
Diabetes, cardiovascular risk, sleep apnea
What to actually do if you are on Medicare
Find out what else you have. Not as a workaround — as medicine.
A very large number of older adults with obesity have undiagnosed obstructive sleep apnea, undiagnosed prediabetes progressing to diabetes, and undocumented cardiovascular disease. Getting properly assessed is good care in its own right, and it happens to be the only path currently open.
Then ask your Part D plan, in writing, which GLP-1 products are on formulary and under which indications. Plans differ, and the answer changes at every open enrolment.
And keep the clinical record yourself, because the one thing every one of these routes demands is documentation — dated weights, doses, blood pressure, the sleep study, the albumin result. Part D plans change annually and prescribers change more often than that; a file you own survives both. Zenday will export something you can hand to whoever is arguing your case next year.
The part that will age badly
At some point the price will fall, generics and small-molecule orals will arrive, and the arithmetic will change. When it does, the exclusion will be quietly amended and everyone will agree it was obviously wrong.
The people who spent those years unable to get treated will still have spent them.